Joint Region and Nucleus Segmentation for Characterization of Tumor Infiltrating Lymphocytes in Breast Cancer

摘要

Histologic assessment of stromal tumor infiltrating lymphocytes (sTIL) as a surrogate of the host immune response hasbeen shown to be prognostic and potentially chemo-predictive in triple-negative and HER2-positive breast cancers. Thecurrent practice of manual assessment is prone to intra- and inter-observer variability. Furthermore, the inter-play of sTILs,tumor cells, other microenvironment mediators, their spatial relationships, quantity, and other image-based features haveyet to be determined exhaustively and systemically. Towards analysis of these aspects, we developed a deep learning basedmethod for joint region-level and nucleus-level segmentation and classification of breast cancer H&E tissue whole slideimages. Our proposed method simultaneously identifies tumor, fibroblast, and lymphocyte nuclei, along with keyhistologic region compartments including tumor and stroma. We also show how the resultant segmentation masks can becombined with seeding approaches to yield accurate nucleus classifications. Furthermore, we outline a simple workflowfor calibrating computational scores to human scores for consistency. The pipeline identifies key compartments with highaccuracy (Dice= overall: 0.78, tumor: 0.83, and fibroblasts: 0.77). ROC AUC for nucleus classification is high at 0.89(micro-average), 0.89 (lymphocytes), 0.90 (tumor), and 0.78 (fibroblasts). Spearman correlation between computationalsTIL and pathologist consensus is high (R=0.73, p<0.001) and is higher than inter-pathologist correlation (R=0.66,p<0.001). Both manual and computational sTIL scores successfully stratify patients by clinical progression outcomes.