ABSTRACT
A single cell or cell population exhibits the fundamental phenomenon of cell migration during developmental processes or disease progression. Vast literature suggests that, in vitro 2-dimensional or 3-dimensional cell migration assay is one of the most commonly used assays in cancer, wound healing research, and developmental biology research. The data obtained from this assay are often analyzed using various proprietary or open-source programs. Proprietary software are costly and not always accessible to everyone. Whereas the open-source programs are free, easy to access, and user friendly. However, not all researchers are aware of these open-source programs. Despite the increasing availability of these programs, many researchers still rely on proprietary software, due to a lack of comparative analyses and practical guidance on their implementation. Hence, this review aims to provide insights into these open-source tools and serves as a practical guide to both biologists and computational researchers for their specific analytical needs.
KEYWORDS: Cancer, cell migration assay, collective cell migration, image analysis, in vitro studies, Open-source software, scratch assay, wound healing, wound healing assay
Introduction
Cell migration is a basic phenomenon involved in developmental, immunological, wound healing, and cancer metastasis processes [1–4]. Owing to the indigenous or non-indigenous signals, such as physicochemical or biological stress or inherent cellular metabolic signaling cues, cells migrate individually or collectively [2,4]. Collective cell migration is the most commonly observed phenomenon compared to single-cell migration. Due to its involvement in normal physiological processes as well as in disease states, cell migration has been studied using various experimental and simulation techniques [5,6].
In reality, cell migration occurs in 3-dimensions (3D) under normal physiological conditions [1,2]. There are few exceptions to this case; for example, cell migration on bone occurs in 2-dimensions (2D). Hence, in the laboratory setup, the collective cell migration is studied in both 2D and 3D setups. The cell migration observed in a microwell plate or cell culture dish is an example of 2D cell migration [7,8]. Whereas the cell migration studied in Matrigel™ [9] or a similar hydrogel matrix [10] and that in a spheroid/organoid [11] are perfect examples of 3D cell migration assays (Figure 1). As developing and maintaining a 3D cell migration assay setup in the laboratory is tedious, expensive, and time consuming, the usual choice among researchers is to use 2D assay [7]. In the literature, such 2D cell migration assay has other synonyms such as ‘scratch assay’ or ‘wound healing assay’. Performing a 2D cell migration assay is very easy and cheap owing to the availability of several techniques based on methods of creation of cell-free areas. Over the course of time, to improve accuracy, reproducibility, and fluency in the assay procedure, several advanced technologies such as electrical pulses or lasers have been used in this assay [7,8]. Each of these methods has its own advantages and disadvantages. Regardless of these differences, there lies a consensus among these methods, which is referred to as the creation of uniform ‘cell-free area’. In this cell-free area, the migration pattern of cells is often studied either alone or with some treatment. Depending on the need, several treatments ranging from simple one molecule (a potentially lead compound or a standard FDA-approved drug) to a complex delivery system for the same (in the form of nanoparticles/liposomes/micelles) can be employed during cell migration assay. In addition, various growth factors (as positive control) or cell migration inhibitors (as negative control) have been employed to deduce accurate inferences from this simple assay [12]. Thus, in short, the cell migration assay has great applicability in drug discovery, drug delivery technology, developmental biology, pathophysiology, and personalized medicine fields.
Figure 1.
Schematics showing the general procedure for 2D as well as 3D cell migration assay, data collection in the form of images or videos, and data analysis to obtain information of cell migration with respect to typical 5 parameters.
However, the accuracy of the results of the cell migration experiment does not lie only in the methodology or in the correct comparison between various controls and treatments. It also depends on the type of analysis chosen for live-cell imaging or time-lapse imaging data [13]. Thus, it is equally important to select the right image/video analysis platform for the cell migration study. Vast literature shows that several proprietary and open-source (also called free-to-use) analytical software/programs are available for quantifying and analyzing cell migration [14]. It is obvious that proprietary programs such as Image-Pro (https://mediacy.com/image-pro/life-science-package/), MetaMorph (https://www.biovis.com/metamorph.html), Imaris (https://imaris.oxinst.com/products/imaris-for-tracking), and Volocity (https://www.volocity4d.com/) offer a wide variety of features and data analysis tools, but not all scientists, academicians, students, and researchers have access to these programs. In addition, the prices of proprietary software are very high. While proprietary software offers robust features, its high cost, restrictive licensing agreements, and lack of easy access to underlying algorithms pose challenges for many researchers, particularly those in resource-limited settings [15,16]. Additionally, proprietary platforms often restrict data reproducibility, making it difficult to validate results across different laboratories [15,16]. Open-source alternatives provide greater flexibility, allowing researchers to customize workflows, improve transparency, and ensure long-term usability without financial barriers. Thus, the applicability of proprietary programs drastically reduces rendering the open source programs as the best alternative. This review focuses on the key features of such programs, along with their advantages and disadvantages. Previous reviews on cell migration analysis have primarily focused on proprietary platforms or specific methodologies [17–19]. However, few studies have systematically compared open-source software options, leaving researchers with limited guidance on choosing appropriate tools for different experimental conditions. This review fills that gap by providing a comparative evaluation of open-source cell migration analysis platforms, considering their accuracy, usability, and computational requirements. Cell migration assays require accurate tracking of cells over time, which has led to the development of various software tools [17]. Traditional tracking methods rely on manual annotation or algorithm-based detection, which can be prone to segmentation errors in dense cell populations and variability across different imaging conditions [17]. Recent advancements in machine learning and AI-driven image analysis have introduced deep learning-based cell tracking methods, significantly improving accuracy and automation [17–19]. However, these AI-based tools often require high computational power and large annotated datasets, limiting their accessibility to many research groups. This review examines the feasibility of integrating AI-assisted analysis with open-source platforms to enhance efficiency in cell migration studies.
Despite several advances in cell migration analysis as well as associated image analysis [20], several challenges remain. Segmentation errors in dense cell populations, high computational requirements, and variability in imaging conditions continue to limit the accuracy and reproducibility of migration assays [21–23]. Moreover, many open-source platforms require manual parameter adjustments, which may introduce user-dependent variability in results. Addressing these limitations requires integrating AI-assisted segmentation, improving standardized benchmarking datasets, and developing user-friendly interfaces. This review aims to provide a critical evaluation of current open-source platforms, highlighting their strengths, limitations, and future directions.
Methodology and criteria for selection of open-source software for review
This review compares various open-source platforms used for cell migration analysis, focusing on three primary aspects: ease of use, tracking accuracy, and computational efficiency. By evaluating these factors, we aim to provide researchers with a practical comparison to help them select the most suitable tool for their specific experimental workflows. Using PubMed, GitHub, and Scopus search engine as well as using ‘Open-source image analysis’, ‘Cell migration analysis programs’, ‘Cell tracking’ as search strings, the information was collected and elaborated in the presented review.
Open-source programs
To date, various R-based, MATLAB-based, JAVA-based, and Python-based programs (Table 1) have been devised to analyze the results of collective cell migration (Figures 1 and 2) [14]. Without these programs, most biologists would resort to only visual inspection of the cells to describe the migration qualitatively in terms of the movement of cells in cell-free areas. With such programs, the qualitative data can be converted to a meaningful quantitative form to describe the cell migration phenomenon via a series of complex image processing steps. With such analysis, the user can extract important information of migrating cells in a given cell population about their morphology, migration pattern (direction and velocity), gap width in cell-free region and % cell-free area, cell lineages, proliferating cells, and their density (Figure 1). The main advantages of open-source programs are: (i) free availability, (ii) ease of use, (iii) user-friendly graphical user interface, (iv) capability to change analysis parameters or sometimes even code of program depending on the need of use due to its free availability on the web sources (links are given in Table 1), and (v) graphical as well as image outputs for easy interpretation (Figure 3). Irrespective of these advantages, some of these programs suffer from some drawbacks, such as: (i) the need for foundation software to run the program (e.g., MATLAB), (ii) high computing power for computers in the case of analysis of large sets of images or videos, and (iii) inability to segment, identify, and detect two closely spaced yet different cells in the dense cell population.
Table 1.
List of the open-source software/tools/programs available for cell migration study along with the links for their website and user manuals.
Figure 2.
Open-source platforms available for single and collective cell migration studies in 2D as well as 3D format.
Figure 3.
Common types of output graphs and images obtained from analysis of cell migration data using open-source software.
MigraR [24]
In 2022, Shaji et al. developed this R-based analytical platform with very easy-to-use functionalities and a simple graphical user interface. To use it, one does not require in-depth mathematics or programming background. MigraR allows the user to plot cell trajectories, quantify the speed and straightness of cell migration, and helps in plotting the directions and angles of cell migration. The overall workflow of MigraR is divided into four sections: Trajectories, Velocity, Straightness, and Direction. In the first step of this workflow, the line graph for cell movement in all time frames is plotted, and the average is calculated. This line graph is known as the cell trajectory path. In the next step, based on the cell trajectory length (in micrometer or pixel terms) and time frames, the velocity of each individual cell is calculated, from which the average for the whole cell population can be determined. In the third step, the straightness that provides the behavior of individual cells for the first and last time frames can be determined. Finally, a line graph indicating the direction of movement of cells for all time frames can be extracted to observe and quantify the directionality of cells under no or specific treatment. The authors validated their program using data provided as an example by Imaris. The greatest advantage of MigraR is its capability to filter the raw dataset to improve the quality of the output results. The input files must be in the form of an Excel file with X-coordinates, Y-coordinates-, and Z-coordinates information (if 3D migration is analyzed), time, and tracking ID. Hence, to provide raw data to MigraR, it is necessary to first deduce the X and Y coordinates of migrating cells using another third-party software such as ImageJ/FIJI or Imaris.
CellTracksColab [25]
CellTracksColab is another free-to-use cell migration analysis program recently developed by Gómez-de-Mariscal et al. It is a user-friendly web browser-based tool with a simple graphical user interface that requires a Google account to operate. The main advantage of this tool is its web browser-based functionality that eliminates the installation of Python or R-environments on the end-user computers. Owing to the use of the Google Colab feature of Google LLC (California, USA), the user gets free access to extra computing resources such as random access memory (RAM) and the disc space, which in turn allows users to analyze big data sets. Due to its high computing capacity, the authors have stated that CellTracksColab can compute with more than 500,000 tracks. To validate the program, authors have explored T-cell migration patterns. Similar to MigraR, CellTracksColab also needs cell tracking information to analyze the cell migration. The authors used the StarDist and TrackMate algorithms to determine the tracking coordinates and from these datasets, they analyzed the mean speed, directionality, and distribution of cell tracks (cluster of migrating cells) under given treatments within few seconds. Furthermore, using a breast cancer cell line expressing random short hairpin ribonucleic acid (shRNA) or a shRNA targeting the filopodial protein MYO10, as well as filopodia-forming U2OS cells, the authors successfully validated their program. Overall, an easy-to-use program, fast and big dataset computation, and wide applicability (in terms of cell migration, direction, speed, collective migration, and edge migration determination) make this program very useful for biologists who are novices in the field of computation. However, the downside of this program is its inability to calculate the cell migration in 3D setup. As the program is set for measuring cell migration on 2D surface or in 2D direction, at its current level, it fails to accurately measure cell migration patterns on 3D surfaces.
Ibidi chemotaxis tool
Ibidi GmbH, a German company founded in 2001 has specialized products and services pertaining to the cell migration experiments. One of these services is the JAVA language and the ImageJ based plugin called as ‘Chemotaxis and migration tool’. This tool is solely made for the determination of single and collective cell migration under the influence of various treatments, and it has capacity to measure the cell migration parameters for 2D as well as 3D migration. This tool is capable of calculating cell migration velocity, direction, center of mass displacement, forward migration indices in parallel as well as perpendicular modes, and angle distribution of cell endpoints. Based on time-lapse videos, to calculate these parameters, it needs X-, Y-coordinates that can be calculated from manual tracking feature of ImageJ. Upon analysis, this tool gives graphical information in the form of histograms, circular plots, rose plots, density plots, and velocity plots (Figure 3). Additionally, it provides a distinct measurement test called Rayleigh test, which measures the uniformity of a circular distribution of cell endpoints. Ibidi has developed several cell culture vessels specialized for studying chemotaxis, cell migration in cancer cells, agarose gel matrix based cell chambers etc [26]. This ImageJ based tool can be used for the data obtained from time-lapse videos and images obtained from all these cell culture vessels. Hence, the biggest advantage of Ibidi chemotaxis tool is its applicability for analysis of wide variety of wound/cell-free areas having different shapes and sizes. Another advantage of this tool is that it can be applied to fluorescent as well as bright-field gray scale image/video data. Over time, Ibidi scientists have further modified this tool to analyze cell migration patterns in 3D settings.
CelltrackR [13]
In 2021, Wortel et al. developed R-package-based cell-tracking program to deduce the immune cell migration. This program offers all the pipelines necessary for migration data management, quality control of input data, extracting and visualizing the migration statistics in the form of numbers as well as graphs, determination of cell cluster tracks, and finally, simulation of cell migration. This intricate seven-step program starts with the input of cell tracking data obtained from other software such as ImageJ in the form of X-, Y-, and Z-coordinates, along with the time frame number. In the second step, CelltrackR allows to remove image bias and image artifacts and further improves the data quality with the aid of its predefined algorithms. In the next step, the programs offer visualization of cell migration data in the form of track plots (line graphs), as well as common descriptive plots, such as mean squared displacement (MSD) plots or autocorrelation plots. From these plots, the statistical information about cell migration, such as migration speed and directionality, is derived in the next step. Further, it allows the user to cluster the cells with identical migratory behavior along the tracks depending on the treatment. In the next step, based on the input data and output results, simulations can be performed to test the long-term effects of treatment on the cell migration. This feature has not been provided by other open-source programs. In the final step, based on the experimental results, analysis, and simulation, one can draw conclusions and inferences for the performed experiments. Comparative analysis with other programs revealed that CellTrackR has capability to analyze data from 2D as well as 3D cell migration studies. An additional advantage of this program is the angle analysis between the displacement vectors for the cells within number of frames. At present, this program is available as a web browser-based application where user needs to upload the data to analyze it online. This particular feature makes it even simpler to use and do the analysis.
CellMissy [27,28]
Masuzzo et al. developed this program for the correct analysis of single as well as collective cell migration, and downstream image processing. According to the authors, their program offers easy comparative analysis of all tested conditions (under which cell migration has been tested), data filtration, and quality control, along with the visualization and extraction of cell migration parameters with non-parametric statistical tests. Similar to other programs, CellMissy also requires input in the form of X- and Y-coordinates in a text file. After providing the input, the program operates in four sequential steps: (i) data inspection and visualization, (ii) cell migration parameter extraction, (iii) quality control of data, and (iv) statistical data analysis. With the help of two examples, one describing normal physiological process (immune cell migration) and another describing disease condition (cancer cell migration), they validated the utility of the CellMissy program. In the first step, the data of the cell trajectories can be visualized in the form of rose plots, which helps to deduce the differential behavior of different types of cells or the same cells under different treatments. In the next step, the analysis is performed either trajectory-wise or stepwise. In the first type, for each cell, the trajectory path is analyzed to yield the average trajectory parameters (mean or median) for all analyzed cells. On the other hand, in stepwise analysis, the displacement of the cell in each step in a single trajectory is analyzed. In the third step, the artifacts present in the cell migration data are removed using a quality control process. The criteria for quality control are the minimum distance traveled by the cells (either in µm or pixels) and the minimum number of steps per trajectory. In the final step, the statistical analysis is done using non-parametric tests (e.g., Mann-Whitney U test), and then the final interpretation of experiments is performed. Overall, CellMissy offers multi-parametric (cell speed, migration directionality) and multiscale (trajectory, steps per trajectory, number of replicates, experimental condition, and type of cells) analyses, along with the quality control of the data. Hence, CellMissy appears to be a perfect option for single and collective cell migration study.
CellProfiler Tracer [29]
This toolkit is a Python-based package that runs as a Windows application and has its own graphical user interface. Bray and Carpenter et al. developed this package in 2015. The Tracer package of the CellProfiler program is useful for deducing extensive information regarding the morphological, intensity, and textural characteristics of cells in high-throughput assays. The cell-tracking function of this program uses the linear assignment problem (LAP) approach for analyzing the time-lapse images. This feature allows the user to obtain accurate tracking by closing the temporal gaps between the time frames. This program helps in the visualization and quality assessment of large sets of tracking data derived from fluorescent as well as grayscale time-lapse image sets. CellProfiler Tracer provides output in the form of XYT plot, lineage trees, synchrograms, and heatmaps (Figure 3). The XYT plot shows the trajectory of migrating cells in the form of their X- and Y-coordinates with respect to each time frame (T). Lineage tree, on the other hand, focuses on proliferating and migrating cells in time-lapse images to dictate the ancestral relationship among cell population that can be used for cell cycle studies. The synchrogram focuses on individual cells to track them in a given image sequence, which helps in minimizing the errors associated with tracking individual cell migration in each time frame. Finally, the heat maps represent the color-coded matrix that represents the average migration feature over time for a given cell population in the experiment. The color-coded matrix also represents the significant trends in migration and trajectory data for the cell population that is being studied for certain experimental conditions (e.g., drug treatment or gene knockout conditions). Using a Drosophila blastoderm embryo, the authors validated their Tracer tool and successfully identified the cells and cell nuclei. With this validation, they deduced information about the migration tracks over time and measured the morphological features of the area, shape, intensity, and texture.
CellTracker [30]
In 2017, Piccinini et al. developed a MATLAB-based stand-alone open platform called CellTracker to analyze the cell dynamics during migration. The biggest advantage of this tool is its capacity to analyze the fluorescent images, phase contrast images, grayscale brightfield images, and differential interference contrast (DIC) images. In addition, this tool allows users to: (i) automate the image quality using different options such as vignetting or image alignment; (ii) detect and track cells in fully automatic, semi-automatic, or manual ways; (iii) compute various parameters with respect to cell motion; and finally, (iv) generate several graphs as well as 3D images of cell migration. However, the downside of this program is that the user must install a complete MATLAB software along with its image processing and figure processing toolboxes. With the first option (vignetting), the user can correct the image quality by correcting the errors imparted into image capturing owing to unavoidable imperfections in the optical systems and mechanical movable stage of the microscope. Such corrections help the cell tracking algorithm to track cells without any errors. In the next step, by feeding the values of the image dimensions and cell diameter, cells can be tracked automatically or manually. In this step, minor morphological changes in the cell structure during migration are taken into consideration by the adaptive template approach algorithm, and it is nullified to improve the accuracy of the tracking results. In the third step, by clicking on the ‘Statistics’ button and then incorporating the scale bar (µm/pixel) as well as the time frame value (the time points of the experiment at which images were captured, i.e., in terms of min/s/h per frame), various parameters such as cell displacement values, tracking coordinates, minimum, maximum, and average cell velocities along with cell movement directions can be obtained. In the final and fourth step, a 3D MATLAB figure is generated based on the analysis showing the start time image (T = 0 in the experiment) and paths of cells over the other time frames. Overall, CellTracker provides a versatile solution that is easy to use and provides results within a short period of time. A simple graphical user interface allows the user with limited or no image processing experience to work efficiently with small as well as large sets of data.
Icy [31]
In 2012, Chaumont et al. developed an intricate collaborative bioimage analysis platform called Icy, which allows the seamless development of sophisticated analysis pipelines for a variety of biological images and videos. It is a free, open-source software that can be used by anyone to develop reusable imaging protocols regardless of programming knowledge and skills. Initially, it was available in single-window mode, but recently, authors upgraded it to multiple-window mode for graphical user interface which has ‘subgrouping’ and ‘organized tab’ features that make Icy a user friendly and easy-to-use program. In this Java-based software with a multi-operating system bundle, data in more than 100 types of image formats can be processed with the help of approximately 120 plugins to provide a wide range of output options. Icy has its own in-app plugin browser, which keeps all options available just a single click away for the user. Particularly for cell migration studies, Icy works with a pipeline of few steps, such as creating detections (i.e., identifying cells and cell boundaries), linking detections in every image (i.e., tracing the tracks of the cells), collecting several tracks, and finally processing track information using processors (time-clip processors, ROI gate processor, flow track processor, etc.) to determine cell migration parameters such as displacement, migration velocity, migration direction along with graphical output (Figure 3).
iTrack4u [32]
This versatile tool was developed by Cordelières et al. This Java-based program offers single and collective cell migration analysis in manual as well as automated ways. It uses the mean shift algorithm to track cells in phase-contrast images. The images obtained from the experiment are preprocessed for enhancing the contrast and for normalizing the grayscale histogram. In a given image, this program identifies the cell by its boundary in the form of a polygon. After the identification of cells, the program tracks cell movement in a given set of time-lapse images or videos based on the original X- and Y-coordinates defined by the user. The tracking is usually done either manually or automatically. For manual tracking, the algorithm considers cell nuclei as a reference point, which appears as a dark entity inside the cell. In contrast, for automatic tracking, the program considers the cell border and nucleus position. Furthermore, upon successful tracking of the cells, the program automatically computes several variables associated with cell migration, such as cell motility, cell trajectory, cell speed, cell acceleration, pause between two time frames, and the trajectory angle of the cell in any two time frames. The time required by this tool to analyze a video or set of time-lapse images lies in the range of minutes. The main advantages of iTrack4U are reduced bias, automatic calculation of migratory variables specific to a cell line, and the ability to track cells using phase-contrast and/or fluorescence images within a very short period of time. However, irrespective of its advantages and utility, it suffers from two main drawbacks: (i) inability to identify two very close cells or dividing cells, and (ii) the small changes in cell shape being considered as cell migration leading to exaggerated and erroneous migration results.
Several FIJI/ImageJ plugins (TrackMate, wound healing tool, macro script)
TrackMate [33] is one of the multifaceted dynamic plugins of Fiji that can be combined with other inherent features and plugins of Fiji. It is an open platform developed originally for single particle tracking but has found applications in cell division dynamics, cell migration studies, particle motion studies, cell lineage studies, and cell signaling molecule dynamic studies. The independent graphical user interface of this platform guides the user through several sections for performing different steps, such as choosing the right algorithm for detection, tracking the cells and particles, and configuring them for accurate results. The output of each step is immediately displayed in a separate window parallel to the image window, which makes this platform feasible for choosing the correct parameters and seeing the results in real time. Such parallel window features make this platform reliable for obtaining accurate results. To track the migrating cells or particles or trace the dynamics of signaling molecules, the TrackMate program uses three different algorithms: (i) an algorithm derived from the LAP framework algorithm, (ii) the Kalman filter algorithm, and (iii) the simplest linking algorithm. TrackMate can be used for single image analysis as well as batch analysis of several images. The tracking data files can be extracted in several formats that are readable by other software, such as Icy or MATLAB, which proves its interoperability with other platforms. Over time, several researchers have developed various advanced modules based on the TrackMate plugin, which has led to its applicability in a wide variety of biological and healthcare fields such as protein motility, sperm motility, wound healing, microbial biofilm rheology, and cell migration during the development of Drosophila and zebrafish.
In 2020, Suarez-Arnedo et al. [34] developed a Fiji-based plugin called ‘Wound healing size tool’ for easy, fast, and reproducible analysis of scratch or wound areas in grayscale images based on the classic computer vision segmentation algorithm. This tool operates in five steps, where grayscale images are converted into their binary counterparts. After this segmentation, the wound area, also known as the scratch area or cell-free area is measured using Fiji software. For the entire process, the authors used the inherent image-processing algorithms of the Fiji. For segmentation purposes, contrast enhancement, variance filtering, binarization, and hole-filling algorithms were used, whereas the ROI manager was used for area calculation. Unlike other Fiji plugins or other open-source programs, this tool does not require image angle management or image rotation as this tool can measure cell-free areas in horizontal, vertical, and diagonal scratches. The authors have also provided an additional option to analyze the stack of time-lapse images within universal parameters. To validate this tool, authors used HaCaT cell line and performed the cell migration assay. By creating a cell-free area using specially developed 2D molds, the authors measured cell migration in terms of the % closure of the wound (or scratch). The authors also determined the accuracy of their tool by comparing it with two similar older programs called ‘ScratchAssayAnalyzer tool (2012)’ and ‘MRI_Wound_Healing_Tool (2014).’
In our recent publication [12], using a similar principle, we also developed a Fiji/ImageJ compatible JAVA language-based macro script for highly reproducible high-throughput screening of cell migration analysis. In this work, we also reported its superiority over the aforementioned ‘wound-healing size tool’ in terms of batch processing of hundreds of images without forming a stack in a reproducible manner. To develop this macro script, we used several inherent Fiji plugins of Fiji such as Find Edges, Threshold, Invert LUT, and Fill Holes. By copying this macro script into the Batch Process window, the entire directory of several images can be processed in one step.
Usiigaci [35]
In 2019, Tsai et al. developed Python-based software for the accurate tracking of cells or cell populations using stain-free phase contrast microscopy. This tool is particularly useful for analyzing densely packed cell population, where other programs may fail to identify and detect cells separately. This feature makes it one of a kind program for cell migration analysis. The word ‘Usiigaci’ refers to tracing outlines of the objects. As the name suggests, this program performs segmentation and identification of cells among the population in given images based on the mask regional convolutional neural network (Mask R-CNN) algorithm. Usiigaci has a semiautomated image analysis pipeline consisting of (i) the segmentation process, (ii) the tracking process, and (iii) data analysis. Using Mask R-CNN algorithm, the outlines of cells that are placed far apart or in proximity of each other are correctly identified. Cell tracking is performed by assigning unique identifiers to each cell and tracking these identifiers in each time frame in a given image set or video. The side-by-side panels show the original phase-contrast images and segmented images, with which the user can verify and increase the accuracy of tracking in real time. During segmentation and tracking, each identifier is featured with several output parameters, such as cell diameter, cell perimeter, eccentricity, cell orientation in a given plane, and true solidity. Usiigaci is capable of correcting the time frame associated error (the missing track of the same cell in one of the time frames) as well as track-related error (two different tracks assigned to the same cell). After this step, data analysis is performed to extract information related to cell migration velocity, direction, angle, displacement, perimeter, and area occupied by cells. These data can be represented in the form of trajectory plots, box plots, and time-series plots (Figure 3). Using NIH/3T3 fibroblasts and by testing their migration in the presence of an electric field, the authors validated the performance of Usiigaci and confirmed the superiority of its cell segmentation capability over other programs such as Phantast, Fogbank, and Deepcell.
PyScratch [36]
Recently, in 2020, Garcia-Fossa et al. developed a Python-based cell migration analysis platform called PyScratch. It is also an easy-to-use program having user-friendly graphical user interfaces. It can be used without prior background or knowledge of coding, programming, or Python language, enabling biologists to use it in a hassle-free manner. Upon start up, this program asks for the input in terms of sequentially stored images (e.g., ImageName001.tif, ImageName002.tif etc.). In the first step, this program converts all RGB images into grayscale images. In the next step, using various filters, such as the Laplacian filter to highlight the cell borders and Gaussian blur filter to remove background noise, the images are processed. In the last step, the filtered images are converted to binary images, leaving cell-free areas as white areas and cells as black-colored areas. In the final step, the program calculates the white cell-free area. Parallel panels within the same window of the GUI allow the user to visualize the image processing and plots simultaneously (Figure 3). The authors validated this program by performing a cell migration study on PNT1A and PC-3 cells. To prove its robustness and accuracy, the authors compared it with the MRI Wound Healing Tool of the Fiji software and showed that the autonomous computing feature of PyScratch yields equally reproducible results.
To conclude, the reviewed open-source platforms offer various advantages, including flexibility and cost-effectiveness. However, their accuracy in detecting cell boundaries, computational efficiency, and suitability for different migration assays vary significantly. Table 2 provides a structured comparison of these factors to help researchers select the most appropriate tool for their needs. From the comparative analysis, it is evident that MATLAB-based and Python-based programs provide the high accuracy due to intricate analysis pipeline options but come at the cost of high computational requirements. Open-source programs such as CellTrackR, CellProlfiler, MigraR, CellMissy etc. supports cell migration analysis in both 2D and 3D format with reliable analysis options. Meanwhile, Java-based programs such as ImageJ and OpenCFU are well-suited for high throughput image analysis within short duration but require manual adjustments for optimal performance [37]. These findings suggest that the choice of software should be based on the specific needs of the study, balancing accuracy, computational efficiency, and experimental complexity.
Table 2.
Comparative performance analysis of the open-source cell migration tools.
| Software Tool | Supported Migration assays | Program Base | Cell boundaries detection accuracy | Processing time per image (Computational load) | Cell segmentation Algorithm | Output parameters | Advantages | Disadvantages |
|---|---|---|---|---|---|---|---|---|
| MigraR | 2D/3D | R-based | High | Medium | No | Cell velocity, direction, Straightness, Trajectories | Simple Graphical User Interface, No programming background needed, Filter the raw dataset | Requires third-party software for initial data processing |
| CellTracksColab | 2D | Python- based | High | High | Yes | Cell velocity, Direction, Speed, Collective migration, Edge migration | Web-based, Free access to extra computing resources, Fast processing | Limited to 2D migration analysis |
| Ibidi Chemotaxis Tool | 2D/3D | Java-based | High | Medium | Yes | Cell velocity, Direction, Centre of mass displacement, Forward migration Indices, angle distribution, Rayleigh test | Wide applicability, Supports various image types, Specialized for chemotaxis, Applied to fluorescent as well as bright-field gray scale image data | Requires manual tracking for initial data processing |
| CellTrackR | 2D/3D | R-based | High | Medium | No | Cell velocity, Direction, Clustering, Simulation, Angle analysis | Comprehensive analysis pipeline, Supports simulations | Requires third-party software for initial data processing |
| CellMissy | 2D | Java-based | High | Medium | No | Cell velocity, Direction, Trajectory analysis, Quality control, Statistical analysis | Multi-parametric analysis, | Limited to 2D migration analysis |
| CellProfiler Tracer | 2D/3D | Python-based | High | Medium | Yes | Cell velocity, Direction, Lineage trees, Synchrograms, Heatmaps | Accurate tracking, Visualization of large datasets | Requires installation of Python environment |
| CellTracker | 2D/3D | MATLAB-based | High | Medium | Yes | Cell velocity, Direction, Displacement, 3D images | Versatile, supports various image types, Easy to use Graphical User Interface | Requires installation of MATLAB and toolboxes |
| Icy | 2D/3D | Java-based | High | Medium | Yes | Cell velocity, Direction, Displacement, Graphical output | Free, open source, Supports various image formats | Requires installation of Java environment |
| iTrack4U | 2D | Java-based | High | Medium | Yes | Cell motility, Trajectory, Speed, Acceleration, Pause, Trajectory angle | Reduced bias, Automatic calculation, Fast processing | Inability to identify close or dividing cells, Small shape changes considered as migration |
| TrackMate | 2D | Java-based | High | Low | Yes | Cell velocity, Direction, Particle motion, Cell lineage | Multifaceted, Real-time results, Batch analysis | Limited to 2D migration analysis |
| Wound Healing Tool | 2D | Java-based | High | Low | Yes | Percent closure of wound, Scratch area measurement | Easy, Fast, Reproducible, No image angle management needed | Limited to scratch or wound area measurement |
| Macro Script | 2D | Java | High | Low | Yes | Cell migration analysis, Batch processing | Highly reproducible, Batch processing | Limited to 2D migration analysis |
| Usiigaci | 2D | Python | High | Medium | Yes | Cell velocity, Direction, Angle, displacement, Area | Accurate segmentation of densely packed cells, Real-time verification | Complex setup |
| PyScratch | 2D | Python-based | High | Medium | Yes | Cell-free area measurement, Image processing | User-friendly Graphical User Interface, no coding knowledge needed, Reproducible results | Limited to scratch or wound area measurement |
Limitations, scope of development, and future perspective
In general, the open-source platforms for cell migration analysis offer unique advantages; however, there remain several challenges irrespective of their usefulness. First and foremost limitation of many open-source programs is identification and correct segmentation of cell boundaries. The migration phenomenon is often studied for collective cell population and seldomly for single cell. Thus, in dense and highly confluent cell population, cell segmentation becomes tedious task [38]. In such scenario, these open-source programs may lead to either over-segmentation or loss of cell boundaries yielding erroneous results. Although programs such as Usiigaci performs cell segmentation based on neural network algorithm, not all programs have such facility. Furthermore, there are no universal standards or a benchmarking dataset to compare the results obtained from these open-source programs. Additionally, as these programs offer manual tuning of several parameters, such freedom imparts user-dependent variability making it difficult to find accuracy as well as reproducibility among the users and experiments. In order to bring the reproducibility, less errors with respect to cell segmentation and tracking, these open-source programs can be integrated with AI and deep learning algorithms in future. However, such integration and these deep learning AI models require large number of datasets for learning which further limits the accessibility of these open-source programs for many research groups. Additionally, at current level of analysis, these programs either process stored images or videos, but not the real-time cell migration data. In future, these programs can be modulated to handle high throughput real time cell migration data where actual in vitro cell migration experiment can go hand-in-hand with its analysis. In current times, introducing the cloud-based computing solution may help in significant reduction in the cost of computation by allowing the large-scale image analysis without requiring high-performance local machines at the user-end. Additional validation pipeline can be introduced in future to validate the obtained cell migration results against the standard dataset.
Conclusion
In conclusion, the comparative review of all open-source analytical platforms for cell migration shows that almost all programs yield similar and specific valuable information about single or collective cell migration. The three most important parameters, average cell migration speed, directionality, and displacement, were determined by all these programs via a variety of algorithms. Some of these programs have the capacity to deduce minute details about cellular morphology and cell edge changes during migration, while others neglect such minor changes and focus only on the displacement of the cell. Hence, the decision regarding the usage of a particular program depends on the type of information that the user wants to extract from the cell migration study. Most biologists across the globe are unaware of coding and the process of building an analytical pipeline using MATLAB, R, Python, and JAVA. Therefore, crucial information coming from very simple and cheap assay, such as cell migration assay, is often overlooked. The development of such open-source analytical platforms has offered great opportunities to dive deep into the details of cell migration phenomena via analysis of fluorescently labeled as well as label-free cell images. A semi-automatic or fully automatic pipeline allows users to work with large sets of data in a short period. These features make these programs suitable for high-throughput screening. Additionally, the capabilities of some of these programs to compute 2D and 3D cell migration have made them strong toolkits, specifically for cancer biologists. Currently, the introduction of artificial intelligence and machine learning approaches has revolutionized the programming sector. With such approaches and using simple parameters such as cell diameter, size, morphological features, and deformation of the cell structure before, during, and after migration, researchers have improved the accuracy and precision of cell migration assays [39]. For experimental biologists with no knowledge of coding and programming, such programs have proven to be a boon for diving into an intricate world of cell biology with a migration perspective. Each open-source software tool has distinct advantages and the choice of platform should align with the specific needs of the study. While some tools offer ease of use, others provide advanced AI-based tracking capabilities or high-throughput processing. Therefore, based on the comparative analysis, the researchers are recommended to follow Table 2 for selecting the most suitable tool depending on the specific requirements of their experiment.
Supplementary Material
Acknowledgments
A.S.J. expresses gratitude to the Manipal Centre for Biotherapeutics Research (MCBR) and the Manipal Academy of Higher Education (MAHE) for their support.
Funding Statement
This work was supported by the Manipal Academy of Higher Education under Intramural Grant [MAHE/CDS/PHD/IMF/2024].
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data availability statement
Data sharing is not applicable to this study because no new data were generated. The original research articles referred to in this review are available in public domains such as Scopus and PubMed.
Supplementary material
Supplemental data for this article can be accessed online at https://doi.org/10.1080/19336918.2025.2488116
References
- [1].Yamada KM, Sixt M.. Mechanisms of 3D cell migration. Nat Rev Mol Cell Biol. 2019;20(12):738–13. [DOI] [PubMed] [Google Scholar]
- [2].SenGupta S, Parent CA, Bear JE. The principles of directed cell migration. Nat Rev Mol Cell Biol. 2021;22(8):529–547. doi: 10.1038/s41580-021-00366-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Merino-Casallo F, Gomez-Benito MJ, Hervas-Raluy S, et al. Unravelling cell migration: defining movement from the cell surface. Cell Adhes & Migr. 2022;16(1):25–64. doi: 10.1080/19336918.2022.2055520 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Pawluchin A, Galic M. Moving through a changing world: single cell migration in 2D vs. 3D. Front Cell Dev Biol. 2022;10:10. doi: 10.3389/fcell.2022.1080995 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Grada A, Otero-Vinas M, Prieto-Castrillo F, et al. Research techniques made simple: analysis of collective cell migration using the wound healing assay. J Invest Dermatol. 2017;137(2):e11–e16. doi: 10.1016/j.jid.2016.11.020 [DOI] [PubMed] [Google Scholar]
- [6].Matsiaka OM, Baker RE, Shah ET, et al. Mechanistic and experimental models of cell migration reveal the importance of cell-to-cell pushing in cell invasion. Biomed Phys Eng Express. 2019;5(4):045009. doi: 10.1088/2057-1976/ab1b01 [DOI] [Google Scholar]
- [7].Riahi R, Yang Y, Zhang DD, et al. Advances in wound-healing assays for probing collective cell migration. SLAS Technol. 2012;17(1):59–65. doi: 10.1177/2211068211426550 [DOI] [PubMed] [Google Scholar]
- [8].Stamm A, Reimers K, Strauß S, et al. In vitro wound healing assays – state of the art. BioNanomaterials. 2016;17(1–2):79–87. doi: 10.1515/bnm-2016-0002 [DOI] [Google Scholar]
- [9].Audoin M, Søgaard MT, Jauffred L. Tumor spheroids accelerate persistently invading cancer cells. Sci Rep. 2022;12(1):14713. doi: 10.1038/s41598-022-18950-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Trappmann B, Baker BM, Polacheck WJ, et al. Matrix degradability controls multicellularity of 3D cell migration. Nat Commun. 2017;8(1):371. doi: 10.1038/s41467-017-00418-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Zhang Y, Li Y, Thompson KN, et al. Polarized NHE1 and SWELL1 regulate migration direction, efficiency and metastasis. Nat Commun. 2022;13(1):6128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Joshi AS, Madhusudanan M, Mijakovic I. 3D printed inserts for reproducible high throughput screening of cell migration. Front Cell Dev Biol. 2023;11:11. doi: 10.3389/fcell.2023.1256250 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Wortel IMN, Liu AY, Dannenberg K, et al. CelltrackR: an R package for fast and flexible analysis of immune cell migration data. ImmunoInformatics. 2021;1–2:100003. doi: 10.1016/j.immuno.2021.100003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Gonzalez-Beltran AN, Masuzzo P, Ampe C, et al. Community standards for open cell migration data. Gigascience. 2020;9(5). doi: 10.1093/gigascience/giaa041 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Toscano E, Cimmino E, Pennacchio FA, et al. Methods and computational tools to study eukaryotic cell migration in vitro. Front Cell Dev Biol [Internet]. 2024;12:12. doi: 10.3389/fcell.2024.1385991 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Bulut O, Beiting-Parrish M, Casabianca JM, et al. The rise of artificial intelligence in educational measurement: opportunities and ethical challenges. Chin/Engl J Educ Meas & Evaluation | 教育测量与评估双语期刊 [Internet]. 2024. [cited 2025 Mar 22];5(3):3. [Google Scholar]
- [17].Chen B, Yin Z, Ng BW-L, et al. Label-free live cell recognition and tracking for biological discoveries and translational applications. npj Imag [Internet]. 2024;2(1):41. doi: 10.1038/s44303-024-00046-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Mezei T, Kolcsár M, Joó A, et al. Image analysis in histopathology and cytopathology: from early days to current perspectives. J Imag [Internet]. 2024;10(10):252. doi: 10.3390/jimaging10100252 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Juez-Castillo G, Valencia-Vidal B, Orrego LM, et al. FiCRoN, a deep learning-based algorithm for the automatic determination of intracellular parasite burden from fluorescence microscopy images. Med Image Anal [Internet]. 2024;91:103036. [DOI] [PubMed] [Google Scholar]
- [20].Goldstien L, Lavi Y, Atia L. ConfluentFUCCI for fully-automated analysis of cell-cycle progression in a highly dense collective of migrating cells. PLOS ONE [Internet]. 2024;19(6):1–15. doi: 10.1371/journal.pone.0305491 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Wiggins L, O’Toole PJ, Brackenbury WJ, et al. Exploring the impact of variability in cell segmentation and tracking approaches. Microscopy Res & Technique [Internet]. 2025;88(3):716–731. doi: 10.1002/jemt.24715 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Bragantini J, Theodoro I, Zhao X, et al. Ultrack: pushing the limits of cell tracking across biological scales. bioRxiv [Internet]. 2024. doi: 10.1101/2024.09.02.610652 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Chen H, Murphy RF, Lidke D. Evaluation of cell segmentation methods without reference segmentations. MboC [Internet]. 2023;34(6):ar50. doi: 10.1091/mbc.E22-08-0364 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Shaji N, Nunes F, Ines Rocha M, et al. MigraR: an open-source, R-based application for analysis and quantification of cell migration parameters. Comput Methods Programs Biomed [Internet]. 2022;213:106529. doi: 10.1016/j.cmpb.2021.106529 [DOI] [PubMed] [Google Scholar]
- [25].Gómez-de-Mariscal E, Grobe H, Pylvänäinen JW, et al. CellTracksColab — a platform for compiling, analyzing, and exploring tracking data. bioRxiv [Internet]. 2024;2023.10.20.563252. [Google Scholar]
- [26].Zengel P, Nguyen-Hoang A, Schildhammer C, et al. μ-slide chemotaxis: a new chamber for long-term chemotaxis studies. BMC Cell Biol. 2011;12(1):21. doi: 10.1186/1471-2121-12-21 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Masuzzo P, Huyck L, Simiczyjew A, et al. An end-to-end software solution for the analysis of high-throughput single-cell migration data. Sci Rep. 2017;7(1):42383. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Masuzzo P, Hulstaert N, Huyck L, et al. CellMissy: a tool for management, storage and analysis of cell migration data produced in wound healing-like assays. Bioinformatics. 2013;29(20):2661–2663. doi: 10.1093/bioinformatics/btt437 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Bray M-A, Carpenter AE. CellProfiler Tracer: exploring and validating high-throughput, time-lapse microscopy image data. BMC Bioinformatics. 2015;16(1):369. doi: 10.1186/s12859-015-0759-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Piccinini F, Kiss A, Horvath P. CellTracker (not only) for dummies. Bioinformatics. 2016;32(6):955–957. doi: 10.1093/bioinformatics/btv686 [DOI] [PubMed] [Google Scholar]
- [31].de Chaumont F, Dallongeville S, Chenouard N, et al. Icy: an open bioimage informatics platform for extended reproducible research. Nat Methods. 2012;9(7):690–696. doi: 10.1038/nmeth.2075 [DOI] [PubMed] [Google Scholar]
- [32].Cordelières FP, Petit V, Kumasaka M, et al. Automated cell tracking and analysis in phase-contrast videos (iTrack4u): development of Java software based on combined mean-shift processes. PLOS ONE. 2013;8(11):e81266. doi: 10.1371/journal.pone.0081266 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [33].Ershov D, Phan M-S, Pylvänäinen JW, et al. TrackMate 7: integrating state-of-the-art segmentation algorithms into tracking pipelines. Nat Methods. 2022;19(7):829–832. doi: 10.1038/s41592-022-01507-1 [DOI] [PubMed] [Google Scholar]
- [34].Suarez-Arnedo A, Torres Figueroa F, Clavijo C, et al. An image J plugin for the high throughput image analysis of in vitro scratch wound healing assays. PLOS ONE. 2020;15(7):e0232565. doi: 10.1371/journal.pone.0232565 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].Tsai H-F, Gajda J, Sloan TFW, et al. Usiigaci: instance-aware cell tracking in stain-free phase contrast microscopy enabled by machine learning. SoftwareX. 2019;9:230–237. doi: 10.1016/j.softx.2019.02.007 [DOI] [Google Scholar]
- [36].Garcia-Fossa F, Gaal V, de Jesus MB. PyScratch: an ease of use tool for analysis of scratch assays. Comput Methods Programs Biomed. 2020;193:105476. [DOI] [PubMed] [Google Scholar]
- [37].Geissmann Q, Merks RM. OpenCFU, a New free and open-source software to count cell colonies and other circular objects. PLOS ONE [Internet]. 2013;8(2):1–10. doi: 10.1371/journal.pone.0054072 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [38].Loewke NO, Pai S, Cordeiro C, et al. Automated cell segmentation for quantitative phase microscopy. IEEE Trans Med Imaging. 2018;37(4):929–940. doi: 10.1109/TMI.2017.2775604 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [39].Zhang F, Zhang R, Wei M, et al. A machine learning based approach for quantitative evaluation of cell migration in Transwell assays based on deformation characteristics. Analyst. 2023;148(6):1371–1382. doi: 10.1039/D2AN01882A [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data sharing is not applicable to this study because no new data were generated. The original research articles referred to in this review are available in public domains such as Scopus and PubMed.