Cell morphology-based machine learning models for human cell state classification

摘要

a Box and whisker plot of the training dataset. The means of the six features (FSC-A, SSC-A, FSC-H, SSC-H, FSC-W, and SSC-W) were comparable, with the maximal ratio less than 2.0-fold (meanSSC-A/meanFSC-H = 1.81). b Visualization of six-dimensional training dataset (green: live cells, yellow: apoptotic cells) using PCA (principal component analysis). c Visualization of the six-dimensional training dataset (green: live cells, yellow: apoptotic cells) using t-SNE (t-distributed stochastic neighbor embedding).