Blog

QC Scope
Nicolas Stifani posted on Jan 22, 2025


user Accès   time Tarifs   user Services    search-small Instruments   time Réservation   info Blog email Contact


QC Scope is a plugin for FIJI designed to process  microscope images for Quality Control purposes.
Written by Nicolas Stifani from the Centre d'Innovation Biomédicale de la Faculté de Médecine de l'Université de Montréal.

Contents


Installation

  1.  for your operating system
  3. Unzip FIJI.app to your favorite location on your computer (usually your Desktop
  4. Start FIJI (aka FIJI.app, ImageJ-win.exe)
  5. Drag and drop QC_Scope.jar into the FIJI toolbar
  6. Click Save
  7. Click OK
  8. Close FIJI

  9. Reopen FIJI

After installation, QC Scope will be available in the FIJI menu under Plugins>QC Scope.

QC Scope Toolbar

The QC Scope floating toolbar is available for rapid and convenient access. To load it, click the QC Scope Toolbar in the FIJI menu under Plugins>QC Scope>QC Scope Toolbar.

QC Scope Auto-start

Additionally and for an even more convenient usage, it is possible to have the QC Scope Toolbar automatically loaded when FIJI is started. Click the Auto-start button or select the QC Scope Toolbar Auto-start in the FIJI menu under Plugins>QC Scope>QC Scope Toolbar Auto-start.

QC Scope Demo Images

For testing purposes, you will find  that can be used to process and generate results.

Additional information

QC Scope file format compatibility

For now, QC Scope only processes images with the following extensions  ".tif", ".tiff", ".jpg", ".jpeg", ".png", ".czi", ".nd2", ".lif", ".lsm", ".ome.tif", ".ome.tiff"

QC Scope file writing strategy

QC Scope never overwrites files. It checks for the existence of files and increment a number until it can safely write the output file without erasing data.

Filename Convention

If you are well organized when saving and naming your files, QC Scope will help you.
It specifically check for the presence of "_" in the filename and split it into Filename Variables.
For example if you label your image as Date_Microscope-A_Objective-X_Filter-Y_Condition-001.TIF QC Scope will provide in the output one column with each variable: Date, Microscope, Objective, Filter, Condition. Then you can use a Pivot Table in Excel to quickly access your data.

Field Uniformity

Usage

  1. Open FIJI.
  2. Launch the QC Scope Toolbar by navigating to Plugins>QC Scope>QC Scope Toolbar.
  3. Click on Uniformity.
    1. If one or more images are already opened QC Scope will process them.
    2. If no image is open, QC Scope will prompt to select a folder and process all images withing the folder (and subfolders)
  4. QC Scope will try to read the Metadata from the first image and pre-process all the channels with default or the last used processing settings
  5. It will display the metadata, the initial results and the processing options in a dialog
  6. Check the metadata, the results and the binned image. If it looks good you can click OK. 

  7. QC Scope will save files in a folder named Output on your desktop:

    1. At least 2 CSV files:

      1. Field Uniformity_All-Data_Merged.csv gathers all the measured parameters
        Field Uniformity_All-Data_Merged.csv

      2. Field Uniformity_Essential-Data_Merged.csv gathers only the essential information
        Field Uniformity_Essential-Data_Merged.csv

    2. Optionally, if Save Individual Files is selected QC Scope will also save:
      1. 1 CSV file per image ImageName_Uniformity-Data.csv with one row per channel containing all the measured parameters
      2. 1 TIF file per channel for every processed image showing the binned (Iso-density or Iso-Intensity) map with the Reference Center indicated as an overlay

 

Example of iso-density image with the coordinates of the reference center.

 Settings

  • Microscope Settings read from the image metadata or from the preferences: Objective Magnification (character), NA (Numeric>0), and Immersion media 
  • Image Settings: Pixel Width (Numeric>0), Height (Numeric>0), Depth (Voxel) (Numeric>0), Unit (character). QC Scope uses standard space unit (nm, um, mm, cm, m, in, pixels) and will try to convert the entered value into on of those.
  • Channel Settings: For each channel: Name (character) and Emission Wavelength in nm ((Numeric>0)) as well as the source of the values
  • Processing Settings:
    • Binning Method:
      • Iso-Density (preferred): This method divides the image into 10 bins of equal Nb of Pixels. Nb Pixel Per Bin = (Width x Height) / 10 and assign a new pixel value of 25 for all the Nb Pixel Per Bin darkest pixels, 50 for the next darkest Nb Pixel Per Bin etc... until 250 for the brightest Nb Pixel Per Bin pixels.
      • Iso-Intensity: This method divides the image into 10 bins of equal bin intensities. Bin Width = (Max - Min) / 10 and assign a new pixel value of 25 for all the pixels with an intensity between Min and Min + Bin With, 50 for intensities between Min + Bin Width and Min + 2 x Bin Width etc... until 250 for intensities between Min + 9 x Bin Width and Max.
    • Gaussian Blur: Apply a Gaussian blur with the given Gaussian Blur Sigma before processing each channel
    • Channel: The selected channel will be processed with the entered processing parameters and displayed as part of the testing process to define optimal processing parameters.
    • Batch Mode: If activated, QC Scope will re-use the settings without displaying the dialog unless metadata differs
    • Save Individual Files: For each image, QC Scope will save the individual processed images (1 per channel) and a CSV file with all measured parameters.
    • Prolix Mode: Display all the QC Scope actions in the log

    • Test Processing: When selected the QC Scope Field Uniformity Dialog will keep appearing. This is useful to test the Processing Settings. When you are satisfied you can uncheck Test processing to proceed to the next image.

 QC Scope Field Uniformity dialog

Metrics

 Field Uniformity measures how evenly illumination is distributed and how centered is the illumination across the field of view. The following metrics provide different perspectives on uniformity:

  1. Uniformity: Evaluates the overall evenness of illumination. Since a single value can't capture all variations, QC Scope computes several uniformity metrics for a comprehensive assessment.

    1. Standard Deviation (GV): The standard deviation of pixel intensities measures how much individual pixel values deviate from the mean intensity of the image. A low standard deviation indicates that the pixel intensities are close to the mean, suggesting high uniformity and minimal variation across the image. The standard deviation is in the same unit as the pixel intensity values in the image (Grey Value).

    2. Uniformity Std (%): Calculated as ratio between the intensities of the darkest and brightest pixels. This metric evaluates image uniformity by comparing the minimum pixel intensity to the maximum pixel intensity in the image. It ranges between 0 and 1. A value of 1 indicates perfect uniformity, where the minimum and maximum intensities are equal. This simple metric provides a quick indication of how evenly distributed the pixel intensities are. It highlights whether there are extreme intensity variations, such as bright or dark spots, which can suggest uneven illumination or artifacts. While easy to compute, this ratio is sensitive to outliers. A single very bright or very dark pixel can skew the measurement. It also does not take in account the distribution of pixel intensities and instead focuses on the minimun and maximum.
      In MetroloJ_QC, the Field Uniformity function applies a Gaussian blur before calculation. This reduce the impact of potential outliers. QC Scope uses the unmodified image to compute the Uniformity. For this reason QC Scope will always provide lower Uniformity Std values compared to MetroloJ QC as it is representing the worse case scenario.

      $$\text{Uniformity}_{Std} = \frac{\text{Min}}{\text{Max}}$$

       

    3. Uniformity Percentile (%): This metric evaluates image uniformity by analyzing the average intensities of the darkest and brightest pixels within a chosen percentile (e.g., 5%) instead of relying on extreme values. It uses the mean intensities of these percentile subsets to provide a more robust and representative measure of uniformity.

      $$\text{Uniformity}_{Percentile} = \left(1 - \frac{\text{Avg}_{95} - \text{Avg}_{5}}{\text{Avg}_{95} + \text{Avg}_{5}} \right)$$

    4. Coefficient of Variation (CV): The Coefficient of Variation is a standard statistical measure used to quantify the relative variation or dispersion of pixel intensities in an image. It is calculated as the ratio of the Standard Deviation to the Mean of the pixel intensities. The CV provides a dimensionless measure of variation, making it easier to compare variability across images with different intensity scales or ranges. It reflects the extent of variability in relation to the average intensity. A lower CV indicates higher uniformity, with pixel intensities tightly clustered around the mean. The CV adjusts for differences in intensity scale, enabling fair comparisons between images. It is particularly useful in imaging workflows where intensity levels may vary across samples. CV is an excellent metric as long as the mean is not close to 0.

      $$\text{CV} = \frac{\text{Standard Deviation}}{\text{Mean}}$$

    5. UniformityCV (%): This metric is derived from the Coefficient of Variation (CV) and is tailored for microscopy quality control, where users aim to capture highly uniform images. It assumes low variation (CV≪1) and provides an intuitive measure of uniformity expressed as a percentage. A perfectly uniform image, with CV = 0 results in a uniformity of 100%, while increasing variability lowers the value. A uniformity of 0% corresponds to CV = 1, where the standard deviation equals the mean intensity, indicating significant intensity variation. This metric is particularly useful in controlled imaging conditions typical of microscopy quality control. It assumes low CV value. UniformityCV offers a simple robust and user-friendly way to assess image uniformity. 

      $$\text{Uniformity}_{CV} = (1 - \text{CV})$$
  2. Centering Accuracy (%): Assesses how close the brightest region is to the image center. Defined as:
    $$\text{Centering Accuracy} = 1 - \frac{2}{\sqrt{\text{Image Width}^2 + \text{Image Height}^2}} \times \sqrt{(\text{X}_{\text{Ref Pix}} - \frac{\text{Image Width}}{2})^2 + (\text{Y}_{\text{Ref Pix}} - \frac{\text{Image Height}}{2})^2}$$
  3. Field Illumination Index: Combine both UniformityCV and the Centering Accuracy in a single measure. A field Illuminatin index of 100% indicates a perfectly uniform and centered image.
    $$\text{Field Illumination Index} = \text{Weigth}\times \text{Uniformity}_{CV} + (1 - \text{Weight})\times \text{Centering Accuracy}$$

QC Scope Field Uniformity Metrics (in bold the variables included in Essential Data)

Key OrderField NameData ExampleData TypeDescription
1Filename10x_Quad_Exp-01.cziStringName of the processed image
2Channel Nb4IntegerNumber of the Channel from 1 to n
3Channel NameDAPIStringName of the Channel
4Channel Wavelength EM (nm)465IntegerChannel Emission Wavelength
5Objective Magnification10xStringObjective Magnification
6Objective NA0.25FloatObjective Numerical Aperture
7Objective Immersion MediaAirStringObjective Immerion Media
8Gaussian Blur AppliedTRUEBooleanIf Gaussian Blur was applied
9Gaussian Sigma10IntegerSigma of the Gaussian Blur
10Binning MethodIso-DensityStringBinning Method used
11Batch ModeTRUEBooleanBoolean key to process images in batch mode (no Dialog)
12Save Individual FilesFALSEBooleanBoolean key to save individual files (1 csv data per file with 1 row per channel, 1 tif binned image par channel)
13Prolix ModeFALSEBooleanBoolean key display detailed plugin actions in the log
14Image Min Intensity856IntegerRaw Image Minimum of Pixel Intensities
15Image Max Intensity1080IntegerRaw Image Maximum of Pixel Intensities
16Image Mean Intensity959.1FloatRaw Image Mean of Pixel Intensities
17Image Standard Deviation Intensity24.3FloatRaw Image Standard Deviation of Pixel Intensities
18Image Median Intensity959IntegerRaw Image Median Pixel Intensity
19Image Mode Intensity110IntegerRaw Image Mode Pixel Intensity
20Image Width (pixels)1388IntegerImage Width in pixels
21Image Height (pixels)1040IntegerImage Height in pixels
22Image Bit Depth16IntegerImage Bit Depth
23Pixel Width (um)0.645FloatImage pixel width (unit/px)
24Pixel Height (um)0.645FloatImage pixel height (unit/px)
25Pixel Depth (um)1FloatImage voxel depth (unit/voxel)
26Space UnitmicronStringRaw Image Space Unit
27Space Unit StandardumStringStandardize Space Unit (nm, um, cm, m)
28Calibration StatusTRUEBooleanBoolean key displaying the calibration status
29Standard Deviation (GV)24.3FloatRaw Image Standard Deviation of Pixel Intensities
30Uniformity Standard (%)0.793FloatUniformity as calculated by MetroloJ_QC. Uniformity_Standard = (Min / Max)
31Uniformity Percentile (%)0.958FloatUniformity calculated with the average of the 5% and 95% pixel intensities.
Uniformity_Percentile = (1 - (Avg_Intensity95 - Avg_Intensity5) / (Avg_Intensity95 + Avg_Intensity5) )
32Coefficient of Variation0.0253FloatCoefficient of variation. CV = (Std_Dev / Mean)
33Uniformity CV based0.975FloatUniformity calculated from the Coefficient of variation. Uniformity_CV = (1 - CV) 
34X Center (pixels)694IntegerCoordinate in pixel of the center of the Image (Ideal centering). Image Width (pixels) / 2
35Y Center (pixels)520IntegerCoordinate in pixel of the center of the Image (Ideal centering). Image Height (pixels) / 2
36X Ref (pixels)230.4FloatCoordinate in pixels of the centroid of the largest particule identified in the last bin. Used to caculate the Centering Accuracy
37Y Ref (pixels)876.4FloatCoordinate in pixels of the centroid of the largest particule identified in the last bin. Used to caculate the Centering Accuracy
38X Ref (um)148.6FloatCoordinate in scaled unit of the centroid of the largest particule identified in the last bin.
39Y Ref (um)565.3FloatCoordinate in scaled unit of the centroid of the largest particule identified in the last bin.
40Centering Accuracy (%)0.32FloatCentering Accuracy = 1 - (2 / sqrt(Image Width**2 + Image Height**2)) * sqrt ( (X_Ref_Pix - Image Width/2)**2 + (Y_Ref_Pix - Image Height/2)**2)
41
Field Illumination Index (%)0.884
Float
Field Illumination Index  = Weight * UniformityCV+ (1 - Weight) * Centering Accuracy
Weight = 0.5

Key Metrics

  1. UniformityCV (%): 100% perfectly uniform. 0% or less if the Standard Deviation of pixel intensities in the image is equal or higher than the mean.
  2. Centering Accuracy (%): 100% illumination is perfectly centered on the image. Ratio of the distance of the centroid of the Indicates how the 10% brightest pixels to the center image 
  3. Field Illumination Index: Combine both UniformityCV and Centering Accuracy in a single Index. 100% indicates a perfectly uniform and centered image.

Results

  • Convert the Field Uniformity_Essential-Data_Merged.csv created by QC Scope Field Uniformity function into a .xlsx
  • Summarize the data with a pivot table
  • Use the provided spreadsheet template Field Uniformity_Template.xlsx
  • Enter your measurement into the highlighted cells of the template to visualize your results. The cells in grey are automatically computed.
  • Plot the Field illumination Index for each objective and filter combination

The 2x, 20x and 63x objectives have a field illumination index above 70% for all filters indicating an acceptable uniformity and centering accuracy. The 10x objective field illumination index is below 70% indicating a poor uniformity and/or Centering accuracy requiring further investigations.

  • Plot the Uniformity and Centering Accuracy for each objective and filter combination

All objectives show a good Uniformity. The 2x, 10x and 20x objectives have a centering accuracy below 70% indicating a poor illumination centering.

Iso-density map showing the localization of the centroid of the highest intensity bin for 10x and 63x objectives.

Field illumination Index for each objective and filter combination

Conclusion

Centering accuracy could be improved on the 2x and 10x objectives. This is not too much of a concern since Uniformity is well above 70%. 


Channel Alignment

Usage

  1. Open FIJI.
  2. Launch the QC Scope Toolbar by navigating to Plugins>QC Scope>QC Scope Toolbar.
  3. Click on Ch Alignment.
    1. If one or more images are already opened QC Scope will process them.
    2. If no image is open, QC Scope will prompt to select a folder and process all images withing the folder (and subfolders)
  4. QC Scope will try to read the Metadata from the first image and pre-process all the channels with the default or the last used processing settings
  5. It will display the metadata, the initial results and the processing options in a dialog
  6. Check the metadata and the results. If it looks good you can click OK. You can also modify the the metadata and processing options

  7. QC Scope will save files in a folder named Output on your desktop:

    1. At least 2 CSV files are saved :

      1. Channel-Alignment_All-Data_Merged.csv gathers all the measured parameters
        Channel Alignment_All-Data_Merged.csv

      2. Channel Alignment_Essential-Data_Merged.csv gathers only the essential information
        Channel Alignment_Essential-Data_Merged.csv

    2. Optionally, if Save Individual Files is selected QC Scope will also save:
      1. 1 CSV file per image gathering all the measured parameters for each channel pair (Image-Name_Channel-Alignment_All-Data.csv)

Settings

  • Microscope Settings: Read from the image metadata or from the preferences: Objective Magnification (character), NA (Numeric>0), and Immersion media 
  • Image Settings: Read from the image Calibration. Pixel Width (Numeric>0), Height (Numeric>0), Depth (Voxel) (Numeric>0), Unit (character). QC Scope uses standard space unit (nm, um, mm, cm, m, in, pixels) and will try to convert the entered value into one of those.
  • Channel Settings: Read from the image metadata or from the preferences, for each channel: Name (character) and Emission Wavelength in nm ((Numeric>0)) as well as displaying the source of the values
  • Processing Settings:
    • Detection Method:
      • Log Detector: Laplacian of Gaussian (LoG) detector. From Trackmate: The LoG detector is the best detector for Gaussian-like particles in the presence of noise. It is based on applying a LoG filter on the image and looking for local maxima. The Laplacian of Gaussian result is obtained by summing the second order spatial derivatives of the gaussian- filtered image, and normalizing for scale. Reference: Trackmate manual (page 52). 
      • Dog Detector: Difference of Gaussian (DoG) detector. From Trackmate: This detector is based on the Difference of Gaussian (DoG) filter. It approximates the Laplacian of Gaussian (LoG) filter with the aim at offering better speed. It is commonly used when applying a collection of DoG filters tuned to a wide range of scales. Reference: Trackmate manual (page 52). 
    • Threshold: This is a quality threshold. Only spots above the quality threshold will be detected. The quality value is larger for : Bright spots, spots which diameter is close to the specified diameter.
    • Diameter: This is the diameter for the detection of spots with the indicated unit. Usually this is the diameter of the beads used for the channel alignment.
    • Median Filtering: If activated QC Scope will ask Trackmate to use a Median Filtering. Median filtering: will apply a 3 x 3 median filter prior to any processing.
      This can help dealing with images that have a marked salt and pepper noise which generates spurious spots. Reference: Trackmate manual (page 11). 
    • Channel: The selected channel will be processed with the entered processing parameters and displayed as part of the testing process to define optimal processing parameters. Selecting an another channel will automatically enable Test Processing.
    • Batch Mode: If activated, QC Scope will re-use the settings without displaying the dialog unless metadata differs or the detection fails.
    • Save Individual Files: If activated, QC Scope will save an additional CSV File per image named Image-Name_Channel-Alignment_All-Data.csv gathering all the measured parameters for each channel pair (1 row per channel pair = Nb Channel2 rows per file. 
    • Prolix Mode: Display all the QC Scope actions in the log. If used in combination with Save Individual Files, it will save the original Trackmate spot table (1 per channel)
    • Subpixel Precision: If activated QC Scope will ask Trackmate to use a subpixel localization for the detection of spots

    • Test Processing: When selected the QC Scope Dialog will keep appearing. This is useful to test the Processing Settings. When you are satisfied you can uncheck Test processing to proceed to the next image. Changing the selected channel automatically select the Test Processing option.

  • Pre-detection results:
    • Nb of Detected spots: Indicate the number of detected spot per channel. To proceed, exactly one spot must be detected for every channel.
    • Max Quality: Indicate the maximum quality of all detected spots for each channel. The maximum quality can then be used to adjust the threshold value

Metrics

Channel Alignment measures how a single bead appears in different channels. QC Scope Channel Alignment detects one spot per channel and retrieve the xyz coordinates using TrackMate plugin More info about TrackMate. It then computes the distance between two spots and compare it to a reference spot. The position of the reference spot depends on the actual xyz resolution of the system. A Co-localization ratio is then calculated:

$$\text{Colocalisation Ratio} = \frac{\text{Distance}_{\text{Spot Ch1-Spot Ch2}}}{\text{Distance}_{\text{Spot Ch1-Spot Reference}}}$$
  • A Colocalization Ratio above 1 indicates that the Spot detected in Channel 2 is further away from the spot detected in Channel 1 than the reference spot. In other words the spot in channel 2 is further away than the resolution of the system. This indicates that Channel Alignment is not good and corrections should be performed using the pixel shift table.
  • A Colocalization Ratio below 1 indicates that the spot detected in Channel 2 is closer from the spot detected in channel 1 than the reference spot. In other words the distance between the spots detected in channel 1 and 2 is smaller than the resolution of the system. This ratio indicates a good Channel Alignment.

QC Scope Channel Alignment Metrics (in bold the variables included in Essential Data)

Key OrderField NameData ExampleData TypeDescription
1Filename20x_Quad_Exp-01.cziStringName of the processed image
2Objective Magnification20xStringObjective Magnification
3Objective NA0.5FloatObjective Numerical Aperture
4Objective Immersion MediaAirStringObjective Immerion Media
5Immersion Media Refractive Index1.0003FloatObjective Immerion Media Refractive Index
6Detection MethodDog DetectorStringMethod for the detection Log Dectector or Dog Detector
7Spot Diameter (um)4FloatSpot Diameter used for the detection
8Threshold Value20FloatQuality threhsold value used for the detection
9Subpixel LocalizationTrueBooleanSubpixel Localization used for the detection
10Median FilteringFalseBooleanMedian Filtering used for the detection
11Batch ModeTrueBooleanBoolean key to process images in batch mode (no Dialog)
12Save Individual FilesTrueBooleanBoolean key to save individual files (1 csv data per file with 1 row per channel, 1 tif binned image par channel)
13Prolix ModeFalseBooleanBoolean key display detailed plugin actions in the log
14Image Width (pixels)100IntegerImage Width in pixels
15Image Height (pixels)100IntegerImage Height in pixels
16Image Bit Depth16IntegerImage Bit Depth
17Pixel Width (um/px)0.3225FloatImage pixel width (unit/px)
18Pixel Height (um/px)0.3225FloatImage pixel height (unit/px)
19Pixel Depth (um/px)1.24FloatImage voxel depth (unit/voxel)
20Space UnitmicronStringRaw Image Space Unit
21Space Unit StandardumStringStandardize Space Unit (nm, um, cm, m)
22 Time UnitsecStringTime Unit
23Calibration StatusTrueBooleanBoolean key displaying the calibration status
24Channel 11IntegerNumber of the Channel 1 from 1 to n
25Name Channel 1Cy5StringName of the Channel 1
26EM Wavelength Channel 1 (nm)673IntegerChannel 1 Emission Wavelength
27Nb Detected Spots Ch11IntegerNb of Detected Spots for Channel 1
28Spot ID Ch12055IntegerSpot ID for Channel 1
29Spot Quality Ch1132FloatSpot Quality for Channel 1
30X Ch1 (um)16.005FloatCoordinate of the center of the detected spot for Channel 1 along the X axis in the indicated unit
31Y Ch1 (um)16.311FloatCoordinate of the center of the detected spot for Channel 1 along the Y axis in the indicated unit
32Z Ch1 (um)35.171FloatCoordinate of the center of the detected spot for Channel 1 along the Z axis in the indicated unit
33T Ch1 (sec)0FloatTime of the detected spot for Channel 1 in the indicated unit
34Frame Ch10IntegerFrame of the detected spot for Channel 1
35Radius Ch1 (um)2FloatRadius of the detected spot for Channel 1 in the indicated unit
36Visibility Ch1TrueBooleanSpot visibility for Channel 1
37Channel 23IntegerNumber of the Channel 2 from 1 to n
38Name Channel 2GFPStringName of the Channel 2
39EM Wavelength Channel 2 (nm)509IntegerChannel 2 Emission Wavelength
40Nb Detected Spots Ch21IntegerNb of Detected Spots for Channel 2
41Spot ID Ch22075IntegerSpot ID for Channel 2
42Spot Quality Ch2132FloatSpot Quality for Channel 2
43X Ch2 (um)16.05FloatCoordinate of the center of the detected spot for Channel 2 along the X axis in the indicated unit
44Y Ch2 (um)16.243FloatCoordinate of the center of the detected spot for Channel 2 along the Y axis in the indicated unit
45Z Ch2 (um)35.718FloatCoordinate of the center of the detected spot for Channel 2 along the Z axis in the indicated unit
46T Ch2 (sec)0FloatTime of the detected spot for Channel 2 in the indicated unit
47Frame Ch20IntegerFrame of the detected spot for Channel 2
48Radius Ch22FloatRadius of the detected spot for Channel 2 in the indicated unit
49Visibility Ch2TrueBooleanSpot visibility for Channel 2
50Channel PairCy5 x GFPStringCombination of Channel 1 x Channel 2 Names
51X Shift (um)0.046FloatDifference between the coordinates of Channel 2 and Channel 1 along the X axis in the indicated unit
52Y Shift (um)-0.068FloatDifference between the coordinates of Channel 2 and Channel 1 along the Y axis in the indicated unit
53Z Shift (um)0.547FloatDifference between the coordinates of Channel 2 and Channel 1 along the Z axis in the indicated unit
54X Shift (pixels)0.1FloatDifference between the coordinates of Channel 2 and Channel 1 along the X axis in pixels
55Y Shift (pixels)-0.2FloatDifference between the coordinates of Channel 2 and Channel 1 along the Y axis in pixels
56Z Shift (pixels)0.4FloatDifference between the coordinates of Channel 2 and Channel 1 along the Z axis in pixels
57Distance Lateral (um)0.082FloatLateral distance between the spot detected in Channel 2 and Channel 1 in the indicated unit
58Distance Axial (um)0.547FloatAxial distance between the spot detected in Channel 2 and Channel 1 in the indicated unit
59Distance 3D (um)0.553Float3D distance between the spot detected in Channel 2 and Channel 1 in the indicated unit
60Conversion Factor1000FloatConversion factor to convert Emission Wavelength from nm to the same unit than the calibrated image
61EMWavelength Unit Ch1 (um)0.673FloatEmission Wavelength in the indicated unit for Channel 1
62EMWavelength Unit Ch2 (um)0.509FloatEmission Wavelength in the indicated unit for Channel 2
63Nyquist Pixel Size Lateral Ch1 (um)0.337FloatNyquist Lateral pixel Size for Channel 1 in the indicated unit. Formula : Nyquist_Pixel_Size_Lateral = EMWavelength_Unit / (4 * Objective_NA)
64Nyquist Pixel Size Axial Ch1 (um)2.513FloatNyquist pixel Axial Size for Channel 1 in the indicated unit. Formula : Nyquist_Pixel_Size_Axial = EMWavelength_Unit / (2 * Refractive_Index * (1-cos(Theta))) with Theta = asin(Objective_NA / float(Refractive_Index))
65Nyquist Ratio Lateral Ch11FloatNyquist Lateral Ratio for Channel 1 in the indicated unit. Formula : Nyquist_Ratio_Lateral = Pixel_Width / Nyquist_Pixel_Size_Lateral. A ratio above 1 indicate the Pixel Size does not meet the Nyquist sampling criteria (Pixel too big)
66Nyquist Ratio Axial Ch10.5FloatNyquist Axial Ratio for Channel 1 in the indicated unit. Formula : Nyquist_Ratio_Axial = Pixel_Depth / Nyquist_Pixel_Size_Axial. A ratio above 1 indicate the Pixel Depth does not meet the Nyquist sampling criteria (Z Stack steps too big)
67Nyquist Pixel Size Lateral Ch2 (um)0.255FloatNyquist pixel Size for Channel 1 in the indicated unit. Formula : Nyquist_Pixel_Size_Lateral = EMWavelength_Unit / (4 * Objective_NA)
68Nyquist Pixel Size Axial Ch2 (um)1.9FloatNyquist Lateral Ratio for Channel 2 in the indicated unit. Formula : Nyquist_Ratio_Lateral = Pixel_Width / Nyquist_Pixel_Size_Lateral. A ratio above 1 indicate the Pixel Size does not meet the Nyquist sampling criteria (Pixel too big)
69Nyquist Ratio Lateral Ch21.3FloatNyquist Axial Ratio for Channel 2 in the indicated unit. Formula : Nyquist_Ratio_Axial = Pixel_Depth / Nyquist_Pixel_Size_Axial. A ratio above 1 indicate the Pixel Depth does not meet the Nyquist sampling criteria (Z Stack steps too big)
70Nyquist Ratio Axial Ch20.7FloatNyquist pixel Size for Channel 2 in the indicated unit. Formula : Nyquist_Pixel_Size_Lateral = EMWavelength_Unit / (4 * Objective_NA)
71Resolution Lateral Theoretical Ch1 (um)0.686FloatTheoretical Lateral Resolution for Channel 1 in the indicated unit. Formula: Resolution_Lateral_Theoretical = (0.51 * EMWavelength_Unit) / (Objective_NA)
72Resolution Axial Theoretical Ch1 (um)4.766FloatTheoretical Axial Resolution for Channel 1 in the indicated unit. Formula: Resolution_Axial_Theoretical = (1.77 * Refractive_Index * EMWavelength_Unit) / (Objective_NA ** 2)
73Resolution Lateral Practical Ch1 (um)0.686FloatPractical Lateral Resolution for Channel 1 in the indicated unit. If the Nyquist Lateral Ratio is above 1 the Practical Lateral Resolution = Theoretical Lateral Resolution x Nyquist Lateral Ratio otherwise the Practical Lateral Resolution = Theoretical Lateral Resolution
74Resolution Axial Practical Ch1 (um)4.766FloatPractical Axial Resolution for Channel 1 in the indicated unit. If the Nyquist Axial Ratio is above 1 the Practical Axial Resolution = Theoretical Axial Resolution x Nyquist Axial Ratio otherwise the Practical Axial Resolution = Theoretical Axial Resolution
75Resolution Lateral Theoretical Ch2 (um)0.519FloatTheoretical Lateral Resolution for Channel 2 in the indicated unit. Formula: Resolution_Lateral_Theoretical = (0.51 * EMWavelength_Unit) / (Objective_NA)
76Resolution Axial Theoretical Ch2 (um)3.605FloatTheoretical Axial Resolution for Channel 2 in the indicated unit. Formula: Resolution_Axial_Theoretical = (1.77 * Refractive_Index * EMWavelength_Unit) / (Objective_NA ** 2)
77Resolution Lateral Practical Ch2 (um)0.658FloatPractical Lateral Resolution for Channel 2 in the indicated unit. If the Nyquist Lateral Ratio is above 1 the Practical Lateral Resolution = Theoretical Lateral Resolution x Nyquist Lateral Ratio otherwise the Practical Lateral Resolution = Theoretical Lateral Resolution
78Resolution Axial Practical Ch2 (um)3.605FloatPractical Axial Resolution for Channel 2 in the indicated unit. If the Nyquist Axial Ratio is above 1 the Practical Axial Resolution = Theoretical Axial Resolution x Nyquist Axial Ratio otherwise the Practical Axial Resolution = Theoretical Axial Resolution
79X Ref (um)16.143FloatCoordinate in indicated unit of the reference spot in the along the X axis. The reference spot is a calculated as a spot at the interesect between the Line defined by Spot Channel 1 and Spot Channel 2 coordinates and an ellipse centered on Spot Channel 1 with Distance Lateral Ref as a semi-minor axis and Distance Axial Ref as a semi-major axis
80Y Ref (um)16.106FloatCoordinate in indicated unit of the reference spot in the along the Y axis. The reference spot is a calculated as a spot at the interesect between the Line defined by Spot Channel 1 and Spot Channel 2 coordinates and an ellipse centered on Spot Channel 1 with Distance Lateral Ref as a semi-minor axis and Distance Axial Ref as a semi-major axis
81Z Ref (um)36.825FloatCoordinate in indicated unit of the reference spot in the along the Z axis. The reference spot is a calculated as a spot at the interesect between the Line defined by Spot Channel 1 and Spot Channel 2 coordinates and an ellipse centered on Spot Channel 1 with Distance Lateral Ref as a semi-minor axis and Distance Axial Ref as a semi-major axis
82X Ref Shift (um)0.138FloatRelative distance in the indicated unit between the reference spot and the spot detected in Channel 1 along the X axis
83Y Ref Shift (um)-0.205FloatRelative distance in the indicated unit between the reference spot and the spot detected in Channel 1 along the Y axis
84Z Ref Shift (um)1.654FloatRelative distance in the indicated unit between the reference spot and the spot detected in Channel 1 along the Z axis
85Semi Minor Axis (um)0.343FloatThis is the semi minor axis of an ellipse defining the maximum distance in the XY plan calculated by dividing the largest Practical Lateral Resolution by 2.
86Semi Major Axis (um)2.383FloatThis is the semi major axis of an ellipse defining the maximum distance in the Z plan calculated by dividing the largest Practical Axial Resolution by 2.
87Distance Lateral Ref (um)0.247FloatDistance between the reference spot and the spot detected in Channel 1 in the XY plan in the indicated unit
88Distance Axial Ref (um)1.654FloatDistance between the reference spot and the spot detected in Channel 1 in the Z plan in the indicated unit
89Distance 3D Ref (um)1.673FloatDistance between the reference spot and the spot detected in Channel 1 in the 3D in the indicated unit
90Colocalization Ratio0.3FloatRatio between the 3D Distance (Channel 1 Channel 2) and the 3D Reference Distance. A ratio above 1 indicates the spot detected in channel 2 is further than the pratical resolution of the system. Formula: Colocalization_Ratio = Distance_3D /Distance_3D_Ref

Key Metrics

  1. Colocalisation Ratio: A ratio above 1 indicates the distance between the spot detected in channel 1 and 2 is larger than the practical resolution of the system: Channel Alignment is not good and corrections should be performed using the pixel shift table.

Results

  • Convert the Channel Alignment_Essential-Data_Merged.csv created by QC Scope Channel Alignment function into a .xlsx
  • Summarize the data with a pivot table
  • Use the provided spreadsheet template Channel Alignment_Template.xlsx
  • Paste the Colocalization Ratio and the XYZ pixel shifts in the Shift Tab. The cells use conditional formatting to highlight cells with a ratio above 1.0.

These results indicates that the 63x objective requires correction for the DAPI channel. User should be informed to correct the images.

63x DAPI (Cyan) and Cy5 (Magenta) 4um bead raw image.


  63x DAPI (Cyan) and Cy5 (Magenta) 4um bead corrected for chromatic shift.

 

63x DAPI (Cyan) and Cy5 (Magenta) 4um bead  Z-Stack.

63x DAPI (Cyan) and Cy5 (Magenta) 4um bead Z-Stack corrected for chromatic shift.

Plateformes Scientifiques CIB               Biologie structurale    |    Cytométrie    |    Microscopie    |    Histologie   |    Instruments et Services

Microscopy Quality Control
Nicolas Stifani posted on Dec 12, 2024


user Accès   time Tarifs   user Services    search-small Instruments   time Réservation   info Blog email Contact




This page provides a practical guide for microscope quality control. By following the outlined steps, utilizing the provided template files, and running the included scripts, you will have everything needed to easily generate comprehensive report on your microscope's performance.

Equipment used

  • Thorlabs Power Meter (PM400) and sensor (S170C)
  • Thorlabs Fluorescent Slides (FSK5)
  • TetraSpeck™ Fluorescent Microspheres Size Kit (mounted on slide) ThermoFisher (T14792)

Software used

Excel Templates and Scripts


Please note that during quality control, you may, and likely will, encounter defects or unexpected behavior. This practical guide is not intended to assist with investigating or resolving these issues. With that said, we wish you the best of luck and are committed to providing support. Feel free to reach out to us at microscopie@cib.umontreal.ca


Illumination Warmup Kinetic

When starting light sources, they require time to reach a stable and steady state. This duration is referred to as the warm-up period. To ensure accurate performance, it is essential to record the warm-up kinetics at least once a year to precisely define this period. For a detailed exploration of illumination stability, refer to the Illumination Power, Stability, and Linearity Protocol by the QuaRep Working Group 01.

Acquisition protocol

  • Place a power meter sensor on the microscope stage.

  • Center the sensor and the objective.

  • Zero the sensor to ensure accurate readings.

  • Select the wavelength of the light source you want to monitor, using your power meter controller or associated software.

  • Turn on the light source and immediately record the power output over time until it stabilizes.

    I personally use the Thorlab Powermeter Software in Monitoring mode and record every 10 seconds for 24 hours and stops (recording) when it has been stable for 1 h.

  • Save the recorded data as a CSV file.
  • Repeat for each light source.

    Keep the light source turned on at all times. Depending on your hardware, the light source may remain continuously on or be automatically shutdown by the software when not in use.

Results

  • Use the provided spreadsheet template, Illumination_Warmup Kinetic_Template.xlsx
  • Copy and paste the data from the recorded CSV file into the highlighted cells of the template to visualize your results.
  • For each light source, plot the measured power output (in mW) against time to analyze the data.

  • Calculate the relative power using the formula: Relative Power = (Power / Max Power) for each wavelength. Then, plot the Relative Power (%) against Time to visualize the data.

We observe some variability in the power output during the first 20 min. It is especially noticeable for the 385 nm light source.

To assess the stability:

  • Define a Stability Duration Window: This is a time period (e.g., 10 minutes) during which the power output should remain stable.

  • Specify a Maximum Coefficient of Variation (CV) Threshold: This a limit acceptable for the CV (e.g., 0.01%).

  • Calculate the Coefficient of Variation (CV): CV = (Standard Deviation / Mean)

  • Plot the calculated CV over time to analyze the stability of the power output.

We observe that most of light sources stabilize quickly, within less than 10 minutes, while the 385 nm light source takes approximately 41 minutes to reach the stability threshold. The template also calculates Stability Factor (S) using the formula: S (%) = 100 × (1 - (Pmax - Pmin) / (Pmax + Pmin))

  • Report the results in a table

385nm475nm555nm630nm

Stabilisation time (Max CV 0.01% for 10 min)

41338
Stability Factor (%) Before Warmup99.7%99.9%100.0%100.0%
Stability Factor (%) After Warmup100.0%100.0%100.0%99.9%

 

Metrics

  • The Stability Factor indicates a higher stability the closer to 100% and focuses specifically on the range of values (difference between max and min) relative to their sum, providing an intuitive measure of how tightly the system's behavior stays within a defined range.
  • The Coefficient of Variation focuses on the dispersion of all data points (via the standard deviation) relative to the mean. Lower Coefficient indicates a better stability around the mean.

Conclusion

The illumination warmup time for this instrument is approximately 40 minutes. This duration is essential for ensuring accurate quantitative measurements, as the Coefficient of Variation (CV) threshold is strict, with a maximum allowable variation of 0.01% within a 10-minute window.

Illumination Maximum Power Output

This measure assesses the maximum power output of each light source, considering both the quality of the light source and the components along the light path. Over time, we expect a gradual decrease in power output due to the aging of hardware, including the light source and other optical components. These measurements will also be used to track the performance of the light sources over their lifetime (see Long-Term Illumination Stability section). For a detailed exploration of illumination properties, refer to the Illumination Power, Stability, and Linearity Protocol by the QuaRep Working Group 01.

Acquisition protocol

  • Warm up the light sources for the required duration (see previous section).
  • Place the power meter sensor on the microscope stage.
  • Center the sensor and the objective.
  • Zero the sensor to ensure accurate readings.
  • Select the wavelength of the light source you want to monitor, using your power meter controller or associated software.
  • Turn the light source on to 100% power.
  • Record the average power output over 10 seconds.
    I personally re-use the data collected during the warm-up kinetic experiment (when the power has been stable for 1h) for this purpose.
  • Repeat for each light source.

Results

These results are informative on their own but could become even more meaningfull when comapred to the manufacturer’s specifications.

  • Calculate the Relative PowerSpec using the formula: Relative PowerSpec = (Measured Power / Manufacturer Specifications) and plot the Relative Power for each light source.


  • Report the results in a table

Manufacturer
Specifications (mW)		Measurements
2024-11-22 (mW)		Relative PowerSpec (%)
385nm150.25122.281%
470nm110.495.987%
555nm31.92475%
630nm5239.2676%

Metrics

  • The Maximum Power indicates how much light is provided by the instrument.
  • The Relative PowerSpec indicates how much power is provided compared to the specifications.

Conclusion

This instrument provides 80% of the power specified by the manufacturer. These results are consistent, as the manufacturer’s specifications are based on a different objective, and likely different filters and mirrors, which can affect the measured power output.

Illumination Stability

The light sources used on a microscope should remain constant or at least stable over the time scale of an experiment. For this reason, illumination stability is recorded across four different time scales:

  • Real-time Illumination Stability: Continuous recording for 1 minute. This represents the duration of a z-stack acquisition.
  • Short-term Illumination Stability: Recording every 1-10 seconds for 5-15 minutes. This represents the duration needed to acquire several images.
  • Mid-term Illumination Stability: Recording every 10-30 seconds for 1-2 hours. This represents the duration of a typical acquisition session or short time-lapse experiments. For longer time-lapse experiments, a longer duration may be used.
  • Long-term Illumination Stability: Recording once a year or more over the lifetime of the instrument.

For a detailed exploration of illumination stability, refer to the Illumination Power, Stability, and Linearity Protocol by the QuaRep Working Group 01.

Real-time Illumination Stability

Acquisition protocol

  • Warm up the light sources for the required duration (see previous section).
  • Place the power meter sensor on the microscope stage.
  • Center the sensor and the objective.
  • Zero the sensor to ensure accurate readings.
  • Select the wavelength of the light source you want to monitor, using your power meter controller or associated software.
  • Turn the light source on to 100% power.
  • Record the power output every 100 ms for 1 minute. For some microscope dedicated to a fast imaging it might be required to record stability at a faster rate. The Thorlab S170C sensor can record at 1 kHz !

    I personally acquire this data immediately after the warm-up kinetic experiment, without turning off the light source.

  • Repeat for each light source.

 Results

  • Use the provided spreadsheet template Illumination_Stability_Template.xlsx
  • Copy and paste the data from the recorded CSV file into the highlighted cells of the template to visualize your results.
  • For each light source, plot the Measured Power Output (in mW) over Time.

  • Calculate the Relative Power using the formula: Relative Power = (Power / Max Power). Then, plot the Relative Power (%) over Time.

  • Calculate the Stability Factor (S) using the formula: S (%) = 100 × (1 - (Pmax - Pmin) / (Pmax + Pmin)).
  • Calculate the Coefficient of Variation (CV) using the formula: CV = Standard Deviation / Mean.
  • Reports the results in a table.

Stability Factor Coefficient of Variation
385nm99.99%0.002%
475nm99.99%0.002%
555nm99.97%0.004%
630nm99.99%0.002%

From the Stability Factor results, we observe that the difference between the maximum and minimum power is less than 0.03%. Additionally, the Coefficient of Variation indicates that the standard deviation is less than 0.004% of the mean value, demonstrating excellent realtime power stability.

Conclusion

The light sources exhibit a very high stability, >99.9% during a 1-minute period.

Short-term Illumination Stability

Acquisition protocol

  • Warm up the light sources for the required duration (see previous section).
  • Place the power meter sensor on the microscope stage.
  • Center the sensor and the objective.
  • Zero the sensor to ensure accurate readings.
  • Select the wavelength of the light source you want to monitor, using your power meter controller or associated software.
  • Turn the light source on to 100% power.
  • Record the power output every 10 seconds for 15 minutes.

    I personally re-use the data collected during the warm-up kinetic experiment (when the power has been stable for 1h) for this purpose.

  • Repeat for each light source you.

Results

  • Use the provided spreadsheet template Illumination_Stability_Template.xlsx
  • Copy and paste the data from the recorded CSV file into the highlighted cells of the template to visualize your results.
  • For each light source, plot the Measured Power Output (in mW) over Time.

  • Calculate the Relative Power using the formula: Relative Power = (Power / Max Power). Then, plot the Relative Power (%) over Time.

  • Calculate the Stability Factor (S) using the formula: S (%) = 100 × (1 - (Pmax - Pmin) / (Pmax + Pmin)).
  • Calculate the Coefficient of Variation (CV) using the formula: CV = Standard Deviation / Mean.
  • Reports the results in a table.


Stability Factor Coefficient of Variation
385nm100.00%0.000%
475nm100.00%0.002%
555nm100.00%0.003%
630nm99.99%0.004%

From the Stability Factor, we observe that the difference between the maximum and minimum power is less than 0.01%. Additionally, the Coefficient of Variation indicates that the standard deviation is less than 0.004% of the mean value, demonstrating excellent short-term power stability.

Conclusion

The light sources exhibit high stability, maintaining >99.9% stability during a 15-minute period.

Mid-term Illumination Stability

Acquisition protocol

  • Warm up the light sources for the required duration (see previous section).
  • Place the power meter sensor on the microscope stage.
  • Center the sensor and the objective.
  • Zero the sensor to ensure accurate readings.
  • Select the wavelength of the light source you want to monitor, using your power meter controller or associated software.
  • Turn the light source on to 100% power.
  • Record the power output every 10 seconds for 1 hour.

    I personally re-use the data collected during the warmup kinetic experiment.

  • Repeat for each light source

Results

  • Use the provided spreadsheet template Illumination_Stability_Template.xlsx
  • Copy and paste the data from the recorded CSV file into the highlighted cells of the template to visualize your results.
  • For each light source, plot the Measured Power Output (in mW) over Time

  • Calculate the Relative Power using the formula: Relative Power = (Power / Max Power). Then, plot the Relative Power (%) over Time.

  • Calculate the Stability Factor (S) using the formula: S (%) = 100 × (1 - (Pmax - Pmin) / (Pmax + Pmin)).
  • Calculate the Coefficient of Variation (CV) using the formula: CV = Standard Deviation / Mean.
  • Reports the results in a table.

Stability Factor Coefficient of Variation
385nm99.98%0.013%
475nm99.98%0.011%
555nm99.99%0.007%
630nm99.97%0.020%

From the Stability Factor, we observe that the difference between the maximum and minimum power is less than 0.03%. Additionally, the Coefficient of Variation indicates that the standard deviation is less than 0.02% of the mean value, demonstrating excellent mid-term power stability.

Conclusion

The light sources exhibit exceptional stability, maintaining a performance of >99.9% during a 1-hour period.

Long-term Illumination Stability

Long-term illumination stability measures the power output over the lifetime of the instrument. Over time, we expect a gradual decrease in power output due to the aging of hardware, including the light source and other optical components. These measurements are not an experiment per se but it is the measurement of the maximum power output over time.

Acquisition protocol

  • Warm up the light sources for the required duration (see previous section).
  • Place the power meter sensor on the microscope stage.
  • Center the sensor and the objective.
  • Zero the sensor to ensure accurate readings.
  • Select the wavelength of the light source you want to monitor, using your power meter controller or associated software.
  • Turn the light source on to 100% power.
  • Record the average power output over 10 seconds.
    I personally re-use the data collected for the maximal power output section and plot it over time.
  • Repeat for each light source.

Results

  • Use the provided spreadsheet template Illumination_Long-Term_Stability_Log.xlsx
  • Copy and paste the data from the recorded CSV file into the highlighted cells of the template to visualize your results.
  • For each light source, plot the Measured Power Output (in mW) over Time

  • Calculate the Relative Power using the formula: Relative Power = (Power / Max Power). Then, plot the Relative Power (%) over Time.

  • Calculate the Relative PowerSpec by comparing the measured power to the manufacturer’s specifications using the following formula: Relative PowerSpec = Power / PowerSpec
  • Plot the Relative PowerSpec (% Spec) over Time.

We expect a gradual decrease in power output over time due to the aging of hardware. Light sources should be replaced when the Relative PowerSpec falls below 50%.

  • Reports the results in a table.

Relative PowerSpec
385nm80.53%
475nm83.61%
555nm65.83%
630nm67.12%
  • Keep the Log file to append future measurements

Conclusion

The light sources are somehow stable over the last 2 years but a decrease in the maximum power output is seen.

Illumination Stability Conclusions

Stability Factor

Real-time

1 min

Short-term

15 min

Mid-term

1 h

385nm

99.99%100.00%99.98%

475nm

99.99%100.00%99.98%

555nm

99.97%100.00%99.99%

630nm

99.99%99.99%99.97%

The light sources are highly stable (Stability >99.9%).

Metrics

  • The Stability Factor indicates a higher stability the closer to 100% and focuses specifically on the range of values (difference between max and min) relative to their sum, providing an intuitive measure of how tightly the system's behavior stays within a defined range.
  • The Coefficient of Variation focuses on the dispersion of all data points (via the standard deviation) relative to the mean. Lower Coefficient indicates a better stability around the mean.


Illumination Input-Output Linearity

This measure compares the power output as the input varies. A linear relationship is expected between the input and the power output. For a detailed exploration of illumination linearity, refer to the Illumination Power, Stability, and Linearity Protocol by the QuaRep Working Group 01.

Acquisition protocol

  • Warm up the light sources for the required duration (see previous section).
  • Place the power meter sensor on the microscope stage.
  • Center the sensor and the objective.
  • Zero the sensor to ensure accurate readings.
  • Select the wavelength of the light source you want to monitor, using your power meter controller or associated software.
  • Adjust its intensity to 100%, then incrementally decrease the intensity input increase to 90%, 80%, 70%, and so on, until 0%.
    I typically collect this data immediately after the warm-up kinetic phase and once the real-time power stability data has been recorded.
  • Record the power output corresponding to each input level.
  • Repeat for each light source

Results

  • Use the provided spreadsheet template Illumination_Linearity_Template.xlsx
  • Enter your measurement into the highlighted cells of the template to visualize your results.
  • For each light source, plot the Measured Power output (in mW) as a function of the Input (%).

  • Calculate the Relative Power using the formula: Relative Power = Power / MaxPower.
  • Plot the Relative Power (%) as a function of the Input (%).

  • Determine the equation for each curve, which is typically a linear relationship of the form: Output = K × Input
  • Report the Slope (K) and the Coefficient of Determination (R²) for each curve in a table.

Illumination Input-Output Linearity


Slope

R2

385nm

0.9969

1

475nm

0.9984

1

555nm

1.0012

1

630nm

1.0034

1

The slopes demonstrate a nearly perfect linear relationship between the input and the measured output power, with values very close to 1. The coefficient of determination (R²) indicates a perfect linear fit, showing no deviation from the expected relationship.

Metrics

  • The Slope indicates the rate of change between Input and Ouput.
  • The Coefficient of Determination indicates how fitted is the data to a linear relationship.

Conclusion

The light sources are highly linear with an average Slope=0.999 and a perfect fit R2=1

Objectives and Cubes Transmittance

Since we are using a power meter, we can easily assess the transmittance of the objectives and filter cubes. This measurement compares the power output when different objectives and filter cubes are in the light path. It evaluates the transmittance of each objective and compares it with the manufacturer’s specifications. This method can help detect defects or dirt on the objectives. It can also verify the correct identification of the filters installed in the microscope.

Objectives Transmittance

Acquisition protocol

  • Warm up the light sources for the required duration (see previous section).
  • Place the power meter sensor on the microscope stage.
  • Center the sensor and the objective.
  • Zero the sensor to ensure accurate readings.
  • Select the wavelength of the light source you want to monitor, using your power meter controller or associated software.
  • Turn on the light source to 100% intensity.
  • Record the power output for each objective, as well as without the objective in place.

    I typically collect this data after completing the warm-up kinetic phase, followed by the real-time power stability measurements, and immediately after recording the power output linearity.

  • Repeat for each light source and wavelength

Results

  • Use the provided spreadsheet template Objective and cube transmittance_Template.xlsx
  • Enter your measurement into the highlighted cells of the template to visualize your results.
  • For each light source, plot the Measured Power output (in mW) as a function of the wavelength (in nm).

  • Calculate the Relative Transmittance using the formula: Relative Transmittance = Power / PowerNoObjective.
  • Plot the Relative Transmittance (%) as a function of the wavelength (in nm).

  • Calculate the average transmittance for each objective
  • Compare the average transmittance to the specifications provided by the manufacturer
  • Report results in a table.

Average
Transmittance		Specifications
[470-630]

Average Transmittance
[470-630]

2.5x-0.07577%>90%84%
10x-0.25-Ph160%>80%67%
20x-0.5 Ph262%>80%68%
63x-1.429%>80%35%

The measurements are generally close to the specifications, with the exception of the 63x-1.4 objective. This deviation is expected, as the 63x objective has a smaller back aperture, which reduces the amount of light it can receive..

Conclusion

The objectives are transmitting light properly.

Cubes Transmittance

Acquisition protocol

  • Warm up the light sources for the required duration (see previous section).
  • Place the power meter sensor on the microscope stage.
  • Center the sensor and the objective.
  • Zero the sensor to ensure accurate readings.
  • Select the wavelength of the light source you want to monitor, using your power meter controller or associated software.
  • Turn on the light source to 100% intensity.
  • Record the power output for each filter cube.

    I typically collect this data after completing the warm-up kinetic phase, followed by the real-time power stability measurements, and immediately after recording the power output linearity.

  • Repeat for each light source and wavelength

Results

  • Use the provided spreadsheet template Objective and cube transmittance_Template.xlsx
  • Enter your measurement into the highlighted cells of the template to visualize your results.
  • For each light source, plot the Measured Power output (in mW) as a function of the wavelength (in nm)..

  • Calculate the Relative Transmittance using the formula: Relative Transmittance = Power / PowerMaxFilter.
  • Plot the Relative Transmittance (%) as a function of the wavelength (in nm).

  • Calculate the Average Transmittance for each filter at the appropriate wavelengths
  • Report the results in a table.

385475555590630
DAPI/GFP/Cy3/Cy5100%100%100%100%100%
DAPI14%0%0%8%0%
GFP0%47%0%0%0%
DsRed0%0%47%0%0%
DHE0%0%0%0%0%
Cy50%0%0%0%84%

  • The DAPI cube transmits only 14% of the excitation light compared to the Quad Band Pass DAPI/GFP/Cy3/Cy5. While it is still usable, it will provide a low signal. This is likely because the excitation filter within the cube does not match the light source properly. Since an excitation filter is already included in the light source, the filter in this cube could be removed.

  • The GFP and DsRed cubes transmit 47% of the excitation light compared to the Quad Band Pass DAPI/GFP/Cy3/Cy5, and they are functioning properly.

  • The DHE cube does not transmit any light from the Colibri. This cube may need to be removed and stored.

  • The Cy5 cube transmits 84% of the excitation light compared to the Quad Band Pass DAPI/GFP/Cy3/Cy5, and it is working properly.

Conclusion

Actions to be Taken:

  • Remove the excitation filter from the DAPI cube, as it does not properly match the light source and is redundant with the excitation filter already included in the light source.
  • Remove and store the DHE cube, as it does not transmit any light from the Colibri and is no longer functional.

We are done with the powermeter (wink) .

Field Illumination Uniformity

Having confirmed the stability of our light sources and verified that the optical components (objectives and filter cubes) are transmitting light effectively, we can now proceed to evaluate the uniformity of the illumination. This step assesses how evenly the illumination is distributed. For a comprehensive guide on illumination uniformity, refer to the Illumination Uniformity by the QuaRep Working Group 03.

Acquisition protocol 

  1. Place a fluorescent plastic slide (Thorlabs FSK5) onto the stage.
  2. Center the slide and the objective.
  3. Adjust the focus to align with the surface of the slide.

    I typically use a red fluorescent slide and focus on a scratch mark on its surface for alignment.

  4. Slightly adjust the focus deeper into the slide to minimize the visibility of dust, dirt, and scratches.

  5. Modify the acquisition parameters to ensure the image is properly exposed.

    I typically use the auto-exposure feature, aiming for a targeted intensity of 30%.

  6. Capture a multi-channel image.
  7. Repeat steps 5 and 6 for each objective and filter combination.

Processing

You should have acquired several multi-channel images that now need processing to yield meaningful results.

I initially was using the Field Illumination analysis function of the MetroloJ_QC plugin for FIJI but eventually branched away to write my own processing plugin named QC Scope. For more information about the QC Scope please refer to the QC Scope repository on Github. For more information about MetroloJ_QC plugin please refer to manual available on the MontpellierRessourcesImagerie repository on GitHub.

  1. Download FIJI for your operating system
  2. Download QC Scope (aka QC_Scope.jar)
  3. Unzip FIJI.app to your favorite location on your computer (usually your Desktop but not in a System Directory).
  4. Start FIJI (aka FIJI.app, ImageJ-win.exe)
  5. Drag and drop QC_Scope.jar into the FIJI toolbar
  6. Click Save
  7. Click OK
  8. Close FIJI

  9. Reopen FIJI

After installation, QC Scope will be available in the FIJI menu under Plugins>QC Scope. The QC Scope floating toolbar is available for rapid and convenient access. To load it, select the QC Scope Toolbar in the FIJI menu under Plugins>QC Scope>QC Scope Toolbar. Additionally and for an even more convenient usage, it is possible to have the QC Scope Toolbar automatically loaded when FIJI is started. Click the Autostart button  or select the QC Scope Toolbar Autostart in the FIJI menu under Plugins>QC Scope>QC Scope Toolbar Autostart.

  1. Open FIJI.
  2. Launch the QC Scope Toolbar by navigating to Plugins>QC Scope>QC Scope Toolbar.
  3. Click on Uniformity.
    1. If one or more images are already opened QC Scope will process them.
    2. If no image is open, QC Scope will prompt to select a folder and process all images withing the folder (and subfolders)

      QCSCope file format compatibility

      For now, QC Scope only processes images with the following extensions  ".tif", ".tiff", ".jpg", ".jpeg", ".png", ".czi", ".nd2", ".lif", ".lsm", ".ome.tif", ".ome.tiff"

  4. QC Scope will try to read the Metadata from the first image and pre-process all the channels with default or the last used processing settings
  5. It will display the metadata, the initial results and the processing options in a dialog
    1. Microscope Metadata: Objective Magnification, NA, and Immersion media
    2. Image Calibration status, Pixel Width, Height, Voxel Size, Unit
    3. For each channel: Name and Emission Wavelength
    4. Processing Settings:
      1. Binning Method:
        1. Iso-Density (preferred): This method divides the image into 10 bins of equal Nb of Pixels. Nb Pixel Per Bin = (Width x Height) / 10 and assign a new pixel value of 25 for all the Nb Pixel Per Bin darkest pixels, 50 for the next darkest Nb Pixel Per Bin etc... until 250 for the brightest Nb Pixel Per Bin pixels.
        2. Iso-Intensity: This method divides the image into 10 bins of equal bin intensities. Bin Width = (Max - Min) / 10 and assign a new pixel value of 25 for all the pixels with an intensity between Min and Min + Bin With, 50 for intensities between Min + Bin Width and Min + 2 x Bin Width etc... until 250 for intensities between Min + 9 x Bin Width and Max.
      2. Gaussian Blur: Apply a gaussian blur with the given Gaussian Blur Sigma before processing the image channel
      3. Channel: The selected channel will be processed with the entered processing parameters and displayed as part of the testing process to define optimal processing parameters.
      4. Batch Mode: If activated, QC Scope will re-use the settings without displaying the dialog unless metadata differs
      5. Save Individual Files: For each image, QC Scope will save the individual processed images (1 per channel) and a CSV file with all measured parameters.
      6. Prolix Mode: Display all the QC Scope actions in the Log

      7. Test Processing: When selected the Dialog will keep appearing. This is useful to test the Processing Settings
































  6. QC Scope will save files in a folder named Output on your desktop:

    1. At least 2 CSV files:

      1. Field Uniformity_All-Data_Merged.csv gathers all the measured parameters

      2. Field Uniformity_Essential-Data_Merged.csv gathers only the essential information

    2. Optionnally, if Save Individual Files is selected QC Scope will also save:
      1. 1 CSV file per image NameOfYourImage_Uniformity-Data.csv with one row per channel containing all the measured parameters
      2. 1 TIF file per channel for every processed image showing the binned (Iso-density or Iso-Intensity) image map with the Reference Center indicated as an overlay

Note: QC Scope never overwrites files. It will check for the existence of files and increment a number until it can safely write the output file.




Description of QC Scope Field Uniformity Results (in bold the results included in Essential Data)

Key OrderField NameData ExampleData TypeDescription
1Filename10x_Quad_Exp-01.cziStringName of the processed image
2Channel Nb4IntegerNumber of the Channel from 1 to n
3Channel NameDAPIStringName of the Channel
4Channel Wavelength EM (nm)465IntegerChannel Emission Wavelength
5Objective Magnification10xStringObjective Magnification
6Objective NA0.25FloatObjective Numerical Aperture
7Objective Immersion MediaAirStringObjective Immerion Media
8Gaussian Blur AppliedTRUEBooleanIf Gaussian Blur was applied
9Gaussian Sigma10IntegerSigma of the Gaussian Blur
10Binning MethodIso-DensityStringBinning Method used
11Batch ModeTRUEBooleanBoolean key to process images in batch mode (no Dialog)
12Save Individual FilesFALSEBooleanBoolean key to save individual files (1 csv data per file with 1 row per channel, 1 tif binned image par channel)
13Prolix ModeFALSEBooleanBoolean key display detailed plugin actions in the log
14Image Min Intensity856IntegerRaw Image Minimum of Pixel Intensities
15Image Max Intensity1080IntegerRaw Image Maximum of Pixel Intensities
16Image Mean Intensity959.1FloatRaw Image Mean of Pixel Intensities
17Image Standard Deviation Intensity24.3FloatRaw Image Standard Deviation of Pixel Intensities
18Image Median Intensity959IntegerRaw Image Median Pixel Intensity
19Image Mode Intensity110IntegerRaw Image Mode Pixel Intensity
20Image Width (pixels)1388IntegerImage Width in pixels
21Image Height (pixels)1040IntegerImage Height in pixels
22Image Bit Depth16IntegerImage Bit Depth
23Pixel Width (um)0.645FloatImage pixel width (unit/px)
24Pixel Height (um)0.645FloatImage pixel height (unit/px)
25Pixel Depth (um)1FloatImage voxel depth (unit/voxel)
26Space UnitmicronStringRaw Image Space Unit
27Space Unit StandardumStringStandardize Space Unit (nm, um, cm, m)
28Calibration StatusTRUEBooleanBoolean key displaying the calibration status
29Standard Deviation (GV)24.3FloatRaw Image Standard Deviation of Pixel Intensities
30Uniformity Standard (%)79.3FloatUniformity as calculated by MetroloJ_QC. Uniformity_Standard = 100 * (Min / Max)
31Uniformity Percentile (%)95.8FloatUniformity calculated with the average of the 5% and 95% pixel intensities.
Uniformity_Percentile = (1 - (Avg_Intensity95 - Avg_Intensity5) / (Avg_Intensity95 + Avg_Intensity5) ) * 100
32Coefficient of Variation0.0253FloatCoefficient of variation. CV = (Std_Dev / Mean)
33Uniformity CV based97.5FloatUniformity calculated from the Coefficient of variation. Uniformity_CV = (1 - CV) * 100
34X Center (pixels)694IntegerCoordinate in pixel of the center of the Image (Ideal centering). Image Width (pixels) / 2
35Y Center (pixels)520IntegerCoordinate in pixel of the center of the Image (Ideal centering). Image Height (pixels) / 2
36X Ref (pixels)230.4FloatCoordinate in pixels of the centroid of the largest particule identified in the last bin. Used to caculate the Centering Accuracy
37Y Ref (pixels)876.4FloatCoordinate in pixels of the centroid of the largest particule identified in the last bin. Used to caculate the Centering Accuracy
38X Ref (um)148.6FloatCoordinate in scaled unit of the centroid of the largest particule identified in the last bin.
39Y Ref (um)565.3FloatCoordinate in scaled unit of the centroid of the largest particule identified in the last bin.
40Centering Accuracy (%)32.6FloatCentering Accuracy = 100 - 100 * (2 / sqrt(Image Width**2 + Image Height**2)) * sqrt ( (X_Ref_Pix - Image Width/2)**2 + (Y_Ref_Pix - Image Height/2)**2)


  1. Open FIJI.
  2. Load your image by dragging it into the FIJI bar.
  3. Launch the MetroloJ_QC plugin by navigating to Plugins > MetroloJ QC.
  4. Click on Field Illumination Report.
  5. Enter a Title for your report, I typically use the date
  6. Type in your Name.
  7. Click Filter Parameters and input the filter's names, excitation, and emission wavelengths.

  8. Check Remove Noise using Gaussian Blur.

  9. Enable Apply Tolerance to the Report and reject uniformity and accuracy values below 80%.

  10. Click File Save Options.

  11. Select Save Results as PDF Reports.

  12. Select Save Results as spreadsheets.

  13. Click OK.

  14. Repeat steps 4 through 13 for each image you have acquired.

This will generate detailed results stored in a folder named Processed. The Processed folder will be located in the same directory as the original images, with each report and result saved in its own sub-folder.

The following script for R will process the Illumination Uniformity Results generated by MetroloJ_QC Process Illumination Uniformity Results.R. To use it, simply drag and drop the file into the R interface. You may also open it with RStudio and Click the Source button. The script will:

  • Prompt the user to select an input folder.
  • Load all _results.xls files located in the selected folder and subfolders
  • Process and merge all results generated by the MetroloJ_QC Field Illumination.
  • Split the filenames using the _ character as a separator and create columns named Variable-001, Variable-002, etc. This will help in organizing the data if the filenames are formatted as Variable-A_Variable-B, for example.
  • Save the result as Field-Unifromity_Merged-Data.csv in an Output folder on the user's Desktop 


This script will generate a csv file that can be saved as an XLSX and manipulated with a pivot table to generate informative graphs and statistics. From the Pivot table copy and paste your data into the orange cells of the following spreadsheet Illumination_Uniformity_Template.xlsx. 

Results

Plot the uniformity and centering accuracy for each objective.

Metrics

  • The Uniformity indicates the range between the minimum and maximum intensities in the image. U=(Min/Max)*100. 100% Uniformity indicates a perfectly homogeneous image. 50% Uniformity indicates the minimum is half the maximum.
  • The Centering Accuracy indicates how far from the center of the image is the center of the illumination (centroid of the max illumination bin). 100% indicates a perfectly aligns with the center of the image. 0% centering accuracy indicates that the center of the illumination is the farthest from the center of the image.



ObjectiveUniformityCentering Accuracy
2x97.5%92.7%
10x97.0%94.5%
20x97.3%97.1%
63x96.6%96.7%


 Plot the uniformity and centering accuracy for each filter set. 

FilterUniformityCentering Accuracy
DAPI98.3%99.4%
DAPIc95.8%84.9%
GFP98.1%99.1%
GFPc96.5%93.3%
Cy397.6%96.5%
Cy3c96.8%97.9%
Cy597.0%99.6%
Cy5c96.7%91.3%


This specific instrument has a quad-band filter as well as individual filter cubes. We can plot the uniformity and centering accuracy per filter types. 

Filter TypeUniformity

Centering Accuracy

Quad band

97.7%98.7%
Single band96.5%

91.8%

Store the original field illumination images to be able to perform shading corrections after acquisition.

Conclusion

The uniformity and centering accuracy are excellent across all objectives and filters, consistently exceeding 90%. However, the single-band filter cubes exhibit slightly lower uniformity and centering accuracy compared to the quad-band filter cube.

Store the original field illumination images to be able to perform shading corrections after acquisition.

XYZ Drift

This experiment evaluates the stability of the system in the XY and Z directions. As noted earlier, when an instrument is started, it requires a warmup period to reach a stable steady state. To determine the duration of this phase accurately, it is recommended to record a warmup kinetic at least once per year. For a comprehensive guide on Drift and Repositioning, refer to the Stage and Focus Precision by the QuaRep Working Group 06.

Acquisition protocol 

  1. Place 4 µm diameter fluorescent beads (TetraSpec Fluorescent Microspheres Size Kit, mounted on a slide) on the microscope stage.

  2. Center an isolated bead under a high-NA dry objective.

  3. Crop the acquisition area to only visualize one bead but keep it large enough to anticipate a potential drift along the X and Y axes (100 um FOV should be enough)

  4. Select an imaging channel appropriate for the fluorescent beads

    I typically use the Cy5 channel which is very bright and resistant to bleaching, yet this channel has a lower resolution, but it does not really matter here

  5. Acquire a large Z-stack at 1-minute intervals for a duration of 24 hours.

    To ensure accurate measurements, it is essential to account for potential drift along the Z-axis by acquiring a Z-stack that is substantially larger than the visible bead size. I typically acquire a 40 µm Z-stack.

Processing

  1. Open your image in FIJI
  2. If necessary, crop the image to focus on a single bead for better visualization.
  3. Use the TrackMate plugin included in FIJI to detect and track spots over time Plugins\Tracking\TrackMate.
  4. Apply Difference of Gaussians (DoG) spot detection with a detection size of 4 µm
  5. Enable sub-pixel localization for increased accuracy
  6. Click Preview to visualize the spot detection
  7. Set a quality threshold (click and slide) high enough to detect a single spot per frame
  8. Click Next and follow the detection and tracking process
  9. Save the detected Spots coordinates as a CSV file for further analysis

Results

  • Open the spreadsheet template XYZ Drift Kinetic_Template.xlsx and fill in the orange cells.
  • Copy and paste the XYZT and Frame columns from the TrackMate spots CSV file into the corresponding orange columns in the spreadsheet.
  • Enter the numerical aperture (NA) and emission wavelength used during the experiment.
  • Calculate the relative displacement in X, Y, and Z using the formula: Relative Displacement = Position - PositionInitial.
  • Finally, plot the relative displacement over time to visualize the system's drift.

We observe an initial drift that stabilizes over time in X (+2 um), Y (+1.3 um) and Z (-10.5 um). Calculate the displacement 3D Displacement = Sqrt( (X2-X1)2 + (Y2-Y1)2) + (Z2-Z1)2 ) and plot the displacement over time. Calculate the resolution of your imaging configuration, Lateral Resolution = LambdaEmission / 2*NA and plot the resolution over time (constant).

Identify visually the time when the displacement is lower than the resolution of the system. On this instrument it takes 120 min to reach its stability. Calculate the velocity, Velocity = (Displacement2-Displacement1)/T2-T1) and plot the velocity over time.

Calculate the average velocity before and after stabilisation and report the results in a table.

Objective NA0.5
Wavelength (nm)705
Resolution (nm)705
Stabilisation time (min)122
Average velocity Warmup (nm/min)113
Average velocity System Ready (nm/min)14

Metrics

  • The Stabilisation Time indicates the time in minutes necessary for the instrument to have a drift lower than the resolution of the system.
  • The Average Velocity indicates the speed of drift in all directions XYZ in nm/min.

Conclusion

The warmup time for this instrument is approximately 2 hours. After the warmup period, the average displacement velocity is 14 nm/min, which falls within an acceptable range.

Stage Repositioning Dispersion

This experiment evaluates how accurate is the system in XY by measuring the dispersionof repositioning. Several variables can affect repositioning: i) Time, ii) Traveled distance, iii) Speed and iv) acceleration. For a comprehensive guide on Stage Repositioning, refer to the Stage and Focus Precision by the QuaRep Working Group 06 and the associated XY Repositioning Protocol.

Acquisition protocol

  1. Place 4 µm diameter fluorescent beads (TetraSpec Fluorescent Microspheres Size Kit, mounted on a slide) on the microscope stage.

  2. Center an isolated bead under a high-NA dry objective.

  3. Crop the acquisition area to only visualize one bead but keep it large enough to anticipate a potential drift along the X and Y axes (100 um FOV should be enough)

  4. Select an imaging channel appropriate for the fluorescent beads

    I typically use the Cy5 channel which is very bright and resistant to bleaching, yet this channel has a lower resolution.

  5. Acquire a Z-stack at positions separated by 0 um (Drift), 1000 um and 10 000 um in both X and Y direction.

  6. Repeat the acquisition 20 times

  7. Acquire 3 datasets for each condition

  • Your stage might have a smaller range!
  • Lower the objectives during movement to avoid damage

Processing

  1. Open your image in FIJI
  2. If necessary, crop the image to focus on a single bead for better visualization.
  3. Use the TrackMate plugin included in FIJI to detect and track spots over time Plugins\Tracking\TrackMate.
  4. Apply Difference of Gaussians (DoG) spot detection with a detection size of 4 µm
  5. Enable sub-pixel localization for increased accuracy
  6. Click Preview to visualize the spot detection
  7. Set a quality threshold (click and slide) high enough to detect a single spot per frame
  8. Click Next and follow the detection and tracking process
  9. Save the detected Spots coordinates as a CSV file for further analysis

Results

  • Open the spreadsheet template Stage Repositioning_Template.xlsx and fill in the orange cells.
  • Copy and paste the XYZT and Frame columns from the TrackMate spots CSV file into the corresponding orange columns in the spreadsheet.
  • Enter the numerical aperture (NA) and emission wavelength used during the experiment.
  • Calculate the relative position in X, Y, and Z using the formula: Relative PositionRelative = Position - PositionInitial.
  • Finally, plot the relative position over time to visualize the system's stage repositioning dispersion.


We observe an initial movement in X and Y that stabilises. Calculate the displacement 2D Displacement = Sqrt( (X2-X1)2 + (Y2-Y1)2) ) and plot the 2D displacement over time. Calculate the resolution of your imaging configuration, Lateral Resolution = LambdaEmission / 2*NA and plot the resolution over time (constant).


This experiment shows a significant initial displacement between Frame 0 and Frame 1, ranging from 1000 nm to 400 nm, which decreases to 70 nm by Frame 2. To quantify this variation, calculate the dispersion for each displacement using the formula: Dispersion = StandardDeviation(Displacement). Report the results in a table.

Traveled Distance (mm)0 mm1 mm10 mm
X Dispersion (nm)4188121
Y Dispersion (nm)414148
Z Dispersion (nm)103453
Repositioning Dispersion 3D (nm)622791
Repositioning Dispersion 2D (nm)222690

Conclusion

The system has an effective Stage Repositioning Dispersion of 230 nm. The results are higher than expected because of the initial shift in the first frame that eventually stabilizes. Excluding the first frame significantly improves the measurements, reducing the repositioning dispersion to 40 nm. Further investigation is required to understand the underlying cause.

Traveled Distance (um)0 mm1 mm10 mm
X Dispersion (nm)32852
Y Dispersion (nm)36835
Z Dispersion (nm)102640
Repositioning Dispersion 3D (nm)64336
Repositioning Dispersion 2D (nm)24036

Metrics

The Repositioning Dispersion indicates how spread is the repositioning in nm. The lower, the more accurate.


Further Investigation


We observed a significant shift in the first frame, which was unexpected and invites further investigation. These variables can affect repositioning dispersion: i) Traveled distance, ii) Speed, iii) Acceleration, iv) Time, and v) Environment. We decided to test the first three.

Methodology

To test if these variables have a significant impact on the repositioning, we followed the XYZ repositioning dispersion protocol with the following parameters:

  • Distances: 0um, 1um, 10um, 1000um, 10 000um, 30 000um 
  • Speed: 10%, 100%
  • Acceleration: 10%,100%
  • For each conditions 3 datasets were acquired

 Processing

This experimental protocol generated a substantial number of images. To process them automatically in ImageJ/FIJI using the TrackMate plugin, we use the following script Stage Repositioning with Batch TrackMate-v7.py

This script automates the process of detecting and tracking spots using the TrackMate plugin for ImageJ/FIJI. To use it:

  • Drop the script into the FIJI toolbar and click Run.

If images are already opened:

  1. Prompt User for Configuration:

    • The user is prompted to configure settings such as enabling subpixel localization, adjusting spot diameter, setting the threshold, and applying median filtering. These settings can be loaded from previous configurations stored in ImageJ’s preferences or set to default values. It’s recommended to run TrackMate manually on a representative image to fine-tune detection parameters (ideally detecting one spot per frame).

  2. User Selection for Image Processing:

    • The user can choose to process all open images or just the active one.

  3. For Each Image:

    • Configure TrackMate Detector and Tracker: The TrackMate detector is configured using the Difference of Gaussians method, and the tracker is set to SparseLAPTracker.
    • Analyze and Filter Tracks: Features are analyzed, and tracks are filtered according to user-defined settings.
    • Process Image: The image is processed, spots are tracked, and results are visualized in the active window.
    • Export Spot Data: Detected spot data is exported to a uniquely named CSV file stored in an "Output" directory on the desktop.
    • Save Settings: User-defined settings are saved to ImageJ’s preferences for future use.

  4. Summary:

    • A dialog is displayed summarizing the number of processed images and the location of the saved output.

If no image is opened:

  1. Prompt User to Select Folder:

    • If no image is open, the user is prompted to select a folder containing images for processing.

  2. Process the First Image:

    • The first image is opened, and the processing workflow is the same as described for open images. The user can choose to process just this image or all images in the folder.

  3. For Subsequent Images:

    • Batch Mode Handling:
      • If batch mode is enabled, images are processed without additional user input, and the settings are applied automatically.
      • If batch mode is not enabled, the user is prompted to configure settings for each subsequent image.

  4. Summary:

    • A dialog is displayed summarizing the number of processed images and the location of the saved output.

Post-Processing:

  • A summary message is logged, detailing the number of processed images and the location where the CSV files are saved.
  • The script merges and processes CSV files into a single "Merged-Data" file stored in the output directory.
  • The user is notified when the merging and processing are completed successfully.

From v7, the FIJI script performs all the necessary tasks.

Prior v7, it was generating a CSV file for each image, which can be aggregated for further analysis using the accompanying R script, Process Stage Repositioning Results.R. This R script processes all CSV files in a selected folder and save the file as Stage-Repositioning_Merged-Data.csv in an "Output" folder on the user desktop for streamlined data analysis.

This R code automates the processing of multiple CSV files containing spot tracking data.

  1. Installs and loads required libraries
  2. Sets directories for input (CSV files) and output (merged data) on the user's desktop.
  3. Lists all CSV files in the input directory and reads in the header from the first CSV file.
  4. Defines a filename cleaning function to extract relevant metadata from the filenames (e.g., removing extensions, and extracting variables).
  5. Reads and processes each CSV file:
    • Skips initial rows and assigns column names.
    • Cleans up filenames and adds them to the dataset.
    • Calculates displacement in the X, Y, and Z axes relative to the initial position with the formulat  PositionRelative= Position - PositionInitial, and computes both 3D and 2D displacement values with the following formulas:  2D_Displacement = Sqrt( (X2-X1)2 + (Y2-Y1)2)); 3D_Displacement = Sqrt( (X2-X1)2 + (Y2-Y1)2) + (Z2-Z1)2 )
  6. Merges all the processed data into a single dataframe.
  7. Saves the results as Stage-Repositioning_Merged-Data.csv located in the selected input folder.

This script generate a single CSV File that can be further processed and summarized with a pivot table as shown in the following spreadsheet Stage-Repositioning_Template.xlsx

Using the first frame as a reference we can plot the average XYZ position for each frame.

 

As observed earlier, there is a significant displacement between Frame 0 and Frame 1, particularly along the X-axis. For this analysis, we will exclude the first two frames and focus on the variables of interest: (i) Traveled distance, (ii) Speed, and (iii) Acceleration and will come back to the initial shift later.

 Repositioning Dispersion: Impact of Traveled Distance

Results

Plot the 2D displacement versus the frame number for each condition of traveled distance.


The data looks good now with the two first frames ignored. Now, we can calculate the average of the standard deviation of the 2D displacement and plot these values against the traveled distance..


 We observe a power-law relationship, described by the equation: Repositioning Dispersion = 8.2 x Traveled Distance^0.2473

Traveled Distance (um)Repositioning Dispersion (nm)
04
16
1020
10019
100076
1000056
30000107

Conclusion

In conclusion we observe that the traveled distance significantly (one-way ANOVA) affects the repositioning dispersion. However, this dispersion is much lower than the lateral resolution of the system (705nm).

Repositioning Dispersion: Impact of Speed and Acceleration

Results

Generate a plot of the 2D displacement as a function of frame number for each combination of Speed and Acceleration conditions. This visualization will help assess the relationship between displacement and time across the different experimental settings.


As noted earlier, there is a significant displacement between Frame 0 and Frame 1, particularly along the X-axis (600 nm) and, to a lesser extent, the Y-axis (280 nm). To refine our analysis, we will exclude the first two frames and focus on the key variables of interest: (i) Speed and (ii) Acceleration. To better understand the system's behavior, we will visualize the average standard deviation of the 2D displacement for each combination of Speed and Acceleration conditions.

Our observations indicate that both Acceleration and Speed contribute to an increase in 2D repositioning dispersion. However, a two-way ANOVA reveals that only Speed has a statistically significant effect on 2D repositioning dispersion. Post-hoc analysis further demonstrates that the dispersion for the Speed-Fast, Acc-High condition is significantly greater than that of the Speed-Low, Acc-Low condition.


2D Repositioning Dispersion (nm)
Speed-Slow Acc-Low32
Speed-Slow Acc-High49
Speed-Fast Acc-Low54
Speed-Fast Acc-High78

Conclusion

In conclusion we observe that Speed but not Acceleration increase 2D Repositioning Dispersion.

What about the initial shift ?

Right, I almost forgot about that. See below.

Results

Ploting the 3D displacement for each tested conditions from the preivous data.

We observe a single floating point that corresponds to the displacement between Frame 0 and Frame 1. This leads me to hypothesize that the discrepancy may be related to the stage's dual motors, each controlling a separate axis (X and Y). Each motor operates in two directions (Positive and Negative). Since the shift occurs only at the first frame, this likely relates to how the experiment is initiated.

To explore this further, I decided to test whether these variables significantly impact the repositioning. We followed the XYZ repositioning dispersion protocol, testing the following parameters:

  • Distance: 1000 µm
  • Speed: 100%
  • Acceleration: 100%
  • Axis: X, Y, XY
  • Starting Point: Centered (on target), Positive (shifted positively from the reference position), Negative (shifted negatively from the reference position) 
  • For each condition, three datasets were acquired.

Data Stage-Repositining_Diagnostic-Data.xlsx was processed as mentioned before and we ploted the 2D displacement function of the frame for each condition.

When moving along the X-axis only, we observe a shift in displacement when the starting position is either centered or positively shifted, but no shift occurs when the starting position is negatively shifted. This suggests that the behavior of the stage’s motor or the initialization of the experiment may be affected by the direction of the shift relative to the reference position, specifically when moving in the positive direction.

When moving along the Y-axis only, we observe a shift in displacement when the starting position is positively shifted, but no shift occurs when the starting position is either centered or negatively shifted. This indicates that the stage's motor behavior or initialization may be influenced by the direction of the shift, particularly when starting from a positive offset relative to the reference position.

When moving along both the X and Y axes simultaneously, a shift is observed when the starting position is centered. This shift becomes more pronounced when the starting position is positively shifted in any combination of the X and Y axes (+X+Y, +X-Y, -X+Y). However, the shift is reduced when the starting position is negatively shifted along both axes.

Conclusion

In conclusion, the observed shifts in repositioning dispersion are influenced by the initial starting position of the stage. When moving along the X and Y axes simultaneously, the shift is most significant when the starting position is centered and increases with positive shifts in both axes. Conversely, when the starting position is negatively shifted along both axes, the shift is reduced. These findings suggest that the initialization of the stage's position plays a crucial role in the accuracy of movement.

Channel Co-Alignment

Channel co-alignment or co-registration refers to the process of aligning image data collected from multiple channels. This ensures that signals originating from the same location in the sample are correctly overlaid. This process is essential in multi-channel imaging to maintain spatial accuracy and avoid misinterpretation of co-localized signals. For a comprehensive guide on Channel Co-Registration, refer to the Chromatic aberration and Co-Registration the QuaRep Working Group 04.

Acquisition protocol

  1. Place 4 µm diameter fluorescent beads (TetraSpec Fluorescent Microspheres Size Kit, mounted on a slide) on the microscope stage.

  2. Center an isolated bead under a high-NA dry objective.

  3. Crop the acquisition area to only visualize one bead but keep it large enough to anticipate a potential chromatic shift along the XY and Z axes

  4. Acquire a multi-channel Z-stack

    I usually acquire all available channels, with the understanding that no more than seven channels can be processed simultaneously. If you have more than 7 channels you can either split them into sets smaller or equal to 7 keeping one reference channel in all sets.

  5. Acquire 3 datasets

  6. Repeat for each objective

 Processing

You should have acquired several multi-channel images that now need processing to yield meaningful results. To process them, use the Channel Co-registration analysis feature of the MetroloJ_QC plugin for FIJI. For more information about the MetroloJ_QC plugin please refer to manual available at the MontpellierRessourcesImagerie on GitHub.

  1. Open FIJI.
  2. Load your image by dragging it into the FIJI bar.
  3. Launch the MetroloJ_QC plugin by navigating to Plugins > MetroloJ QC.
  4. Click on Channel Co-registration Report.
  5. Enter a Title for your report.
  6. Type in your Name.
  7. Click Microscope Acquisition Parameters and enter:
    1. The imaging modality (Widefield, Confocal, Spinning-Disk, Multi-photon) 
    2. The Numerical Aperture of the Objective used.
    3. The Refractive index of the imersion media. (Air: 1.0; Water 1.33; Oil: 1.515)
    4. The Emission Wavelength in nm for each channel.

  8. Click OK

  9. Click Bead Detection Options
    1. Bead Detection Threshold: Legacy
    2. Center Method: Legacy Fit ellipses

  10. Enable Apply Tolerance to the Report and Reject co-registration ratio above 1.

  11. Click File Save Options.

  12. Select Save Results as PDF Reports.

  13. Select Save Results as spreadsheets.

  14. Select Save result images.
  15. Select Open individual pdf report(s).

  16. Click OK.

  17. Repeat steps 4 through 16 for each image you have acquired

 Good to know: Settings are preserved between Channel Co-alignment sessions. You may want to process images with the same objective and same channels together.

This will generate detailed results stored in a folder named Processed. The Processed folder will be located in the same directory as the original images, with each report and result saved in its own sub-folder.

The following script for R processes channel co-alignment results generated by MetroloJ_QC Channel Co-registration function Process Chanel Co-Registration Results.R.

To use it, simply drag and drop the file into the R interface. You may also open it with RStudio and Click the Source button. The script will:

  • Prompt the user to select an input folder.
  • Load all _results.xls files located in the selected folder and subfolder
  • Process and merge all results generated by the MetroloJ_QC Channel Co-registration.
  • Split the filenames using the _ character as a separator and create columns named Variable-001, Variable-002, etc. This will help in organizing the data if the filenames are formatted as Variable-A_Variable-B, for example.
  • Save the result as Channel_Co-Registration_Merged-Data.csv in an Output folder on the user's Desktop

 

After much work, I ended up writing my own script for ImageJ/FIJI to detect and compute Channel Registration Data. This script is available here Channel_Registration_Batch-TrackMate_v12.py

To use it, simply drop the script into the FIJI toolbar and click Run.

  1. If images are already opened it will process them

  2. If no images are opened it will prompt to select a folder and will process all images available in the folder and subfolders

  3. It reads files metadata or load information from previously run instances

  4. The first image is pre-detected and setttings are displayed in a dialog

  5. The dialog will keep being displayed until the detection is properly done (1 spot per channel)

  6. Once the detection is correct it will compute the channel registration information
  7. Data is saved as CSV files Channel-Registration_Essential-Data_Merged and Channel-Registration_Full-Data_Merged in an Output directory on the user desktop
  8. In batch mode the settings set up for the first image are re-used for subsequent images unless discrepency is found or detection fails

Importantly this script does not use the same exact method than MetroloJ_QC Channel Registration algorithm:

The Nyquist sampling Ratio and the Resolution are caculated for both Lateral and Axial axes the same way than MetroloJ_QC does. However, if the Nyquist sampling Ratio is above 1 we compute an Maximal Achievable Resolution (named Lateral Practical Resolution and Axial Practical Resolution) by multiplying the Theoretical Resolution by the Nyquist Ratio.

We use the Pratical Resolutions to compute the Colocalization Ratios. We calculate 3 Ratios:

  • Lateral Colocalization Ratio: Distance in the XY plan between the spots idenfitifed in Ch2 - Ch1 divided by half of the Practical Lateral Resolution.
    RatioLateral = Dxy / ( PraticalResolutionLateral / 2 )
  • Axial Colocalization Ratio: Distance in the Z plan between the spots idenfitifed in Ch2 - Ch1 divided by half of the Practical Axial Resolution.
    RatioAxial = Dz / ( PraticalResolutionAxial / 2 )
  • 3D Colocalization Ratio: The Ratio is calculated by dividing the 3D Distance betwen the spots identified in Ch2 - Ch1 divided by a Reference Distance. The Reference Distance is the length between the spot 1 and a point projected from the line formed by Spot 1 and Spot 2 on the ellipse centered on Spot 1 with semi-minor axis being half of the practical lateral resolution and semi major axis as half of the practical axial resolution. The coordinates of the Reference Point is computed iteratively by moving along the Spot1-Spot2 line and minimizing the distance to the ellipse surface. It stops when it is on the surface.
    Ratio3D = Distance3D / ( DistanceRef)

Metrics

  • 3D Colocalization Ratio: A ratio above 1 indicates that the Spot of Channel 2 is further away than the effective resolution of the system. Lateral and Axial Colocalization ratios should be checked.
  • Lateral and Axial Colocalization Ratios: A ratio above 1 indicates that the Spot of Channel 2 is further away than the effective resolution of the system in the respective plan
    • A Lateral Colocalization Ratio above 1 indicates that the Ch2 image should be shifted in X and Y by the values indicated by the Pixel Shift Table.
    • An Axial Colocalization Ratio above 1 indicates that the Ch2 image should be shifted in Z by the values indicated by the Pixel Shift Table.

This method is the approach I am using here.

Results

The following spreadsheet provides a dataset that can be manipulated with a pivot table to generate informative graphs and statistics Channel_Co-registration_Template.xlsx.

Metrics

  • The Nyquist ratio evaluates how closely the images align with the Nyquist sampling criterion. It is calculated as: Nyquist Ratio = Pixel Dimension / Nyquist Dimension
    • A ratio of 1 indicates that the image acquisition complies with the Nyquist criterion.
    • A ratio above 1 signifies that the pixel dimensions of the image exceed the Nyquist criterion.
    • A ratio below 1 is the desired outcome, as it ensures proper sampling.
  • The Co-Registration Ratios measure the spatial alignment between two channels by comparing the distance between the centers of corresponding beads in both channels to a reference distance. The reference distance is defined as the size of the fitted ellipse around the bead in the first channel.
    • A ratio of 1 means the center of the bead in the second channel is located on the edge of the ellipse fitted around the bead in the first channel.
    • A ratio above 1 indicates the center of the bead in the second channel lies outside the ellipse around the first channel's bead center.
    • A ratio below 1 is the desired outcome, indicating that the center of the bead in the second channel is within a range smaller than the system's 3D resolution.

This method is the approach used in the MetroloJ QC Channel Co-Registration function.

Let's look at the 3D Colocalization Ratio for all pairs of channels.

For the 2x Objective we see that the 3D Colocalization Ratio is above 1 for the DAPI x GFP and DAPI x Cy5 pairs. This indicates that the chromatic shift is higher than the effective resolution of the system. Correction should be applied to images after acquisition. It somethimes possible to correct it before acquisition directly in the acquisition software. The correction values are provided by the Pixel Shift tables. Values highlighted correspond to a 3D Colocalization Ratio above 1.

 

These results shows a widefield instrument using a quadband pass filter: A single cube filtering 4 wavelengths. This instrument also possess individual filter cubes. Obviously the Colocalization Ratio are higher because of the mechanical shift induced by the filter turret.

With the corresponding Pixel Shift Table





Why should you care? Well when you are acquiring a multi-channel image you might see a significant shift between the two channels. This is particularly true for the combination of DAPI and Cy3 channels with the 10x Objective.


Report the Pixel Shift Table For each objective and each filter combination. This table can (should) be used to correct a multi-channel image by displacing the Channel 2 relative to the Channel 1 by the XYZ pixel coordinates indicated.




Channel_2
ObjectiveChannel_1AxisDAPIGFPCy3Cy5
2xDAPIX
0.89-0.14-0.35
Y
0.191.632.00
Z
0.893.671.58
GFPX-0.89
-1.04-1.25
Y-0.19
1.441.81
Z-0.89
2.780.70
Cy3X0.141.04
-0.21
Y-1.63-1.44
0.37
Z-3.67-2.78
-2.08
Cy5X0.351.250.21
Y-2.00-1.81-0.37
Z-1.58-0.702.08
10xDAPIX
0.46-0.85-1.16
Y
0.501.792.27
Z
4.224.441.91
GFPX-0.46
-1.31-1.61
Y-0.50
1.291.77
Z-4.22
0.22-2.31
Cy3X0.851.31
-0.30
Y-1.79-1.29
0.48
Z-4.44-0.22
-2.53
Cy5X1.161.610.30
Y-2.27-1.77-0.48
Z-1.912.312.53
20xDAPIX
0.58-0.77-1.06
Y
0.131.231.54
Z
3.313.952.09
GFPX-0.58
-1.35-1.64
Y-0.13
1.101.41
Z-3.31
0.64-1.22
Cy3X0.771.35
-0.29
Y-1.23-1.10
0.31
Z-3.95-0.64
-1.86
Cy5X1.061.640.29
Y-1.54-1.41-0.31
Z-2.091.221.86
63xDAPIX
0.13-1.52-2.03
Y
0.131.191.66
Z
0.791.310.93
GFPX-0.13
-1.65-2.16
Y-0.13
1.061.53
Z-0.79
0.520.13
Cy3X1.521.65
-0.51
Y-1.19-1.06
0.47
Z-1.31-0.52
-0.39
Cy5X2.062.220.51
Y-1.73-1.59-0.47
Z-0.92-0.120.39


Conclusion


Legend (Wait for it)...

For a comprehensive guide on Detectors, refer to the Detector Performances of the QuaRep Working Group 02.

Acquisition protocol 


 Results


Conclusion


Legend (Wait for it...) dary

For a comprehensive guide on Lateral and Axial Resolution, refer to the Lateral and Axial Resolution of the QuaRep Working Group 05.

Acquisition protocol 


 Results


Conclusion



List of Templates

  File Modified
Microsoft Excel Spreadsheet Channel_Co-registration_Data_Template.xlsx Dec 25, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Channel_Co-registration_Tempalte.xlsx Dec 20, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Channel_Co-registration_Template.xlsx Dec 25, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Field_Uniformity_Template.xlsx Dec 18, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Illumination_Linearity_Template.xlsx Dec 19, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Illumination_Long-Term_Stability_Log.xlsx Jan 06, 2025 by Nicolas Stifani
Microsoft Excel Spreadsheet Illumination_Maximum Power Output_Template.xlsx Jan 06, 2025 by Nicolas Stifani
Microsoft Excel Spreadsheet Illumination_Stability_Template.xlsx Jan 06, 2025 by Nicolas Stifani
Microsoft Excel Spreadsheet Illumination_Uniformity_Template.xlsx Dec 18, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Illumination_Warmup Kinetic_Template.xlsx Dec 19, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Illumination Power Linearity_Template.xlsx Dec 17, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Illumination Stability_Template.xlsx Dec 18, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Illumination Warmup Kinetic_Template.xlsx Dec 17, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Maximum Illumination Power Output_Template.xlsx Dec 18, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Objective and cube transmittance_Template.xlsx Jan 06, 2025 by Nicolas Stifani
Microsoft Excel Spreadsheet Objective and cube transmittance.xlsx Dec 12, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Stage-Repositining_Diagnostic-Data.xlsx Dec 26, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Stage Repositioning_Template.xlsx Jan 06, 2025 by Nicolas Stifani
Microsoft Excel Spreadsheet Stage-Repositioning_Template.xlsx Dec 30, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet Stage Repositioning Dispersion_Template.xlsx Dec 30, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet XY Repositioning Accuracy_Template_All-Files.xlsx Dec 15, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet XY Repositioning Accuracy_Template.xlsx Dec 19, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet XYZ_Repositining_Diagnostic_Data.xlsx Dec 19, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet XYZ_Repositioning-Accuracy_Template.xlsx Dec 19, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet XYZ_Repositioning Dispersion_Template.xlsx Dec 19, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet XYZ_Repositioning Dispersion_Traveled Distance_Template.xlsx Dec 19, 2024 by Nicolas Stifani
Microsoft Excel Spreadsheet XYZ Drift Kinetic_Template.xlsx Dec 19, 2024 by Nicolas Stifani

References

The information provided here is inspired by the following references:

Plateformes Scientifiques CIB               Biologie structurale    |    Cytométrie    |    Microscopie    |    Histologie   |    Instruments et Services

Tiling and Stiching
Nicolas Stifani posted on Jan 12, 2024

A microscope can capture a defined area of a sample. This area is called Field-of-View (FOV) and depends on the optical configuration and microscope acquisition device. This is a limiting feature of microscopy. To be able to observe with a higher resolution the total visualized area is reduced. This can be an issue when trying to visualize feature that are bigger than the FOV.

One way to deal with this issue is to acquire multiple images and stitch them together after acquisition. Instead of acquiring adjacent FOV it is best to have partially overlapping regions. These regions will help to stitch images together.

While many softwares provide proprietary stitching solution we will focus here on the free and versatile plugin for ImageJ named Grid Collection Stitching.

Developed by Stephan Preibisch, this plugin is also part of FIJI distribution of ImageJ.

Stitching process

Stitching usually occurs in 3 steps:

  1. The first is the "layout" which finds the adjacent images for each given image. This step approximatively place the images in relation to each other.
  2. The second step finely transform (rotation, translation) one image to the adjacent ones. It matches detected features in one image to the same feature in the adjacent image
  3. The last step blends the images so the results appears smooth


Layout

Three pieces of information can be used to define the layout.

1. Images metadata

Modern microscopes use motorized stages to move the sample in X and Y. These coordinates can be stored in the image metadata and used during stitching. Knowing the approximate position of each tile greatly help stitching as you just have to compute the fine matching between the different images.

2. Tiles configuration and acquisition order

if you have 25 images (Image 1, Image 2,....) and know that it comes from a 5 x 5 acquisition from the top left to the bottom righ, by row from left to right; then you can quickly place your images to their approximate positions. Sometimes the tile configuration is directly saved into the file names (Image X1Y1, Image X2Y1 etc.), this can also be used to define the approximate tile layout

3. Images themselves

The data in the image can also be used to define the layout. It requires computing power as it usually parse all possible pairwise combination and compute a correlation coefficient. It then matches images with highest correlation.

Transformation

Once the layout is defined, the images need to be finely adjusted one to another. Because microscopes are not perfect some translation and rotation can be used to finely match identified features in adjacent images. To do this images are usually acquired with a 10 to 20% overlapping region. This region will be used to finely match adjacent images.

Blending

A blending can be applied to the overlapping region to ensure a smooth tiled result.


Protocol

  • Open up FIJI
  • Open the Grid/Collection stitching plugin Menu Plugins>Stitching>Grid/Collection stitching

Your files are saved under Tile_x001_y001.tif, you know the grid size and the percentage overlap

  • Type: Filename defined position
  • Order: Defined by filename
  • Click OK
  • Indicate the grid size (for example 5x5 if you have 25 images)
  • Under directory click Browse
  • Select the folder containing your files
  • Click Choose
  • Under file names for tiles Tile_x{xxx}_y{yyy}.tif

Several options are available I recommend using the following:

  • Add tiles as ROIs (to check tiling quality
  • Compute Overlap
  • Display fusion


Your files are saved under Tile_001.tif, you know the grid size, the acquisition order and the percentage overlap

  • Type: Grid: Column by column
  • Order: Down & Right
  • Click OK
  • Indicate the grid size (for example 5x5 if you have 25 images)
  • Under directory click Browse
  • Select the folder containing your files
  • Click Choose
  • Under file names for tiles Tile_{iii}.tif

Several options are available I recommend using the following:

  • Add tiles as ROIs (to check tiling quality
  • Compute Overlap
  • Display fusion
  • Subixel accuracy


Your are capturing images with a manual stage. If you read this before the acquisition I would suggest to acquire your tiles using a given size and scheme (for example 3x3 snake horizontal right). This will allow to use the process above (except Type Grid: Snake by rows). Make sure to have some overlap between images to be able to finely place them. Since you are probably reading this after your acquisition you would have file saved as Tile_001.tif... but you do not know the grid size or the acquisition order nor the percentage overlap

  • Type: Unknown position
  • Order: All files in Directory
  • Click OK
  • Under directory click Browse
  • Select the folder containing your files
  • Check Confirm files

Several options are available I recommend using the following:

  • Add tiles as ROIs (to check tiling quality
  • Ignore Z stage position
  • Subpixel accuracy
  • Display fusion
  • Computation parameters: Save computation time


If you choose Type: Grid the plugin except sequential continuous files i i+1 etc...

If you choose Positions from File the plugin expect one single file multiseries file. To Combine several files into one single T series, open the images individually or all together Then combine them as a stack

Image>Stacks>Images to Stack

Check use tiltes as labels

then convert the Stack to a T serie

Image>Hyperstacks>Stacks to Hyperstacks

Slices (z)=1 Frames (t)=number that was in z and you have replaced by 1

In my experience proprietary softwares do not encode X and Y values in the metadata properly so this method is not often used



Type: Positions from file
Use this is you want to use the image metadata to define the tile positions. This only works if you have one single input file with all the tiles inside. In my experience this works when the file issaved under the acquisition software proprietary format.
You can also use this type of stitching if you have an additional text files defining the position of each image. You can also create this file yourself 


Example of Tile Configuration File
# Define the number of dimensions we are working on
dim = 3
# Define the image coordinates (in pixels)
img_01.tif; ; (0.0, 0.0, 0.0)
img_05.tif; ; (409.0, 0.0, 0.0)
img_10.tif; ; (0.0, 409.0, 0.0)
img_15.tif; ; (409.0, 409.0, 0.0)
img_20.tif; ; (0.0, 818.0, 0.0)
img_25.tif; ; (409.0, 818.0, 0.0)


Notes

The higher the overlap the more computing power required.
Percentage overlap is approximate. Starts low and increase until result is satisfying

It is much easier to stitch when the layout is known.

It is much easier to stitch images when there are many visible features: slide of tissue is easier than sparse cell culture; bright field images are easier than fluorescence images 


Renaming your files

You can easily rename files on a mac using Automator.

On a PC you can use Bulk Rename Utility



More information

https://imagej.net/plugins/image-stitching




Quality Control
Nicolas Stifani posted on Dec 04, 2023

Introduction

Microscopy is an approximation of reality. The point spread function is the better example of this: A single point will appear as a blurry ellipse using photonic microscopy. This transformation depends on the optical components which varies with time.

Quality control monitors this transformation over time.





Download MetroloJ QC from GitHub - MontpellierRessourcesImagerie/MetroloJ_QC

Download iText Library v 5.5.13 from https://repo1.maven.org/maven2/com/itextpdf/itextpdf/5.5.13.2/itextpdf-5.5.13.2.jar

Download ImageJ Download (imagej.net) or FIJI Fiji Downloads (imagej.net)


Open ImageJ or FIJI

Install MetroloJ QC and iTextPDF by dropping the .jar file into ImageJ status bar



Start up the microscope

Follow the set up procedure (load the test sample and focus with the lowest magnification objective).

Adjust to obtain a Kholer illumination

Remove the sample from the microscope

Nikon Ti2 without occular a shadow appears on the upper left corner of the FOV



Take a BF image for Widefield illumination Homogeneity

Analyze the image using MetroloJ QC and Check centering accuracy and total homogeneity

4x


Here the centering is off to the left side and the intensity is falling below 30%.


replace the illumination arm by the application screenlight


removing the objective:


If image is similar then the issue comes from the detection side.

Test all items on the detection path: Objective, filter cubes, lightpath selector, camera

Objective 20x-075

It improve the hopmogeneity but centering is still off

60x1.4 gives a perfect homogeneity



Repeat for each objective

Change objective

Adjust Kholer

Take a BF image for Widefield illumination Homogeneity


If illumination is not homogeneous then 

Conclusion: Camera may be re-aligned to obtain an homogenous illumination in bright field.


5408nm laser line not working used for YFP. Laser is shining outside the fiber but not reachingh the objective. Has the cube been changed?


Microscopy consummables
Nicolas Stifani posted on Mar 03, 2023


Immersion Oil

Very important component. Mostly used for oil immersion objectives, the refractive index should match the RI of glass from the coverslip and the objectives used. Different types or grades are available A (Low viscosity), B (high viscosity), N, F, FF (for fluorescence) etc... as well as different viscosity. Because the refractive index (and viscosity) vary with the temperature you should buy immersion oil matching the room temperature. Usually 23C but you could also need 30C and 37C.

To date Cargille provide the best options

Cargille

  • Cargille Immersion Oil type FF #16212, 16 oz (473mL), 94$, 0.2$/mL RI=1.48
  • Cargille Immersion Oil type HF #16245, 16 oz (473mL), 94$, 0.2$/mL RI=1.51
  • Cargille Immersion Oil type LDF #16241, 16 oz (473mL), 94$, 0.2$/mL RI = 1.51

Extremely low fluorescence is achieved by Type LDF and Type HF. Type FF is virtually fluorescence-free, though not ISO compliant. Type HF is slightly more fluorescent than Type LDF, but is halogen-free.

Thorlabs

  • MOIL-30 Olympus Type F, 30mL, $84, 2.8$/mL
  • MOIL-20LN Leica Type N, 20mL, $75, 3.75$/mL
  • OILCL30 Cargille Type LDF, 30mL, $28, 1.1$/mL RI=1.51
  • MOIL-10LF Leica Type F, 10mL, $59, 5.9$/mL

https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=5381

Zeiss

Edmund Optics

Lens Cleaner

Use a lens cleaner to clean objective lenses from oil.

Tiffen

Tiffen is a great product designed for camera lenses but it works great for microscope optics as well

Edmund Optics

  • Lens Cleaner #54-828 (8oz, 236mL) 14$; 6 cts/mL
  • Purosol #57-727 (4oz, 115mL), 29$, 25 cts/mL

Zeiss, Nikon, Olympus, Leica

  • Home made recipe: 85% n-hexane analytical grade, 15% isopropanol analytical grade

Lens tissue

Edmund Optics

  • Lens Tissue #60-375 500 sheets 36$ 7cts/sheet 

Thorlabs

  • Lens Tissue #MC-50E 1250 sheets 93$ 7.6cts/sheet

Tiffen

  • Lens Tissue #EK1546027T 250 sheets 112$ 44cts/sheet

Liquid light guides

Excelitas

  • 3mm core diameter, 1.5m length sold by Digikey #1601-805-00038-ND 682$

Thorlabs

  • 3mm core diameter, 1.2m length #LLG03-4H 410$
  • 3mm core diameter, 1.8m length #LLG03-6H 490$

Edmund Optics

  • 3 mm core diameter, 1.8m length #53-689 3mm 700$
  • Adapters available #66-905

Microscope world

  • Liquid Light Guide 3mm core diameter, 1.5m length #805-00038 445$

Bulbs

AVH Technologies

HXP R 120W/45C 780$


OSRAM

XP R 120 W/45 #69119

Others consumables

  • Cotton swab
  • Absorbent polyester swabs for cleaning optical components, Alpha, Clean Foam or Absorbond series TX743B) from www.texwipe.com
  • Rubber Blower GTAA 1900 from Giottos www.giottos.com
Image Analysis
Nicolas Stifani posted on Jul 09, 2022

Let's say you have many images taken the same way from two different samples: One Control group and One test group.


What will/should you do to analyse them?


The first thing I would do will be to open a random pair of image (one from the control and one from the test) and have a look at them...

Control                                                   Test


Then I would normalize the brightness and contrast for each channel and across images to make the display settings  the same for both images.

ImageJ Macro to Equalize Brightness and Contrast for all opened images 
// This macro was created by nstifani@gmail.com
// Feel free to reach out for any question or improvement suggestions
// This macro will look for Min and Max values on all open images and then apply the Min and the Max to adjust the display brightness of all opened images.
 
// Create Arrays to store image info
ListOfCh=newArray(nImages);
ListOfSlices=newArray(nImages);
ListOfFrames=newArray(nImages);
 
// Get the dimensions of the opened Images
for(ImageI=1; ImageI<nImages+1; ImageI++){
selectImage(ImageI);
getDimensions(width, height, channels, slices, frames);
ListOfCh[ImageI-1]=channels;
ListOfSlices[ImageI-1]=slices;
ListOfFrames[ImageI-1]=frames;
}// end for ImageI
 
 
// Get some statistics
Array.getStatistics(ListOfCh, MinCh, MaxCh, MeanCh, StdDevCh);
Array.getStatistics(ListOfSlices, MinSlice, MaxSlice, MeanSlice, StdDevSlice);
Array.getStatistics(ListOfFrames, MinFrame, MaxFrame, MeanFrame, StdDevFrame);
 
 
// Process all chanels using two functions
for(ChI=1; ChI<MaxCh+1;ChI++){
MinAndMax=GetBnCValues(ChI);
MinMin=MinAndMax[0];
MaxMax=MinAndMax[1];
ApplyBnC(ChI,MinMin, MaxMax);
}
 
 
function GetBnCValues(ChI) {
ListOfMin=newArray(nImages);
ListOfMax=newArray(nImages);
 
// Measure Min and Max for all open images
for(ImageI=1; ImageI<nImages+1; ImageI++){
selectImage(ImageI);
Stack.setChannel(ChI);
resetMinAndMax();
//run("Enhance Contrast", "saturated=0.35");
getMinAndMax(min, max);
ListOfMin[ImageI-1]=min;
ListOfMax[ImageI-1]=max;
}// end for ImageI
 
// Get Statistics
Array.getStatistics(ListOfMin, MinMin, MaxMin, MeanMin, StdDevMin);
Array.getStatistics(ListOfMax, MinMax, MaxMax, MeanMax, StdDevMax);
return newArray(MinMin, MaxMax);
}
 
function ApplyBnC(ChI, MinMin, MaxMax) {
for(ImageI=1; ImageI<nImages+1; ImageI++){
selectImage(ImageI);
Stack.setChannel(ChI);
setMinAndMax(MinMin, MaxMax);
}// end for ImageI
ImageJ Macro Equalize BnC for all images in a folder 
//It might be easier to process images from a folder
//You would need to customize this path to your computer
InputDir="/Users/nicolas/Desktop/Input/";
 
// You could also get a prompt to select the InputDir
// InputDir = getDirectory("Choose a Directory ");
 
 
// And to save results in an output folder
// You would need to customize this path to your computer
//OutputPath="/Users/nicolas/Desktop/Output/";
 
//You could also create a new folder based on the name of the input folder
ParentPath=File.getParent(InputDir);
InputDirName=File.getName(InputDir);
OutputDir=ParentPath+File.separator+InputDirName+"_Results";
i=1;
 
while(File.exists(OutputDir)){
OutputDir=ParentPath+File.separator+InputDirName+"_Results"+"-"+i;
i++;
}
File.makeDirectory(OutputDir);
OutputPath=OutputDir+File.separator;
 
 

function GetBnCValues(ChI) {
ListOfMin=newArray(ListFile.length);
ListOfMax=newArray(ListFile.length);
  
// Measure Min and Max for all open images
for(Filei=0; Filei<ListFile.length; Filei++){
FilePath=InputDir+ListFile[Filei];
open(FilePath);
ImageName=getTitle();
//selectWindow(ImageName);
Stack.setChannel(ChI);
resetMinAndMax();
run("Enhance Contrast", "saturated=0.35");
getMinAndMax(min, max);
ListOfMin[Filei]=min;
ListOfMax[Filei]=max;
selectWindow(ImageName); run("Close");
}// end of for FileI
// Get Statistics
Array.getStatistics(ListOfMin, MinMin, MaxMin, MeanMin, StdDevMin);
Array.getStatistics(ListOfMax, MinMax, MaxMax, MeanMax, StdDevMax);
return newArray(MinMin, MaxMax);
}// End of GetBnCValues function



function ApplyBnC(ChI, MinMin, MaxMax) {
for(Filei=0; Filei<ListFile.length; Filei++){
FilePath=InputDir+ListFile[Filei];
open(FilePath);
ImageName=getTitle();
//selectWindow(ImageName);
Stack.setChannel(ChI);
setMinAndMax(MinMin, MaxMax);
saveAs("Tiff", OutputPath+ImageName);
selectWindow(ImageName); run("Close");
}// end for Filei
}//end of function

 
ListFile=getFileList(InputDir);
run("Set Measurements...", "area mean standard modal min centroid center perimeter bounding fit shape feret's integrated median skewness kurtosis area_fraction stack display redirect=None decimal=3");
run("Clear Results");
 
// It might be faster to work in batchmdoe
setBatchMode(true);
 

// Create Arrays to store image info
ListOfCh=newArray(ListFile.length);
ListOfSlices=newArray(ListFile.length);
ListOfFrames=newArray(ListFile.length);
 
for (Filei=0; Filei<ListFile.length; Filei++){
FilePath=InputDir+ListFile[Filei];
open(FilePath);
ImageName=getTitle();
//selectWindow(ImageName);
getDimensions(width, height, channels, slices, frames);
ListOfCh[Filei]=channels;
ListOfSlices[Filei]=slices;
ListOfFrames[Filei]=frames;
selectWindow(ImageName); run("Close");
}// end for FileI


// Get some statistics
Array.getStatistics(ListOfCh, MinCh, MaxCh, MeanCh, StdDevCh);
Array.getStatistics(ListOfSlices, MinSlice, MaxSlice, MeanSlice, StdDevSlice);
Array.getStatistics(ListOfFrames, MinFrame, MaxFrame, MeanFrame, StdDevFrame);


//for (Filei=0; Filei<ListFile.length; Filei++){
//FilePath=InputDir+ListFile[Filei];
//open(FilePath);
//ImageName=getTitle();
////selectWindow(ImageName);

// Process all chanels using two functions
for(ChI=1; ChI<MaxCh+1;ChI++){
MinAndMax=GetBnCValues(ChI);
MinMin=MinAndMax[0];
MaxMax=MinAndMax[1];
ApplyBnC(ChI,MinMin, MaxMax);
}


Then I would look at each channel individually and use a LUT Fire to have better appreciate the intensities.


Macro Apply LUT Fire to all opened images 
for(ImageI=1; ImageI<nImages+1; ImageI++){
selectImage(ImageI);
getDimensions(width, height, channels, slices, frames);
for(ChI=1; ChI<channels+1;ChI++){
Stack.setChannel(ChI);
Property.set("CompositeProjection", "null");
Stack.setDisplayMode("color");
run("Fire");
}
}


Finally I will use the syncrhonize Windows features to navigate the two images at the same time

ImageJ>Windows>Synchronize Windows

ou dans une macro

run("Synchronize Windows");



                           Control  Ch1                                         Test Ch1                        

                           Control  Ch2                                         Test Ch2                        


To my eyes the Ch1 and Ch2 are slighly brighter in the Test conditions.

Here we can't compare Ch1 to Ch2 because the range of display are not the same. It seems than Ch2 is much weaker than Ch1 but it is actually not accurate. The best way to sort this is to add a calibration bar to each channel ImageJ>Tools>Calibration Bar. The calibration bar is non-destructive as it will be added to the overlay. To "print" it on the image you can flatten it on the image ImageJ>Image>Overlay>Flatten.





                               Control  Ch1                                         Control Ch2       


If you apply the same display to Ch1 and Ch2 then you can see that Ch1 overall more intense while Ch2 has few very strong spots.


                         Control  Ch1                       Control Ch2 with same display than Ch1



Looking more closely

In Ch1 we can see that there is some low level intensity and a high level circular foci whereas in Ch2 there is a bean shaped structure. In the example below the Foci seems stronger in the control than the Test condition.


                       Control Ch1                                            Test Ch1




                       Control Ch2                                                    Test Ch2


But we need obviously to do some quantification to confirm or infirm these first observations.


The first way to address it would be in a bulk fashion: By measuring the mean intensity for example for all the images.



Macro ImageJ Collect Global measurements 
 

//It might be easier to process images from a folder
//You would need to customize this path to your computer
InputDir="/Users/nicolas/Desktop/Input/";

// You could also get a prompt to select the InputDir
// InputDir = getDirectory("Choose a Directory ");


// And to save results in an output folder
// You would need to customize this path to your computer
//OutputPath="/Users/nicolas/Desktop/Output/";

//You could also create a new folder based on the name of the input folder
ParentPath=File.getParent(InputDir); 
InputDirName=File.getName(InputDir);
OutputDir=ParentPath+File.separator+InputDirName+"_Results";
i=1;

while(File.exists(OutputDir)){
OutputDir=ParentPath+File.separator+InputDirName+"_Results"+"-"+i;
i++;
}
File.makeDirectory(OutputDir);
OutputPath=OutputDir+File.separator;


//Then you can measure all values for all ch and all images

ListFile=getFileList(InputDir); 
run("Set Measurements...", "area mean standard modal min centroid center perimeter bounding fit shape feret's integrated median skewness kurtosis area_fraction stack display redirect=None decimal=3");
run("Clear Results");

// It might be faster to work in batchmdoe
setBatchMode(true);


for (Filei=0; Filei<ListFile.length; Filei++){
FilePath=InputDir+ListFile[Filei];
open(FilePath);
ImageName=getTitle();
selectWindow(ImageName);
run("Select None");
run("Measure Stack...");
selectWindow(ImageName); run("Close");
}
selectWindow("Results");
saveAs("Text", OutputPath+"Overall_Measurements.csv");
selectWindow("Results"); run("Close");
  


If all works fine you should have a CSV file you can open with your favorite spreadsheet applications. This table should give one line per image and all available measurements for the whole image and for each channel of the image. Of course some measurements will be all the same because the images were taken in the same way.

What to do with the file? Explore the data and see if there is any relevant information.

My view would be to use a short script in R to plot all the data and to some basic statistics


R Plot ImageJ Global Measurements 
# Prompt for the input CSV file from ImageJ measurements
Input<-file.choose()

# You must cange this path to match your computer
Output<-"/Users/nicolas/Desktop/Output/"
#Output <- "C:\\Users\\stifanin\\OneDrive - Universite de Montreal\\Bureau\\Output\\"


data<-read.csv(Input)

require("ggpubr")

List<-strsplit(as.character(data$Label), ':')

#Add Filename and Group to the data
data$Filename<-as.factor(sapply(List, "[[", 1))
Group<-strsplit(as.character(data$Filename), '_')
data$Group<-as.factor(sapply(Group, "[[", 1))
# MAke Ch as a factor
data$Ch<-as.factor(data$Ch)

#Create a list of plots
ListofPlots<-list()
i=1;

for(ColI in 3:(ncol(data)-2)){
  if(!is.factor(data[[ColI]])){
    for (ChI in 1:nlevels(data$Ch)){
   Graph<-   ggboxplot(
        data[data$Ch==ChI,], x = "Group", y = colnames(data[ColI]),
        color = "Group", palette = c("#4194fa", "#db51d4"),
        add = "jitter", title=paste0(colnames(data[ColI])," of Channel ", ChI)
      )+stat_compare_means(method = "t.test")
   ListofPlots[[i]]<-Graph
i=i+1
  }
  }
}

#Export the graphs
ggexport(
  plotlist = ListofPlots, filename = paste0(Output,"Graphs.pdf"))

This should give you a pdf file with one plot per page. You can scroll it and look at the data. p-values from t-test are indicated on the graphs. As you can see below the mean intensity in both Ch1 and Ch2 are higher in the test than the control. What does it mean?



It means than the average pixel intensity is higher in the test conditions than the control condition.

Other values that are significantly different:

  • Mean Intensity Ch1 and Ch2 (Control<Test)
  • Maximum Intensity of Ch1 (Control>Test) Brightest value in the image
  • Integrated Intensity of Ch1 and Ch2(Control<Test) It is equal to the Mean x Area
  • Median Ch1 and Ch2 (Control<Test)
  • Skew of Ch1 (Control>Test): The third order moment about the mean. Relate to the distribution of intensities. If=0 then intensities are symmetrically distributed around the mean. if<0 then distribution is asymmetric to the Left of the mean (lower intensities), if>0 then it is to the right (higher intensities).
  • Raw Integrated Intensities of Ch1 and Ch2 (Control<Test) Sum of all pixel intensities


Now we start to have some results and statistically relevant information about the data. The test condition have a higher mean intensity (integrated intensity, median and raw integrated intensities are all showing the same result) for both Ch1 and Ch2. This is surprising because I had the opposite impression while looking at the image with normalized intensities and LUT fire applied (see above). Another surprising result is the fact that Control images have a higher maximum intensity than Test images but only for Ch1. This is clearly seen in the picture above.

One thing that can explain these results is that the number of cells can be different in the control vs the test images. If there are more cells in one condition then there are more pixels stained (and less background) and the mean intensity would be higher not because the signal itself is higher in each cell but because there are more cell...

To solve this we need to count the number of cells per image. This can be done manually or using segmentation based of intensity.

Looking at the image it seems that Ch1 is a good candidate to segment each cell.

ImageJ Macro Segment and Measure 
// It might be easier to process all images ina folder
// You can specify this folder. You need to customize this path to your computer
InputDir="/Users/nicolas/Desktop/Input/";
 
// Or you can get a prompt to select the InputDir
// InputDir = getDirectory("Choose a Directory ");
 
 
// To save results in an output folder
// You can specify the OutputPath. You need to customize this path to your computer
//OutputPath="/Users/nicolas/Desktop/Output/";
 
//Or you can create a new folder based on the name of the input folder
//This is the method I prefer
ParentPath=File.getParent(InputDir);
InputDirName=File.getName(InputDir);
OutputDir=ParentPath+File.separator+InputDirName+"_Results";
i=1;
 
while(File.exists(OutputDir)){
OutputDir=ParentPath+File.separator+InputDirName+"_Results"+"-"+i;
i++;
}
File.makeDirectory(OutputDir);
OutputPath=OutputDir+File.separator;
File.makeDirectory(OutputPath+"Cropped Cells");
OutputCellPath=OutputPath+"Cropped Cells"+File.separator;

//// End of creating a new ouput folder

 
// Prepare some measurements settings and clean up the place
run("Set Measurements...", "area mean standard modal min centroid center perimeter bounding fit shape feret's integrated median skewness kurtosis area_fraction stack display redirect=None decimal=3");
run("Clear Results");


//Then you can start to process all images
 
ListFile=getFileList(InputDir);
 run("ROI Manager...");
// It might be faster to work in batchmode
setBatchMode(true);


 
for (Filei=0; Filei<ListFile.length; Filei++){
FilePath=InputDir+ListFile[Filei];
open(FilePath);
ImageName=getTitle();
ImageNameNoExt=File.getNameWithoutExtension(FilePath);
getDimensions(width, height, channels, slices, frames);

//Remove the Background
selectWindow(ImageName);
ImageNameCorrected=ImageNameNoExt+"_Corrected";
run("Duplicate...", "title=&mageNameCorrected duplicate");
rename(ImageNameCorrected);
selectWindow(ImageNameCorrected);
run("Subtract...", "value=500");
run("Subtract Background...", "rolling=22 sliding stack");

//Adjust the display
for(ChI=1; ChI<channels+1; ChI++){
Stack.setChannel(ChI);
resetMinAndMax();
run("Enhance Contrast", "saturated=0.35");
//setMinAndMax(500, 3000);
}
Stack.setChannel(1);
Property.set("CompositeProjection", "Sum");
Stack.setDisplayMode("composite");

//Combine both color for full cell segmentation
run("RGB Color");
rename("RGB");
run("Duplicate...", "title=Mask duplicate");
run("16-bit");
resetThreshold();
setAutoThreshold("Otsu dark no-reset");
setOption("BlackBackground", true);
run("Convert to Mask");
run("Open");
run("Dilate");
run("Fill Holes");
run("Analyze Particles...", "size=2-12 circularity=0.60-1.00 exclude add");
selectWindow("Mask");
saveAs("TIFF", OutputPath+ImageNameNoExt+"_Cell-Mask.tif");
MaskImage=getTitle(); 
selectWindow(MaskImage);run("Close");
selectWindow("RGB");
run("Remove Overlay");
run("From ROI Manager");
saveAs("TIFF", OutputPath+ImageNameNoExt+"_Segmentation-Control.tif");
ControlImage=getTitle(); 
selectWindow(ControlImage);run("Close");

selectWindow(ImageNameCorrected);
count = roiManager("count");
for (i = 0; i < count; i++) {
roiManager("select", i);
run("Measure Stack...", "channels slices frames order=czt(default)");
}
RoiManager.select(0);

selectWindow("Results");
saveAs("Results", OutputPath+ImageNameNoExt+"_Measurements_Cells.csv");
run("Clear Results");

selectWindow(ImageNameCorrected);
saveAs("TIFF", OutputPath+ImageNameNoExt+"_Background-removed.tif");
ImageCorrected =getTitle();
selectWindow(ImageCorrected);

NbROIs=roiManager("size");
AllROIs=Array.getSequence(NbROIs);
roiManager("Select", AllROIs);
OutputCellName=OutputCellPath+ImageNameNoExt+"_";
RoiManager.multiCrop(OutputCellName, " save tif");
roiManager("Save", OutputPath+ImageNameNoExt+"_Cell-ROIs.zip");
roiManager("Deselect");
roiManager("Delete");
selectWindow(ImageCorrected);run("Close");
selectWindow(ImageName);run("Close");

}// end for FileI


This macro starts to be a bit long but to summarize here are the steps:

  • Open each image
  • Remove the offset from the camera (500 GV) and apply a rolling ball background subtraction
  • Create a RGB composite regrouping both channels
  • Convert this RGB to a 16-bit image
  • Threshold the RGB image using the Otsu alogrythm
  • Process the binary to improve detection (Open, Dilate, Fill Hole)
  • Analyze particles to detect cells with size=2-12 circularity=0.60-1.00
  • Add the results to the ROI manager
  • Save the Mask for the control of segmentation
  • Save an RGB image with the detection overlay as a control for the good detection
  • Save the ROIs
  • Use the ROIs to collect all measurements available and save the result as a csv
  • Save the image with the background removed
  • Crop each ROI from the image with the Background removed to isloate each cell

Few notes:

The camera offset is a value the camera adds to avoid having negative value due to noise. The best way to measure it is to take a dark image and have the mean intensity of the image. If you don't have that in hand you can choose a value sligly lower than the lowest pixel value found in your images.

The rolling ball background subtraction is powerfull tool to help with the segmentation

The values for the Analyze particles detection are the tricky part here. How I choose them? I use the thresholded image (binary) and I select the average guy: the spot that looks like all the others. I use the wand tool to create a selection and then I do Analyze>Measure This will give me a good estimate of the size (area) and the circularity. Then I select the tall/skinny and the short/not so skinny guys: I use again the wand tool to select the spots that would be my upper and lower limits. This will give me the range of spot size (area) and circularity. This is really a critical step that should be performed by hand prior running the script. You should do this on few different images to make sure the values are good enough to account for the variability you will encounter.

Now it is time to check that the job was done properly.

Looking at the control Images the detection isn't bad at all.


The only thing missing are the individual green foci seen below. Those cells look different as the FOci is very strong but there is not much fluorescence elsewhere (no diffuse green and no red). I might discuss with the scientist to see if it is OK to ignore them. If not I would need to change the threshold values and the detection parameters but let's say it is fine for now.

So now you should have a list of files (images, ROIs, and CSV files). We will focus on the CSV files has they contain the number of cells we are looking for and a lot more information we can also use.

We will reuse the previous R script but will add a little part to merge all the CSV files from the input folder.


Script R Merge CSV files and plot all data 
#Select the Input and the Output folder
Input <- "/Users/nicolas/Desktop/Input_Results-1/"
Output<-"/Users/nicolas/Desktop/Output/"
require("ggpubr")
#Get the list of CSV files
FileList<-list.files(path=Input, full.names=TRUE, pattern=".*csv")

#Merge the file into one big data
for (FileI in 1:length(FileList)){
  data<-read.csv(FileList[FileI])
  if (FileI==1){
    BigData<-data
  }else{
    BigData<-rbind(BigData,data)
  }
  
}
# then the rest is the same than the previous script
data<-BigData

List<-strsplit(as.character(data$Label), ':')

#Add Filename and Group to the data
data$Filename<-as.factor(sapply(List, "[[", 1))
Group<-strsplit(as.character(data$Filename), '_')
data$Group<-as.factor(sapply(Group, "[[", 1))
# Make Ch as a factor
data$Ch<-as.factor(data$Ch)

write.csv(data, paste0(Output,"Detected-Cells_All-Measurements.csv"), row.names = FALSE)

#Create a list of plots
ListofPlots<-list()
i=1;

for(ColI in 3:(ncol(data)-2)){
  if(!is.factor(data[[ColI]])){
    for (ChI in 1:nlevels(data$Ch)){
   Graph<-   ggboxplot(
        data[data$Ch==ChI,], x = "Group", y = colnames(data[ColI]),
        color = "Group", palette = c("#4194fa", "#db51d4"),
        add = "jitter", title=paste0(colnames(data[ColI])," of Channel ", ChI)
      )+stat_compare_means(method = "t.test")
   ListofPlots[[i]]<-Graph
i=i+1
  }
  }
}

#Export the graphs
ggexport(
  plotlist = ListofPlots, filename = paste0(Output,"Graphs_individual.pdf"), 
  verbose=TRUE
)


This script will save the merged data in a single csv file. Usign your favourite spreadsheet manager you will be able to create a table and summarize the data with a pivot table to get the number of cells per group.


ControlTest
12251360

There are slighly more cells in the test than in the control. If we look at the number of detected cells per image we can confirm that there are more cells in the test conditions than in the control conditions. There are two images that have less celss than others in the control. We can go back to the detection to check those.


Looking at the detection images we can confirm that the two images from the control have less cells, so it is not a detection issue.


Together these results show that there is no more cells in one condition than the other.

On avera between 60 and 65 cells are detected by image with a total of 1200 cells per condition detected.


Then we can look at the graphs and look for what is statistically different, here is the short list

  • Area Test>Control
  • Mean Ch2 Test>Control
  • Min Ch2 Test>Control
  • Max Ch1 Control>Test
  • Max Ch2 Test>Control
  • Perimeter Test>Control
  • Width and Height Test>Control
  • Major and Minor Axis Test>Control
  • Feret Test>Control
  • Integrated Density Ch1 et Ch2 Test>Control
  • Skew et Kurt Ch1 Control>Test
  • Raw Integrated density of Ch2 Test>Control
  • Roundness Control>Test


Looking at p-values (statistical significance) is good approach but it is not enough. We should also look at the biological significance of the numbers. For example the roundess is statiscially different between the control and test but this difference is really small. What does it mean biologically that the test cells are a tiny bit less circular. In this specific case : nothing much. So we can safely forget about it to focus on more important things. 


This can easily be done since we have generated a CSV file gathering all the data Detected-Cells_All-Measurements.csv


R Script Create Graph with Descriptive Statistics 
# Prompt for the input CSV file from ImageJ measurements
Input<-file.choose()

# You must cange this path to match your computer
Output<-"/Users/nicolas/Desktop/Output/"

data<-read.csv(Input)

data$Ch<-as.factor(data$Ch)



require("ggpubr")
#Create a list of plots
ListofPlots<-list()
i=1;

for(ColI in 3:(ncol(data)-2)){
  if(is.numeric(data[[ColI]]) || is.integer(data[[ColI]])){
    for (ChI in 1:nlevels(data$Ch)){
      Graph<-ggsummarystats(
        data[data$Ch==ChI,], x = "Group", y=colnames(data[ColI]),
        ggfunc = ggviolin, digits = 2,
        color = "Group", palette = c("#4194fa", "#db51d4"),
        summaries=c("n", "mean","sd","ci"), add="mean_sd", title=paste0(colnames(data[ColI])," of Channel ", ChI))
      ListofPlots[[i]]<-Graph
      i=i+1
    }
  }
}

#Export the graphs
ggexport(
  plotlist = ListofPlots, filename = paste0(Output,"Graphs_Individual_Descriptive Stats.pdf"), 
  verbose=TRUE
)


Then for all the variables that are statically different between control and test groups we can have a closer look at the data.

  • Area Control 5.03um2 ; Test 5.35um2 p-value = 1.7e−07
    This is a relatively small increase 6%. If we bring this back to the diameter it is even smaller 3% increase. Yet the most relevant value in my view is the volume because cells are spheres in the real life. This increase is about a 10% increase in the cell volume. This is relevant information. Cells in the Test conditions are 10% bigger than in the control condition. 
  • Mean Ch2 Control 1078 GV; Test 1197 GV p-value < 2.2e−16
    This means that the bean shape structure labelled by Ch2 is 10% brighter in the test cells than in the controls. Yet the segmentation was performed on the full cell using Ch1 as a proxy. This result prompt for a segmentation based on Ch2. Are the bean shaped bigger or brighter? 
  • Min Ch2 Control 198 GV; Test 213 GV p-value 1e−05
    The difference in the minimum of Ch2 represent about 7.5% increase. If the minimum is higher it suggests that there is a global increase in the fluorescent of Ch2 and not a redistribution (ie clusterization). The segmentation on the Ch2 might sort out which option is really occurring
  • Max Ch1 Control 10111 GV; Test 8393 GV p-value < 2.2e−16
    This difference represent a 17% decrease of the maximum intensity of Ch1. If we remember Ch1 has two fluorescent levels, low within the nuclei and an intense foci. Here the maximum represent the Foci only. So we can conclude that the Foci are less intense in the test compare to the control.
  • Max Ch2 Control 3522 GV; Test 3761 GV p-value = 0.00035
    Here we have the opposite situation where in the bean shape structure the maximum intensity is higher in  the test condition than the control 
  • Perimeter Control 8.45um ; Test 8.74um p-value = 1.7e−08
    Here we have another measurement about the size of the detected cells. Using the area we can estimate about a 3.1% perimeter increase. Here the increase is about 3.5%. This is consistant meaning the increase in area and perimeter are about the same.
  • Width and Height Control 2.49um; Test 2.57um p-value = 8e−07
    Again here another measurement of the size of the detected cells which are larger in the test than the control
  • Major and Minor axis length Control 2.73um; Test 2.84um  p-value = 2.7e−10
    Similar to above excepted that instead of measuring the width and heigh of a rectangle around the cell it is an ellipse fitting the cell. In my view this would be more relevant variable than the previous one but since all data converge to slightly larger cells in the test than the control we can focus on the most interesting ones (area or perimeter)
  • Feret Control 2.95um; Test 3.05um p-value = 5.1e−10
    This is the maximum distance between two points of the detected cells. It relates to the shape of the cell
  • Integrated Density Ch1 Control 11296 GV; Test 11886 GV p-value
    This is the product of area and mean intensity. We know that the area is larger but not the mean intensity. Yet the integrated density of Ch1 is sligtlhy larger (5%) in test vs control.  
  • Integrated Density Ch2 Control 5570 GV; Test 6548 p-value < 2.2e−16
    Here we have a 17% increase which represent the combination of increase of cell size and mean intensity of Ch2
  • Skew of Ch1 Control 2.47; Test 1.77 p-value < 2.2e−16
    The skewness refers to the asymmetrical shape of the distribution of the intensities. Since values are above 0 it means that the distribution is skewed towards the higher intensities. Since this number is lower in the test cells, it means that the distribution of Ch1 intensity is less asymmetric. This can be easily explained by looking at Ch1 which has a low intensity within the nuclei and a strong foci. The foci high intensities provides a higher skew number. Since the skew is less closer to 0 in test conditions this could mean that the foci is less intense or that the fluorescence of Ch1 is more evenly distributed. Since we don't have an increase nor a decrease of mean intensity in Ch1 it is likely that the fluorescence is more evenly distributed in the test condition Thant the control. In other words it seems that in the control conditions the foci is brighter by taking some of the low fluorescence that is present in the nuclei. In other words, it seems that the test conditions can't make very bright foci and that the Ch1 fluorescence is more evenly distributed.
  • Kurt of Ch1 Control 11.3; Test 7.24 p-value < 2.2e−16
    The Kurt account for the flatness of the distribution. If Kurt=0 we have a normal distribution, if <0 it is flatter than normal; if >0 it is more peaked. In this case the data is very very peaked in both control and test. This is quite interesting especially when looking at the violin graphs below. In the control we can see 2 peaks one close to 0 and one close to 11. This means that there are two kinds of cells in the control condition, the one with a foci and the one with only low level distributed Ch1 fluorescence. In the test conditions the Kurt are entered and the two peaks are so close that they almost merged.


  • Raw Integrated density of Ch2 Control 313299 vs Test 368339; p-value < 2.2e−16
    This is the sum of the pixel intensities. It is higher in the test cell than the control cell. This can be because the intensity is higher and/or the area is larger. In this case we have seen previously that the area is larger. Interestingly Raw intensity is higher for Ch2 but is the same for Ch1. 

  • Roundness Control 0.84; Test 0.83 p-value = 0.0032
    As said before even though the value is significant the difference here is minimal and we can easily go to focus on another 


As we have seen here it seems that Ch1 and Ch2 have defined structrures that differs between the Control and test group. Our analysis looking at each cell does not provide enough detail on each structure. More specifically it can't discriminate between the low intensity Ch1 and the high intensity Foci. Also it looks at the overall fluorescence of Ch2 in the cell while we can clearly identify a bean shape structure. The next step would be to segment the images in 3 ways: high Ch1 intensities would correspond the the foci, high Ch2 intensities would correspond to the bean shape structure and low Ch1 intensities (all Ch1 intensities but excluding the high intensities from the Foci) would relate to the overall nuclei








All about Objectives
Nicolas Stifani posted on Feb 08, 2022

What are the parameters to choose a good objective

What is the  strehl ratio?

Screw type

RMS (Royal Microscopical Society objective thread)

M25 (metric 25-millimeter objective thread)

M32 (metric 32-millimeter objective thread).

Less common or older

M27 x0.75

M27x1.0


Working distance


Immersion


Numerical aperture


Optical Correction

Transittance


Others

Tubelength 

Adjustment 

Publishing Microscopy (MicCheck)
Nicolas Stifani posted on Jan 20, 2022

I get often questions about how to write the Marterial and Method section of a manuscript including microscopy data. The diversity of technologies and applications in the field of microscopy may be complicated for some users and help of a microscopy specialist is often required.

Here I would like to present a usefull tool called MicCheck that provides a checklist of what to include in the material and method section of a manuscript including microscopy data. MicCheck is a free tool to help you. It is described in details in the following article. 

Montero Llopis, P., Senft, R.A., Ross-Elliott, T.J. et al. Best practices and tools for reporting reproducible fluorescence microscopy methods. Nat Methods 18, 1463–1476 (2021). https://doi.org/10.1038/s41592-021-01156-w


Photo-Toxicity
Nicolas Stifani posted on Nov 04, 2021


user Accès   time Tarifs   user Services    search-small Instruments   time Réservation   info Blog email Contact


What is photo-toxicity?

Under appropriate conditions of temperature (37°C), high humidity, and a controlled atmosphere (5% CO2), it is possible to grow cells in vitro. However, observing these cells under a microscope can have unintended effects. This situation is analogous to exposing yourself to sunlight: if the sunlight is too intense or the exposure lasts too long, you may suffer a sunburn. Similarly, cells can be affected by prolonged or intense light exposure during microscopy. While transmitted light (bright-field, phase-contrast, DIC) generally does not interfere with cell biology, fluorescence excitation light can cause significant phototoxic effects.

Healthy cells

What does phototoxicity look like?

Prolonged exposure to intense illumination can cause significant cellular damage, resulting in retraction, detachment, and eventual cell death. The example below demonstrates this process (click on the image to watch the movie).

Click on the image to see the movie

Cells subjected to 1 second of 59 mW, 395 nm light and 1 s 750 mW 550nm every 5 minutes. The total duration of the movie is approximately 8 hours.



Click on the image to see the movie
Cells subjected to 1 second of 70 mW, 550 nm light every 5 minutes. The total duration of the movie is approximately 3 hours, during which the cells gradually retract, detach, and eventually die 

How can I identify if phototoxicity has occurred?

The most effective way to detect phototoxicity is to capture a larger field-of-view image after your acquisition. Since phototoxicity is confined to the illuminated area, comparing adjacent non-illuminated cells with those in the exposed region offers a good—though not perfect—approach.

By stepping back, you can observe that the damage is limited to the illuminated area. However, it’s important to note that the recorded region may be smaller than the actual exposed area.

It’s also crucial to understand that this method is not a flawless control. A more ideal control would involve using a separate dish with cells maintained under identical conditions but without illumination. This is because, in the image above, we cannot definitively conclude whether the affected circular region has no impact on nearby cells. It’s possible that cell death in the illuminated region may release molecules that influence the surrounding cells. Therefore, the most reliable control would be a completely separate dish with unexposed cells.

 Taking an overview

Effects of Photo-toxicity

6 x 6 Tiles around the image area. 

What are the important factors to consider when discussing phototoxicity?

Several key factors influence phototoxicity:

  • Amount of light: Strong illumination causes more damage compared to dimmer light.
  • Wavelength: The energy carried by light depends on its wavelength. Shorter wavelengths carry higher energy and tend to be more harmful.
  • Illuminated area: Concentrating the same amount of light on a smaller area results in more localized damage.
  • Duration of illumination: Prolonged exposure (e.g., 1 second vs. 10 ms) increases the risk of damage.
  • Repetition of illumination: Frequent exposure (e.g., 10 ms pulses applied 20 times per minute) is more damaging than less frequent exposure (e.g., once per minute).

How can you detect phototoxicity?

Empirically, phototoxicity can be identified by observing cell behavior. If cells are not dividing, retracting, or detaching, it may indicate phototoxicity.

To assess phototoxicity:

  • Acquire a larger field-of-view image of the recorded area to ensure no phototoxicity has occurred and to evaluate photobleaching.
  • Use a power meter to measure the energy your cells are exposed to.
    1. Measure the power at the objective using your usual imaging settings.
    2. Record the value in mW (milliwatts = Joules/second).
    3. Divide this value by the field of view area (in cm²) to calculate the irradiance in mW/cm².

How to determine the maximum acceptable irradiance:
Finding an irradiance level that is stress-free for your cells is critical:

  1. Expose your cells continuously to a defined irradiance.
  2. Observe them over several hours. If they show no signs of phototoxicity, gradually increase the irradiance and repeat the observation.
  3. Identify the maximum continuous irradiance that does not cause damage.

Keep in mind that this value provides a baseline. Since most experiments do not involve continuous exposure, it is possible to exceed this threshold briefly. However, doing so may induce temporary stress in the cells. It is up to you to decide whether this level of stress is acceptable for your specific experiment and whether it might interfere with the biological processes you are studying.



Subtle example 

Click on the image to see the movie

Cells were imaged for 100ms with 70mW of 550nm light every 5 minutes. Cells are dividing faster than the effect of photo-toxicity that is occurring. 

How to proceed with my experiment to minimize phototoxicity?

To ensure minimal phototoxicity during your experiment, follow these steps:

  1. Estimate the light output of your instrument:
    Use a power meter to measure the light intensity (in mW) provided by your instrument. Ideally, your microscopy platform manager has recently conducted a complete quality control and can supply data on the light output at the sample. (For more information on microscopy quality control, refer to the "Power Linearity" section.)

  2. Run a pilot experiment:
    Determine the Maximum Acceptable Continuous Irradiance (mW/cm²/s) your sample can tolerate without showing signs of phototoxicity. This helps define safe imaging parameters for your experiment.

  3. Include proper controls:
    Incorporate controls in your experimental design. Ideally, image independent regions at different frequencies to confirm that imaging frequency does not affect your results. While this may not always be feasible due to software or experimental constraints, the goal is to have a control condition with the lowest light intensity required to successfully image your sample.

  4. Assess photobleaching:
    At the end of your experiment, acquire a 5x5 tile scan of the region to estimate photobleaching and ensure that your imaging conditions did not cause significant photo-damage.


Plateformes Scientifiques CIB               Biologie structurale    |    Cytométrie    |    Microscopie    |    Histologie   |    Instruments et Services

List of Software for Microscopy
Nicolas Stifani posted on Sept 12, 2021

All-in-one software (acquisition, processing, analysis)

Proprietary

  • Zeiss
  • manufactured by for  . It comes in 2 flavours: Basic Research, Advanced Research plus some options Confocal, High-Content and Artificial intelligence (Ai) add-ons.
  • Leica
  • Evident / Olympus
  • PicoQuant
  • MBF BioScience
  • Visitron

Acquisition software

Free and Open-source

  • by Nico Stuurman and Mark Tsuchida

Visualization, Processing and Analysis software

Free and Open-source

Proprietary

Utilities

search All About Microscopy: The Microscopy FacMed Blog

search All About Microscopy: The Microscopy FacMed Blog

Expansion Microscopy
Nicolas Stifani posted on Sept 12, 2021

This is not a strictly a microscopy technic. It is more a method for sample preparation in which the sample is embedded in a Gel matrix which is digested and physically expended. Conventional microscopy acquisition is then applied to the expanded sample.


See

https://pubmed.ncbi.nlm.nih.gov/25592419/

https://en.wikipedia.org/wiki/Expansion_microscopy





search The Microscopy FacMed Blog

search All About Microscopy: The Microscopy FacMed Blog