Original title: Nondestructive, quantitative viability analysis of 3D tissue cultures using machine learning image segmentation
Authors: Kylie J. Trettner, Jeremy Hsieh, Weikun Xiao, Jerry S.H. Lee, Andrea M. Armani
In this article, the authors discuss the limitations of current methods for assessing the viability of cells in different cell culture conditions. Traditionally, scientists have relied on colorimetric indicators, which provide simple binary readouts. However, recent research has combined viability assessment techniques with image-based deep-learning models to automate the characterization of cellular properties.
The authors propose a new image processing algorithm that can quantify cellular viability in 3D cultures without the need for assay-based indicators. They demonstrate that their algorithm performs just as well as human experts in analyzing whole-well images over multiple days and culture matrix compositions.
To showcase the potential usefulness of their algorithm, the authors conduct a longitudinal study investigating the effects of a therapeutic on pancreatic cancer spheroids. Using high content imaging, the algorithm successfully tracks viability at both the individual spheroid and whole-well levels.
Not only does this method significantly reduce analysis time compared to human experts, but it also has the advantage of being independent of the microscope or imaging system used. This approach can accelerate progress and improve the robustness and reproducibility of 3D culture analysis in biological and clinical research.
Original article: https://arxiv.org/abs/2311.09354