Focalizing regions of biomarker relevance facilitates biomarker prediction on histopathological images

Jiefeng Gan; Hanchen Wang; Hui Yu; Zitong He; Wenjuan Zhang; Ke Ma; Lianghui Zhu; Yutong Bai; Zongwei Zhou; Alan Yullie; Xiang Bai; Mingwei Wang; Dehua Yang; Yanyan Chen; Guoan Chen; Joan Lasenby; Chao Cheng; Jia Wu; Jianjun Zhang; Xinggang Wang; Yaobing Chen; Guoping Wang; Tian Xia

doi:10.1016/j.isci.2023.107243

Focalizing regions of biomarker relevance facilitates biomarker prediction on histopathological images

iScience. 2023 Jun 29;26(10):107243. doi: 10.1016/j.isci.2023.107243. eCollection 2023 Oct 20.

Authors

Jiefeng Gan^{1

2}, Hanchen Wang^{3

4}, Hui Yu⁵, Zitong He⁶, Wenjuan Zhang⁷, Ke Ma¹, Lianghui Zhu⁸, Yutong Bai⁶, Zongwei Zhou⁶, Alan Yullie⁶, Xiang Bai^{1

9}, Mingwei Wang¹⁰, Dehua Yang¹⁰, Yanyan Chen¹¹, Guoan Chen¹², Joan Lasenby³, Chao Cheng¹³, Jia Wu¹⁴, Jianjun Zhang^{15

16}, Xinggang Wang⁸, Yaobing Chen¹, Guoping Wang¹, Tian Xia^{1

2}

Affiliations

¹ Institute of Pathology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China.
² School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan, China.
³ Department of Engineering, University of Cambridge, Fitzwilliam House 32 Trumpington Street, Cambridge CB2 1QY, UK.
⁴ Computing + Mathematical Sciences Department, California Institute of Technology, 1200 East California Boulevard, Pasadena, CA 91125, USA.
⁵ Wuhan Children's Hospital, Tongji Medical College, Wuhan, Hubei 430000, China.
⁶ Department of Computer Science, Johns Hopkins University, 3400 N Charles St, Baltimore, MD 21218, USA.
⁷ Department of Pathology, Maternal and Child Hospital of Hubei Province, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei 43000, China.
⁸ School of Electronic Information and Communications, Huazhong University of Science and Technology, Wuhan, Hubei 430000, China.
⁹ Department of Radiology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei 43000, China.
¹⁰ The National Center for Drug Screening, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai 201203, China.
¹¹ Department of Information Management, Tongji Hospital, Huazhong University of Science and Technology, Wuhan, Hubei 430000, China.
¹² Wuhan Blood Center, Wuhan, Hubei 43000, China.
¹³ Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA.
¹⁴ Department of Radiation Oncology, Stanford University School of Medicine, 875 Blake Wilbur Dr, Palo Alto, CA 94304, USA.
¹⁵ Thoracic/Head and Neck Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA.
¹⁶ Translational Molecular Pathology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA.

Abstract

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.

Keywords: Cancer; Histology; Machine learning; Pathology.

Abstract

Grants and funding