Focalizing regions of biomarker relevance facilitates biomarker prediction on histopathological images

• Published in: iScience Vol. 26; no. 10; p. 107243
• Main Authors: Gan, Jiefeng, Wang, Hanchen, Yu, Hui, He, Zitong, Zhang, Wenjuan, Ma, Ke, Zhu, Lianghui, Bai, Yutong, Zhou, Zongwei, Yullie, Alan, Bai, Xiang, Wang, Mingwei, Yang, Dehua, Chen, Yanyan, Chen, Guoan, Lasenby, Joan, Cheng, Chao, Wu, Jia, Zhang, Jianjun, Wang, Xinggang, Chen, Yaobing, Wang, Guoping, Xia, Tian
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 20.10.2023; Elsevier
• Subjects: Cancer; Histology; Machine learning; Pathology; Pathology; Histology; Machine learning; Cancer
• ISSN: 2589-0042; 2589-0042
• DOI: 10.1016/j.isci.2023.107243

Image-based AI has thrived as a potentially revolutionary tool for predicting molecular biomarker statuses, which aids in categorizing patients for appropriate medical treatments. However, many methods using hematoxylin and eosin-stained (H&E) whole-slide images (WSIs) have been found to be inefficient because of the presence of numerous uninformative or irrelevant image patches. In this study, we introduced the region of biomarker relevance (ROB) concept to identify the morphological areas most closely associated with biomarkers for accurate status prediction. We actualized this concept within a framework called saliency ROB search (SRS) to enable efficient and effective predictions. By evaluating various lung adenocarcinoma (LUAD) biomarkers, we showcased the superior performance of SRS compared to current state-of-the-art AI approaches. These findings suggest that AI tools, built on the ROB concept, can achieve enhanced molecular biomarker prediction accuracy from pathological images. [Display omitted] •SRS was proposed to predict molecular biomarker status based on ROB concept•SRS can adaptively discover ROB regions targeting specific biomarkers•The generalization performance of SRS was verified on four biomarkers Histology; Pathology; Cancer; Machine learning