Deep Learning Model for Cell Nuclei Segmentation and Lymphocyte Identification in Whole Slide Histology Images

Anti-cancer immunotherapy dramatically changes the clinical management of many types tumours towards less harmful and more personalized treatment plans than conventional chemotherapy or radiation. Precise analysis spatial distribution immune cells in tumourous tissue is necessary to select patients that would best respond treatment. Here, we introduce a deep learning-based workflow for cell nuclei segmentation subsequent identification routine diagnostic images. We applied our on set hematoxylin eosin (H&E) stained breast cancer colorectal images detect tumour-infiltrating lymphocytes. Firstly, segment all tissue, multiple-image input layer architecture (Micro-Net, Dice coefficient (DC) $0.79\pm 0.02$). supplemented Micro-Net with an introduced texture block increase accuracy (DC = $0.80\pm preserved shallow network only 280 K trainable parameters (e.g. U-net ∼1900 parameters, DC $0.78\pm 0.03$). Subsequently, added active contour ground truth further performance $0.81\pm Secondly, discriminate lymphocytes from segmented nuclei, explored multilayer perceptron achieved 0.70 classification f-score. Remarkably, binary was significantly improved (f-score 0.80) by colour normalization. To inspect model generalization, have evaluated trained models public dataset not put use during training. conclude proposed promising results and, little effort, can be employed multi-class tasks.