Fully functional bundle: To use the fully functional ImageJ bundle, just download the platform-dependent version of ImageJ bundled with Java file and containing all the tools needed for the package embedded:
JEasyTFM command at the right of the ImageJ menu for using it.
![]() Manual installation: Proceed with the manual installation by downloading the ImageJ software and unziping it.
Alternatively the download of all the previously described files can simply be replaced by downloading the JEasyTFM_files file and unzip it in the plugins/jars folder. Finally one of the platform-dependent library files need to be downloaded into the plugins/jars folder:
|
The JEasyTFM tool is divided into two submenus: ![]()
Splitting the JEasyTFM tool into two steps, with on one side a job creation and on the other side, a job launching operation, provides two main advantages. The first one is to offer traceability of all the analysis settings used to generate the different analysis results. In the case of heavily booked dedicated analysis computers, the second one is to rapidly define and protect the analysis settings, which can then quickly be launched later on once the machine is available. When launching the JEasyTFM->Create_job tab, the following window will be displayed: ![]() We will now explain each point in details.
|
![]() Debug: When the Debug checkbox is selected, the analysis is performed by launching the different used plugins going through their GUI, often resulting in images being displayed at each stage for the user to visualize the results of the different processes going on. On the contrary, when the Debug checkbox is unselected, all the analyses are performed launching the additional plugins by using some back-doors (i.e. static methods) that have been created to short-cut the plugins GUI resulting in all the images produced in memory and not displayed for a gain of machine time. Data_folder: On clicking on the Data_folder Browse button, a "choose directory" dialog is displayed for the user to choose the directory in which the data to be analyzed are saved.
Upon selection validation of the chosen folder, its path will be updated within the field Data_folder together with the prefixes file name of the acquired images to be analyzed in the File field and the one for the images of the beads at equilibrium (i.e. after the addition of trypsin also called reference images in this document) in the Reference_images field.
Sequence_images: On clicking on the Sequence_images Browse button, a "file open" dialog is displayed for the user to choose an image file to be analyzed.
Upon selection validation of an image file, the file name prefix of the acquired images to be analyzed will be updated within the Sequence_images field.
Reference_images: On clicking on the Reference_images Browse button, a "file open" dialog is displayed for the user to choose a reference images file to be analyzed.
Upon selection validation of the image file, the file name prefix of the acquired images to be analyzed will be updated within the Reference_images field.
The reference images selected in Reference_images have to be previously saved within a folder whose name is the one defined within Data_folder field with the addition of "_-_Reference".
|
![]()
Number_of_acquired_cells: Slider defining the number of cells that have been recorded during the acquisitions.
Note that the value specified by the Number_of_acquired_cells slider represents the maximum value for the sliders:
Slices: Slider defining the number of z-slices that have been acquired during the acquisitions. Frames: Slider defining the number of time-frames that have been acquired during the acquisitions. Frames_begin: Slider whose maximum value will be defined by the Frames slider and defining the first time-frames number of the data that will be analyzed.
|
![]() Beads1: The selection of the Beads1 and/or Beads2 checkboxes implies that an experiment of high-resolution TFM has been performed, i.e. with the use of beads with two different colors.
Beads2: The selection of the Beads2 and/or Beads1 checkboxes implies that an experiment of high-resolution TFM has been performed, i.e. with the use of beads with two different colors.
Ch1: The selection of the Ch1 and/or Ch2 and/or Ch3 checkboxes implies that more than one channel has been acquired for imaging the cells.
Ch2: The selection of the Ch1 and/or Ch2 and/or Ch3 checkboxes implies that more than one channel has been acquired for imaging the cells.
Ch3: The selection of the Ch1 and/or Ch2 and/or Ch3 checkboxes implies that more than one channel has been acquired for imaging the cells.
Number of Ch: Popup menu defining the number of channels that have been acquired.
Kurto1: The Kurto1 checkbox will apply a best Kurtosis selection algorithm to the images of the Ch1 in the case it is selected and the "Extended Depth Field"1 plugin in the case it is unselected.
Kurto2: The Kurto2 checkbox will apply a best Kurtosis selection algorithm to the images of the Ch2 in the case it is selected and the "Extended Depth Field"1 plugin in the case it is unselected.
Kurto3: The Kurto3 checkbox will apply a best Kurtosis selection algorithm to the images of the Ch3 in the case it is selected and the "Extended Depth Field"1 plugin in the case it is unselected.
Young's_modulus_Pa: Numerical field defining the Young's modulus in Pascal units of the support used for cell plating. Precision_for_the_Regularization_factor_calculation: Popup menu defining the precision for the calculation of the regularization or Lagrange parameter λ value that is used to convert the measured beads displacements to the forces applied from the cells onto their support.
(1) Forster, B., Van De Ville, D., Berent, J., Sage, D. & Unser, M. Complex Wavelets for Extended Depth-of-Field: A New Method for the Fusion of Multichannel Microscopy Images. Microsc. Res. Tech. 65, 33-42 (2004).
|
![]() Cells_images: The checkbox defines whether cell images should be extracted from all the Slices for all the defined Frames and for the cells numbers defined between the Cells_starting_folder and Cells_ending_folder.
Cells_starting_folder: Defines the first cell number or folder for which cell images should be extracted from all the Slices and for all the defined Frames.
Cells_ending_folder: Defines the last cell number or folder for which cell images should be extracted from all the Slices and for all the defined Frames.
Image_chosen_from_slices: The automatic selection of the best focused image may be performed by putting the value 0 in this slider. In this case, a Kurtosis algorithm will be applied on the Slices images in order to extract the one giving the highest value and this for all the defined Frames. On the other hand, manual inspection of the z-slices can be performed and the slice number of the selected best focused image can be defined in the Image_chosen_from_slices slider by choosing which cell image number between 1 and the number indicated in Slices should be extracted from all the Slices for all the defined Frames. |
![]() Beads_Reference_images: The checkbox defines whether images corresponding to the best focused image of the beads extracted within the z-series acquired with the support at equilibrium (i.e. after the cells have been detached from the surface) should be extracted from all the Slices.
Beads_reference_starting_folder: Defines the first cell number or folder for which images corresponding to the best focus of the beads within the images acquired with the support at equilibrium (i.e. after the cells have been detached from the surface) should be extracted from all the Slices.
Beads_reference_ending_folder: Defines the last cell number or folder for which images corresponding to the best focus of the beads within the images acquired with the support at equilibrium (i.e. after the cells have been detached from the surface) should be extracted from all the Slices.
Beads_Reference_images_find_focused_slices: The checkbox defines whether the "Find focused slices"2 plugin should be applied prior to applying the "Extended Depth Field"1 plugins on the images of the beads with the support at equilibrium from all the Slices in order to extract the best focused image.
(2) Tseng, Q. Find Focused Slices. ImageJ plugin available at: https://sites.google.com/site/qingzongtseng/find-focus
(1) Forster, B., Van De Ville, D., Berent, J., Sage, D. & Unser, M. Complex Wavelets for Extended Depth-of-Field: A New Method for the Fusion of Multichannel Microscopy Images. Microsc. Res. Tech. 65, 33–42 (2004). |
![]() Beads_Sequence_images: The checkbox defines whether images corresponding to the best focus of the beads extracted within the images through the acquired time-lapse should be extracted from all the Slices for all the defined Frames.
Beads_sequence_starting_folder: Defines the first cell number or folder for which images corresponding to the best focus of the beads extracted within the images through the acquired time-lapse should be extracted from all the Slices for all the defined Frames.
Beads_sequence_ending_folder: Defines the last cell number or folder for which images corresponding to the best focus of the beads extracted within the images through the acquired time-lapse should be extracted from all the Slices for all the defined Frames.
Beads_Sequence_images_find_focused_slices: The checkbox defines whether the "Find focused slices"2 plugin should be applied prior to applying the "Extended Depth Field"1 plugin on the images of the beads through the acquired time-lapse, defining the Frames from all the Slices for all the defined Frames to extract the best focused image.
(2) Tseng, Q. Find Focused Slices. ImageJ plugin available at: https://sites.google.com/site/qingzongtseng/find-focus
(1) Forster, B., Van De Ville, D., Berent, J., Sage, D. & Unser, M. Complex Wavelets for Extended Depth-of-Field: A New Method for the Fusion of Multichannel Microscopy Images. Microsc. Res. Tech. 65, 33–42 (2004). |
![]()
Traction_force: The checkbox defines whether the whole or part of the force calculation processes should be launched.
This is the main checkbox for the analysis, which means that if it is not selected, none of the following analysis processes:
If the checkboxes Beads1 and/or Beads2 is/are activated, the launched analysis will then be applied to all the selected beads colors. Traction_starting_folder: Defines the first cell number or folder for which the whole or part of the force calculation processes should be calculated.
Traction_ending_folder: Defines the last cell number or folder for which the whole or part of the force calculation processes should be calculated.
Traction_force_alignment: The checkbox defines whether the whole or part of the images alignment processes should be calculated.
This means that if the checkbox Traction_force_alignment is not selected none of the following images alignment processes: will be launched whether or not they are selected. If the checkboxes Beads1 and/or Beads2 is/are activated, the launched analysis will then be applied to all the selected beads colors. Alignment_make: The checkbox defines whether an alignment algorithm should be launched on the images.
Alignment_crop: The checkbox defines whether the algorithm that will get rid of the image borders generated by the Alignment_make algorithm should be launched on the images.
Traction_force_calculation: The checkbox defines whether the whole or part of the images forces calculation processes should be launched.
This means that if the checkbox Traction_force_calculation is not selected none of the following force calculation processes: will be launched whether or not they are selected. If the checkboxes Beads1 and/or Beads2 is/are activated, the launched analysis will then be applied to all the selected beads colors. PIV_calculation: The checkbox defines whether a PIV (Particle Image Velocimetry) algorithm should be launched on the images of the aligned and cropped beads in order to generate beads displacement maps.
Lambda_calculation: The checkbox defines whether the regularization or Lagrange parameter λ values should be calculated.
Force_calculation: The checkbox defines whether a Fourier Transform Traction Cytometry (FTTC) algorithm should be launched on the beads displacement maps previously generated with the PIV_calculation feature in order to create force maps.
Force_superposition: The checkbox defines whether the force maps previously generated with the Force_calculation feature should be superimposed with the images of the cells that have been previously generated with the Cells_images feature.
|
![]()
Cells_analysis: The checkbox defines whether the whole or part of the cells segmentation and force integration processes should be launched. This is the main checkbox for the analysis, which means that if it is not selected none of the following analysis processes:
will be launched whether or not they are selected.
If the checkboxes Beads1 and/or Beads2 is/are activated, the launched analysis will then be applied to all the selected beads colors. Cells_analysis_starting_folder: Defines the first cell number or folder for which the whole or part of the cells segmentation and force integration processes should be launched.
Cells_analysis_ending_folder: Defines the last cell number or folder for which the whole or part of the cells segmentation and force integration processes should be launched.
Cells_selection (manual user step): The checkbox defines whether a cells segmentation algorithm with some manual user steps should be launched on the images of the cells that have been previously aligned and cropped with the Alignment_make and Alignment_crop algorithms and pre-filtered with the Cells_segmentation algorithm.
Cells_segmentation: The checkbox defines whether a cells segmentation algorithm should be launched on the images of the cells that have been previously aligned and cropped with the Alignment_make and Alignment_crop algorithms.
Force_integration: The checkbox defines whether a whole cells force integration algorithm should be launched on the traction force images using the cells segmentation ROI (Region Of Interest) data that have been obtained either automatically through a Cells_selection (manual user step) algorithm or manually and the result together with the ROI drawing superimposed with the images of the cells that have been previously generated with the Cells_images feature.
|
![]()
Traction_high_force: The checkbox defines whether the whole or part of the beads positions tracking processes should be launched.
This is the main checkbox for the analysis, which means that if it is not selected none of the following analysis processes: will be launched whether or not they are selected. If the checkboxes Beads1 and/or Beads2 is/are activated, the launched analysis will then be applied to all the selected beads colors. Traction_high_starting_folder: Defines the first cell number or folder for which the whole or part of the beads positions tracking processes should be calculated.
Traction_high_ending_folder: Defines the last cell number or folder for which the whole or part of the beads positions tracking processes should be calculated.
Traction_high_enhance_contrast: The checkbox defines whether a Subtract_Background... followed by an Enhance_Contrast... filter should be applied on the images in order to enhance their contrast before applying the High_PIV_calculation regression.
High_PIV_calculation: The checkbox defines whether a PIV (Particle Image Velocimetry) algorithm should be launched on the images of the aligned and cropped beads in order to generate beads displacement maps.
Particle_tracker: The checkbox defines whether a particle tracking algorithm should be launched on the beads displacement maps previously generated with the High_PIV_calculation feature in order to track the beads displacements between images that have been previously generated with the Beads_Sequence_images.
|
![]() FA_segmentation (manual user step): The checkbox defines whether a FA_segmentation (manual user step) algorithm should be launched on fluorescent images of the cells that have been previously aligned and cropped with the Alignment_make and Alignment_crop algorithms.
FA_segmentation_starting_folder: Defines the first cell number or folder for which the FA_segmentation (manual user step) algorithm should be applied.
FA_segmentation_ending_folder: Defines the last cell number or folder for which the FA_segmentation (manual user step) algorithm should be applied.
|
![]() Traction_force_applied_on_FA: The checkbox defines whether a Traction_force_applied_on_FA7-8 algorithm should be launched which will calculate the forces applied on the focal adhesion positions that had previously been determined by the FA_segmentation algorithm using the beads displacements data obtained by the Particle_tracker algorithm.
Traction_force_applied_on_FA_starting_folder: Defines the first cell number or folder for which the Traction_force_applied_on_FA algorithm should be applied.
Traction_force_applied_on_FA_ending_folder: Defines the last cell number or folder for which the Traction_force_applied_on_FA algorithm should be applied.
(7) Schwarz, U. S., Balaban, N. Q., Riveline, D. Bershadsky, A., Geiger, B., Safran, S. A.
Calculation of forces at focal adhesions from elastic substrate data: the effect of localized force and the need for regularization, Biophys. J. 83, 1380-1394 (2002) (8) Sabass, B., Gardel, M. L., Waterman, C. and Schwarz, U. S. high-resolution traction force microscopy based on experimental and computational advances, Biophys. J., 94,207-220, (2008) |
![]() Generate_file_and_Exit: The button will validate the selections within the JEasyTFM>Create_job window, save the data of the analysis configuration job into a batch file (JEasy.txt) within the ImageJ plugins folder and close the JEasyTFM>Create_job window. Cancel: The button will cancel the selections within the JEasyTFM>Create_job window, without saving the data of the analysis configuration job into a batch file (JEasy.txt) and close the JEasyTFM>Create_job window. |
![]() |
For a TFM time-lapse experiment, z-stack and time-series images need to be gathered for each acquired channel (within them need at least to be a bead and cell image) and xy position (i.e. cell).
Once the time-series images are acquired, z-stack images of the beads at equilibrium (named reference images later in this document) need to be taken as well for each xy position. The input files must be in OME TIF format with the z-stack (slices) images followed by the t-series (frames) images within a single file. Each xy stage position (corresponding to a cell) and channel have to be saved within a single file. Additionally, the filenames need to finish with the cell number as a string followed by an underscore (i.e. "_") and an iterated channel number padded to two digits (i.e. the images for the cell 1 with 3 acquired channels will finish with "1_01.tif", "1_02.tif" and "1_03.tif"). For example for the cell number 12 with 3 acquired channels, the filenames should finish with:
The conditions on the filenames and formats are similar to the one described for the kinetic acquisition images. |
If we have for example image filenames for the cell number 12 being:
In the example, this will be: ProtocolCombination_14.11.2018_15_51_21_Cell_ given that the ends of the filenames
|
Each time a job is launched, a copy of the job file "JEasyTFM.txt" file (saved in the "ImageJ/plugins" folder) used for the given launch is saved in a folder named "jobs" with the name "JEasyTFM_y_begin.txt"; y corresponding to the number of launches made.
This means that on the first job launch within the given experiment folder, the copied job name will be "JEasyTFM_1_begin.txt", on the second launch it will be "JEasyTFM_2_begin.txt" and so on, giving thus traceability of the different launches made.
|
The Cells_images analysis creates a folder for each cell number defined within Cells_starting_folder and Cells_ending_folder within which all following analysis data will be saved.
In this part of the analysis, the best focused images for all the frames of the acquired channels for all the acquired cells will be extracted. The outputted images will be named
The images will be saved in a folder named Cell z; z corresponding to the given cell number. |
In this part of the analysis the best focused images of the beads at equilibrium (i.e. after the addition of trypsin also called reference images) for all the acquired cells will be extracted.
|
In this part of the analysis the best focused images for all the frames of the beads for all the acquired cells will be extracted.
|
In this section of the package, the main features of the traction force analysis will be performed divided into different sections.
Traction force alignment output files: The Traction_force_alignment part of the analysis is composed of an Alignment_make section which aligns the best focused images for all the frames of the beads taking as reference the best focused image of the beads at equilibrium and this for all the acquired cells.
The translation of images with respect to the reference image done by the Alignment_make section will introduce borders that have to be eliminated. To do so, the Alignment_crop section opens the images of the beads, measure the position of the borders and crop them out. Finally the images of the beads created by the Alignment_make algorithm will be overwritten. Similarly, all the images of the cells created by the previous algorithm are opened, cropped and overwritten.
Alignment make output files: The Alignment_make section aligns the best focused images for all the frames of the beads taking as reference the best focused image of the beads at equilibrium and this for all the acquired cells.
The output images will be named
Beads1 the images will be named
Beads2 the images will be named
In the case of the activation of the checkbox Beads1 the reference images will be named
In the case of the activation of the checkbox Beads2 the reference images will be named
The translation transformations that have been applied on the
The outputted images will be named
Beads1 the images will be named
Beads2 the images will be named
Alignment crop output files: The translation of images with respect to the reference image done by the Alignment_make section will introduce borders that have to be eliminated.
To do so, the Alignment_crop section opens the images of the beads, measures the position of the borders, crops them out and adds the border positions data to the previously generated Transformation.txt file. Finally the images of the beads created by the Alignment_make algorithm will be overwritten. Similarly, all the images of the cells created by the previous algorithm are opened, cropped and overwritten.
Traction force calculation output files: The Traction_force_calculation part of the analysis is composed of a PIV_calculation section which applies a PIV (Particle Image Velocimetry) algorithm between the reference images of the beads and the images for all the frames of the beads generated at the Traction_force_alignment section.
Next, the Force_calculation section will generate the corresponding force maps using a FTTC (Fourier Transform Traction Cytometry) algorithm. And the Force_superposition section will combine all the cells with all the force maps images.
PIV calculation output files: The PIV_calculation section which will apply a PIV (Particle Image Velocimetry) algorithm between the reference images of the beads and the images for all the frames of the beads generated at the Traction_force_alignment section.
The results are outputted in the form of text in a file named
Beads1 the output files will be named
Beads2 the output files will be named
Lambda calculation output files: The Lambda_calculation section will read all the previously generated
The results are outputted in the form of text in a file named
Beads1 the file will be saved in a folder named
Beads2 the file will be saved in a folder named
Force calculation output files: The Force_calculation section first reads all the previously generated
These values are then stored in a text file named
The obtained force outputs are then saved in a text file named
Beads1 the output files will be named
Beads2 the output files will be named
Force superposition output files: The Force_superposition section will combine all the cells with all the force maps images.
Thus the outputted images will be named:
Beads1 the images will be named:
Beads2 the images will be named:
|
In the Cells_analysis part of the analysis, the segmentation of the images of the cells as well as the force integration over the cells over all the frames will be performed.
The Cells_segmentation part of the analysis applies several filters to the acquired images of the cells and the Cells_selection part of the analysis requires the user to draw a ROI (Region Of Interest) delimiting the displacements of the cell of interest over all the frames as well as to define a threshold window for the given cell over all the frames in order to generate cells segmentation ROIs. The Force_integration part of the analysis makes a superposition of the different cell images together with the vector and magnitude force representations images and the cell ROI together with the measured integrated values.
Cells selection output files: The Cells_selection part of the analysis requires the user to draw a ROI (Region Of Interest) delimiting the displacements of the cell of interest over all the frames as well as to define a threshold window for the given cell over all the frames.
The output will be a ROI file for all the frames named
In the case of the activation of the checkbox Beads1 the output ROI file
Beads2 it will be saved in the folder
Cells segmentation output files: The Cells_segmentation part of the analysis applies several filters to the acquired images of the cells and output images named
Force integration output files:
The Force_integration part of the analysis makes a copy of the previous PIV output text file
The generation of all these files in a new folder may be redundant, but it is a very fast process and above all, it is able to compartmentalize the creation of
In the case of the activation of the checkbox Beads1 the files will be named
Beads2 the files will be named
Next the superposition of the different cell images together with the vector and magnitude force representations images and the cell ROI as well as measured integrated values are generated and named
Beads1 the images will be named:
Beads2 the images will be named:
All the integration data are resumed in a table saved in a text file named
|
In the Traction_high_force part of the analysis, the tracking of the beads between the reference images of the beads and the images for all the frames of the beads generated at the Traction_force_alignment section will be performed.
The Traction_high_force part of the analysis is composed of a High_PIV_calculation section which takes the data previously obtained from the PIV_calculation section of the Traction_force_calculation part of the analysis to push the calculation further to have a more refined definition of the beads displacement map. The Particle_tracker section will use the more refined definition of the beads displacement map to compute the displacement of the beads.
High PIV calculation output files: The High_PIV_calculation section which will take the data previously obtained from the PIV_calculation section of the Traction_force_calculation part of the analysis, saved in the
The results are outputted in the form of text in a file named
Beads1 the output files will be named
Beads2 the output files will be named
Particle tracker output files: The Particle_tracker section uses the previously generated
The obtained outputs are saved in a text file named
Beads1 the output files will be named
Beads2 the output files will be named
|
Within the FA_segmentation section, a segmentation of focal adhesion algorithm outputs a ROI file named
Beads1 the files will be named
Beads2 the files will be named
|
The Traction_force_applied_on_FA7-10 algorithm outputs
If the checkboxes Ch1 and/or Ch2 and/or Ch3 is/are activated, this algorithm will then be applied to the beads displacements and FA positions data merge of all the selected beads colors.text files named
(7) Balaban, N. Q., Schwarz, U. S., Riveline, D., Goichberg, P., Tzur, G., Sabanay, I., Mahalu, D., Safran, S., Bershadsky, A., Addadi, L. and Geiger, B.
Force and focal adhesion assembly: a close relationship studied using elastic micro-patterned substrates. Nat. Cell Biol. 3, 466-472 (2001) (8) Schwarz, U. S., Balaban, N. Q., Riveline, D. Bershadsky, A., Geiger, B., Safran, S. A. Calculation of forces at focal adhesions from elastic substrate data: the effect of localized force and the need for regularization, Biophys. J. 83, 1380-1394 (2002) (9) Schwarz, U. S., Balaban, N. Q., Riveline, D., Addadi, L., Bershadsky, A., Safran, S. A., Geiger, B. Measurement of cellular forces at focal adhesions using elastic micro-patterned substrates, Mat. Sci. Eng. C 23, 387-394 (2003) (10) Sabass, B., Gardel, M. L., Waterman, C. and Schwarz, U. S. high-resolution traction force microscopy based on experimental and computational advances, Biophys. J., 94,207-220, (2008) |
The gel deformation is measured by the technique of PIV (Particle Image Velocytometry) which measure the change of beads spatial position between constrained and unconstrained conditions and thus tracks the fluorescent beads in different interrogation windows. Therefore, the accuracy of this method is dependent upon parameters such as (1) the number of beads per image, (2) the signal to noise ratio of the images of the beads and (3) the precision of the bead tracking algorithm. |
1 - Minimal number of beads per images [top]The number of beads within the images is as well an important parameter to consider in order to properly reconstruct the traction force maps exerted by the cells on a surface. To determine the lower bead density required for traction force reconstruction within JEasyTFM, we tested images with different beads density. We started with a low bead density of 3000 beads for a 1124 × 1164 pixels image, and then artificially decreased the number of the beads of the reference and sequence images of the beads. Note that the acquired original images had a size of 1200 × 1200 pixels which have been slightly reduced through the Traction force alignment algorithm (i.e. Alignment make and Alignment crop). To do so, the intensity of circular ROIs of 10 pixels centered at the positions of the beads was set to the intensity value of the background. The choice of the beads to be erased in the reference and sequence images had then been chosen using a random number generator. The figure below shows force maps images obtained for different beads density. As expected, decreasing the beads density to 2500 beads / 1124 × 1164 pixels decreases the resolution of the traction force reconstruction map (compare figure 3000 beads with 2000 beads) without altering the overall reconstruction map. However, reducing the bead density to 1500 beads / 1124 × 1164 pixels clearly starts to alter the traction force reconstruction map. Please note that by dragging the mouse over the force maps images, the image is replaced by a movie illustrating the switch between the two images of the beads (i.e. before and after addition of Reference) that have been used to generate the given force map. These movies clearly show the diminution of the beads density. |
![]() |
![]() |
3000 beads | 2500 beads |
![]() |
![]() |
2000 beads | 1500 beads |
![]() |
![]() |
1000 beads | 500 beads |
2 - Minimal signal to noise ratio of the images of the beads [top]Given that the positions of the beads are the fiducial markers used to reconstruct the traction forces exerted by the cells on their support, it is mandatory to acquire images of the beads by respecting a minimum signal to noise ratio value. To evaluate this number, we first measured the mean and standard deviation of the signal-to-noise ratio of the images of the beads by calculating the mean and standard deviation of the intensity ratio integrated over all the images of an experiment (i.e. all stack-z and time-series images). Depending on the system used to acquire the images of the beads, we found a signal-to-noise ratio value up to 7 for the best system we tested and in the chosen example (see below) we measured a signal to noise ratio of 3.4153 ± 0.9159. To test the output results of the force reconstruction algorithm with images having lower signal to noise ratio values, we artificially degraded their quality by using an Enhance Contrast filter with different Saturated pixels values. The figure below shows force maps images obtained for images of the beads having different values of the signal-to-noise ratio. We notice that there is no degradation within the force maps resolution until a signal-to-noise ratio value lower than 2. Please note that by dragging the mouse over the force maps images, the image is replaced by a movie illustrating the switch between the two images of the beads (i.e. before and after addition of trypsin) that have been used to generate the given force map. These movies clearly show the alteration of the beads signal upon applying the Enhance Contrast filter. |
![]() |
![]() |
S/N = 3.4153 ± 0.9159 | S/N = 2.3005 ± 0.5999 |
![]() |
![]() |
S/N = 2.0005 ± 0.4576 | S/N = 1.9003 ± 0.2625 |
![]() |
![]() |
S/N = 1.7001 ± 0.2173 | S/N = 1.1006 ± 0.1549 |
3a - Bead tracking - Build upon the Particle_Tracker 1.5. version released September 2006 [top]The ParticleTracker_2D code developed by the MOSAIC group and published in:
was used as a starting point for the bead tracking routine.
We did not change the first part of the code, which is used for "Particle detection". The "Particle linking" part of the code using a Particle Tracking Velocimetry (PTV) method is used to track the displacement of the beads between strained and unstrained positions. This part of the code is based on scoring between the differences of the positions of the beads, grey values and derivations of the grey values upon the definition of a “Displacement” value (see Figure below). |
![]() |
![]() |
Graphical User Interface of the original ParticleTracker_2D plugin with on left the interface before launching the algorithm and on right the one obtained after the tracking completed |
However, the setting of a "Displacement" value does not always allow an accurate tracking of the beads over the whole image. In fact, beads couples (i.e. the associations of the same bead in the reference and given sequence image) displaying large displacements will be missed if the displacement value has been set to a value that is smaller than the largest bead displacement found in the image. Moreover, if the displacement value has been set to a value equal or higher to the largest bead displacement, the algorithm then generates often wrong beads couples associations. This is well documented in the Figures below, which show results obtained with the original ParticleTracker_2D plug-in using different "Displacement" values within the plugin GUI. As observed when the displacement value is set to 5 pixels, some of the large bead displacements are not taken into account (green circle), whereas when the displacement value is set to 15 pixels, wrong beads couple association can be observed (blue circle). |
![]() |
![]() |
ParticleTracker_2D output with Displacement set to 5 pixels | ParticleTracker_2D output with Displacement set to 10 pixels |
![]() |
![]() |
ParticleTracker_2D output with Displacement set to 15 pixels | ParticleTracker_2D output with Displacement set to 20 pixels |
To improve the quality of the beads tracking routine, we modified the original ParticleTracker_2D algorithm by implementing the results of the displacements maps obtained by the PIV algorithm as starting positions for the tracking algorithm. This allows to use a minimal displacement value and thus eliminate all wrong beads couples associations. The use of a small displacement value requires using a displacement map with a small grid size of 8 pixels. The Figures below show results obtained with our Particle_Tracker_PIV_and_Trajectories_Inputs plugin implemented within JEasyTFM using as input a PIV calculation map with a grid size of 8 pixels and "Displacement" value set to 5 within the plugin GUI. |
![]() |
![]() |
Graphical User Interface of our Particle_Tracker_PIV_and_Trajectories_Inputs plugin with on left the interface before launching the algorithm and on right the one obtained after the tracking completed. |
![]() |
![]() |
PIV calculation with a grid size of 8 pixels | Particle_Tracker_PIV_and_Trajectories_Inputs output with Displacement set to 5 pixels |
3b - Bead tracking - - Build upon the MosaicSuite 1.0.25 version released on 20 September 2022 [top]Starting February 2010 the MOSAIC group started to incorporate the ParticleTracker_2D tool code within a MosaicSuite:
|
![]() |
![]() |
Graphical User Interface of the original Particle_Tracker_2D/3D tool of the MosaicSuite package with on left the interface before launching the algorithm and on right the one obtained after the tracking completed |
Similarly, the setting of a "Displacement" value does not always allow an accurate tracking of the beads over the whole image, either missing large displacements or generating wrong beads couples associations. This is shown within the Figures below, which show results obtained with the original Particle_Tracker_2D/3D tool plug-in using different "Displacement" values within the plugin GUI. As observed when the displacement value is set to 5 pixels, some of the large bead displacements are not taken into account (green circle), whereas when the displacement value is set to 15 pixels, wrong beads couple association can be observed (blue circle). |
![]() |
![]() |
Particle_Tracker_2D/3D tool output with Displacement set to 5 pixels | Particle_Tracker_2D/3D tool output with Displacement set to 10 pixels |
![]() |
![]() |
Particle_Tracker_2D/3D tool output with Displacement set to 15 pixels | Particle_Tracker_2D/3D tool output with Displacement set to 20 pixels |
To improve the quality of the beads tracking routine, we modified the original Particle_Tracker_2D/3D tool algorithm by implementing the results of the displacements maps obtained by the PIV algorithm as starting positions for the tracking algorithm. This allows to use a minimal displacement value and thus eliminate all wrong beads couples associations. The use of a small displacement value requires using a displacement map with a small grid size of 8 pixels. The Figures below show results obtained with our MosaicSuite-1.0.25_Full_extended package update implemented within JEasyTFM using as input a PIV calculation map with a grid size of 8 pixels and "Displacement" value set to 5 within the plugin GUI. |
![]() |
![]() |
Graphical User Interface of our Particle_Tracker_2D/3D tool within our MosaicSuite-1.0.25_Full_extended package update with on left the interface before launching the algorithm and on right the one obtained after the tracking completed |
![]() |
![]() |
PIV calculation with a grid size of 8 pixels | Particle_Tracker_2D/3D tool using the PIV outputs with Displacement set to 5 pixels |
The reconstruction routine used in JeasyTFM provides (1) the Extended Depth of Field plugin to extract the best focused images of the beads for both the reference and all the sequence images and (2) the possibility to run the Find focused slices plugin prior launching the Extended Depth of Field plugin to reduce the number of z-slices images. In addition, the traction force reconstruction routine used in JeasyTFM is mainly based on the one published in Martiel, J.-L. et al. (2015) Measurement of cell traction forces with ImageJ. Methods Cell Biol., 125, 269–287, which used the Fourier Transform Traction Cytometry (FTTC) method to measure the traction force exerted by the cell inducing the gel deformation and bead displacement. The original code was simplified and improved. Maps, exporting LUTs and style possibilities such as "Draw magnitude and vector", Draw X magnitude" and "Draw Y magnitude" were added. The force maps can also now be generated in Newtons and/or in Pascals (versus only in Pa previously). This reconstruction depends mainly on two parameters i.e the grid size value and the regularization factor. |
![]() |
![]() |
Graphical User Interface of the original plugin with on the left and on right our updated version. |
1 - Grid size values [top]The grid size value is pre-defined in the JeasyTFM plugin. This value can be changed by modifying the source code of the plugin. For the force interpolation, a grid size of 16 pixels was selected which means that the distance between 2 vectors is equal to 16 pixels. The "grid size" has to be a value multiple of 2. Increasing or decreasing this value will either increase or decrease the maps vector spacing and thus decrease or increase the spatial resolution of the traction force map. Below are figures showing PIV calculation maps obtained with different grid size values and the corresponding mean and standard deviation calculation times obtained over 20 runs. We pre-defined a grid size value of 16 pixels in JEasy TFM because it is the best compromise between spatial resolution and calculation time. Nevertheless as input for the bead tracking algorithm, used mainly for high-resolution TFM, we use a PIV map generated with a grid size of 8 pixels to be thus able to reduce the allowed beads displacements to 5 pixels. |
![]() |
![]() |
Grid size of 64 pixels - Calculation time : 3.97 ± 0.19 s | Grid size of 32 pixels - Calculation time : 20.93 ± 0.63 s |
![]() |
![]() |
Grid size of 16 pixels - Calculation time : 30.03 ± 1.16 s | Grid size of 8 pixels - Calculation time : 181.95 ± 3.52 s |
2 - Regularization factor [top]Within the present version of the JEasyTFM GUI, the value of the regularization factor is set through a numeric field with a default choice set to "1.0e-10". To analyzed how the smoothing coefficient affects the value of the forces, we generated 2200 values (i.e. 100 measured values per decade which gives 22 × 100 = 2200 values) of the regularization factor logarithmically distributed between 10-21 and 10 and calculated the corresponding forces (see Figures below). The optimal choice for the regularization factor corresponds to the point of inflection of regularization versus force curve. For the example given, this value is ≈ 3.45.10-10. |
Force and force ratio vs regularization factor plots generated for 2200 values of the regularization factor logarithmically distributed between 10-21 and 10. |