Segmentation-Free Cell Phenotype Classification using Deep Residual Neural Networks

Cell phenotype classification can be used to characterize complex cell states associated with chemical compound treatment, therefore has great potential in drug discovery. Previous work on image-based cell phenotype classification required a routine yet cumbersome step of single cell segmentation before the classification task. In this paper, we present a segmentation-free method for image-based cell phenotype classification using deep residual neural networks (ResNets). The cell images are samples treated with annotated compounds that can be mainly grouped into three clusters, giving three classes to be classified. Instead of single-cell phenotype classification, we use the raw images without segmentation for our training and evaluation directly. Compared to previous reference work, we significantly simplify the data preprocessing steps while still achieving high accuracy. Our trained ResNets achieve a 98.2% accuracy rate based on five-fold cross-validation.