Dictionary-based Pathology Mining with Hard-instance-assisted Classifier Debiasing for Genetic Biomarker Prediction from WSIs.

Abstract

Prediction of genetic biomarkers, e.g., microsatellite instability in colorectal cancer is crucial for clinical decision making. However, two primary challenges hamper accurate prediction: (1) It is difficult to construct a pathology-aware representation involving the complex interconnections among pathological components. (2) Whole slide images contain a large proportion of areas unrelated to genetic biomarkers, which make the model easily overfit simple but irrelevant instances. We hereby propose a dictionary-based hierarchical pathology mining with hard-instance-assisted classifier debiasing framework to address these challenges, dubbed as D²Bio. Our first module, dictionary-based hierarchical pathology mining, is able to mine diverse and very fine-grained pathological contextual interaction without the limit to the distances between patches. The second module, hard-instance-assisted classifier debiasing, learns a debiased classifier via focusing on hard but task-related features, without any additional annotations. Experimental results on five cohorts show the superiority of our method, with over 4% improvement in AUROC compared with the second best on the TCGA-CRC-MSI cohort. Our analysis further shows the clinical interpretability of D²Bio in genetic biomarker diagnosis and potential clinical utility in survival analysis. Code, patient-level dataset splits, patch coordinates, and preprocessing scripts will be available at https://github.com/DeepMed-Lab-ECNU/D2Bio.