Advancements in microscopy techniques permit us to acquire endless datasets of images. A major bottleneck in cell imaging is how to analyze petabytes of data in an effective, reliable, objective, and effortless way. Quantitative imaging is becoming crucial to disentangle the complexity of many biological and pathological processes. For instance, cell shape is a summary readout of a myriad of cellular processes. Changes in cell shape use to reflect changes in growth, migration mode (including speed and persistence), differentiation stage, apoptosis, or gene expression, serving to predict health or disease. However, in certain contexts, e.g., tissues or tumors, cells are tightly packed together, and measurement of individual cellular shapes can be challenging and laborious. Bioinformatics solutions like automated computational image methods provide a blind and efficient analysis of large image datasets. Here we describe a detailed and friendly step-by-step protocol to extract various cellular shape parameters quickly and accurately from colorectal cancer cells forming either monolayers or spheroids. We envision those similar settings could be extended to other cell lines, colorectal and beyond, either label/unlabeled or in 2D/3D environments.
Keywords: Cell size; Colon cancer cells; HCT116; Machine learning; Monolayers; Quantification; Segmentation; Spheroids.
© 2023. The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature.