Tracking of cell populations to understand their spatio-temporal behavior in response to physical stimuli

Abstract

We have developed methods for segmentation and tracking of cells in time-lapse phase-contrast microscopy images. Our multi-object Bayesian algorithm detects and tracks large numbers of cells in presence of clutter and identifies cell division. To solve the data association problem, the assignment of current measurements to cell tracks, we tested various cost functions with both an optimal and a fast, suboptimal assignment algorithm. We also propose metrics to quantify cell migration properties, such as motility and directional persistence, and compared our findings of cell migration with the standard random walk model. We measured how cell populations respond to the physical stimuli presented in the environment, for example, the stiffness property of the substrate. Our analysis of hundreds of spatio-temporal cell trajectories revealed significant differences in the behavioral response of fibroblast cells to changes in hydrogel conditions.

Understanding the Optics to Aid Microscopy Image Segmentation

“…For example, thresholding on local intensity value and variation has a long history on cell image segmentation ( [13] and references therein). Morphological operations on the gradient or intensity images are widely applied to segment specimen pixels [5,8]. Using the artifact of microscopy images, Laplacian of Gaussian filter is also used to extract object blob and contour [12].…”

“…Using the artifact of microscopy images, Laplacian of Gaussian filter is also used to extract object blob and contour [12]. After obtaining binary masks indicating whether each pixel is a specimen or background pixel, connected component labelling or marker-controlled watershed algorithms are often performed to group specimen pixels into specimen objects [5,14].…”

“…Over time, biologists have learned how to overcome or even exploit those artifacts for interpreting images. When computer-based microscopy image analysis began to relieve humans from tedious manual labelling [5,8,12,14], it is unsurprising that those artifacts cause the major difficulty in automated image processing. In particular, they hinder the process of segmenting images into cells and background, which is the most critical step in almost all cell analysis and tracking systems.…”

Detection of mitosis within a stem cell population of high cell confluence in phase-contrast microscopy images

“…Computer vision for automated analysis of cell populations has gained increasing attention due to its enormous potential for discoveries in cell biology and pharmacology as well as stem cell engineering [5,11,6]. Among such visionbased systems, the ones adopting phase-contrast time-lapse microscopy allow for long-term continuous monitoring of live and intact cells because phase-contrast microscopy is a non-destructive imaging modality.…”

Cell segmentation in microscopy imagery using a bag of local Bayesian classifiers

“…Automated tracking of cell populations in vitro in microscopy enables applications such as optimizing cell culture conditions in stem cell manufacturing to meet research and clinical demands [5]. When developing a computer vision-based tracking system capable of tracking cells in a large population, cell segmentation plays an important role for shape analysis, cell detection and cell association in spatiotemporal context [4,5,12].…”