Deep learning supported discovery of biomarkers for clinical prognosis of liver cancer

Liang，Junhao1; Zhang，Weisheng1; Yang，Jianghui2; Wu，Meilong3,6,7; Dai，Qionghai4; Yin，Hongfang2; Xiao，Ying2; Kong，Lingjie1,5

2023

10.1038/s42256-023-00635-3

Nature Machine Intelligence   Impact Factor & Quartile

2522-5839

Tissue biomarkers are crucial for cancer diagnosis, prognosis assessment and treatment planning. However, there are few known biomarkers that are robust enough to show true analytical and clinical value. Deep learning (DL)-based computational pathology can be used as a strategy to predict survival, but the limited interpretability and generalizability prevent acceptance in clinical practice. Here we present an interpretable human-centric DL-guided framework called PathFinder (Pathological-biomarker-finder) that can help pathologists to discover new tissue biomarkers from well-performing DL models. By combining sparse multi-class tissue spatial distribution information of whole slide images with attribution methods, PathFinder can achieve localization, characterization and verification of potential biomarkers, while guaranteeing state-of-the-art prognostic performance. Using PathFinder, we discovered that spatial distribution of necrosis in liver cancer, a long-neglected factor, has a strong relationship with patient prognosis. We therefore proposed two clinically independent indicators, including necrosis area fraction and tumour necrosis distribution, for practical prognosis, and verified their potential in clinical prognosis according to criteria derived from the Reporting Recommendations for Tumor Marker Prognostic Studies. Our work demonstrates a successful example of introducing DL into clinical practice in a knowledge discovery way, and the approach may be adopted in identifying biomarkers in various cancer types and modalities.

SCI

English

Others

STI2030-Major Projects[2022ZD0212000] ; National Natural Science Foundation of China (NSFC)["61831014","32021002"] ; Tsinghua-Foshan Innovation Special Fund (TFISF)[2021THFS0207] ; Guoqiang Institute, Tsinghua University[2021GQG1024] ; Beijing Tsinghua Changgung Hospital Fund[12021C1009]

Computer Science

Computer Science, Artificial Intelligence ; Computer Science, Interdisciplinary Applications

WOS:000962715900001

NATURE PORTFOLIO

2-s2.0-85151423540

Scopus

1.State Key Laboratory of Precision Measurement Technology and Instruments,Department of Precision Instrument,Tsinghua University,Beijing,China
2.Department of Pathology,Beijing Tsinghua Changgung Hospital,School of Clinical Medicine,Tsinghua University,Beijing,China
3.School of Clinical Medicine,Tsinghua University,Beijing,China
4.Department of Automation,Tsinghua University,Beijing,China
5.IDG/McGovern Institute for Brain Research,Tsinghua University,Beijing,China
6.Division of Hepatobiliary and Pancreas Surgery,Department of General Surgery,Shenzhen People’s Hospital,The Second Clinical Medical College,Jinan University,Shenzhen,China
7.Division of Hepatobiliary and Pancreas Surgery,Department of General Surgery,The First Affiliated Hospital,Southern University of Science and Technology,Shenzhen,China

Liang，Junhao,Zhang，Weisheng,Yang，Jianghui,et al. Deep learning supported discovery of biomarkers for clinical prognosis of liver cancer[J]. Nature Machine Intelligence,2023.

Liang，Junhao.,Zhang，Weisheng.,Yang，Jianghui.,Wu，Meilong.,Dai，Qionghai.,...&Kong，Lingjie.(2023).Deep learning supported discovery of biomarkers for clinical prognosis of liver cancer.Nature Machine Intelligence.

Liang，Junhao,et al."Deep learning supported discovery of biomarkers for clinical prognosis of liver cancer".Nature Machine Intelligence (2023).