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

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 WangYaobing Chen, Guoping Wang, Tian Xia

Thoracic Head & Neck Medical Oncology

Research output: Contribution to journal › Article › peer-review

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.

Original language	English (US)
Article number	107243
Journal	iScience
Volume	26
Issue number	10
DOIs	https://doi.org/10.1016/j.isci.2023.107243
State	Published - Oct 20 2023

Keywords

Cancer
Histology
Machine learning
Pathology

ASJC Scopus subject areas

General

Access to Document

10.1016/j.isci.2023.107243

Cite this

Gan, J., Wang, H., Yu, H., He, Z., Zhang, W., Ma, K., Zhu, L., Bai, Y., Zhou, Z., Yullie, A., Bai, X., Wang, M., Yang, D., Chen, Y., Chen, G., Lasenby, J., Cheng, C., Wu, J., Zhang, J., ... Xia, T. (2023). Focalizing regions of biomarker relevance facilitates biomarker prediction on histopathological images. iScience, 26(10), Article 107243. https://doi.org/10.1016/j.isci.2023.107243

Gan, J, Wang, H, Yu, H, He, Z, Zhang, W, Ma, K, Zhu, L, Bai, Y, Zhou, Z, Yullie, A, Bai, X, Wang, M, Yang, D, Chen, Y, Chen, G, Lasenby, J, Cheng, C, Wu, J , Zhang, J, Wang, X, Chen, Y, Wang, G & Xia, T 2023, 'Focalizing regions of biomarker relevance facilitates biomarker prediction on histopathological images', iScience, vol. 26, no. 10, 107243. https://doi.org/10.1016/j.isci.2023.107243

@article{9d0bd326c42b428dbcf87ed30e415f70,

title = "Focalizing regions of biomarker relevance facilitates biomarker prediction on histopathological images",

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",

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

note = "Publisher Copyright: {\textcopyright} 2023 The Author(s)",

year = "2023",

month = oct,

day = "20",

doi = "10.1016/j.isci.2023.107243",

language = "English (US)",

volume = "26",

journal = "iScience",

issn = "2589-0042",

publisher = "Elsevier Inc.",

number = "10",

}

TY - JOUR

T1 - Focalizing regions of biomarker relevance facilitates biomarker prediction on histopathological images

AU - Gan, Jiefeng

AU - Wang, Hanchen

AU - Yu, Hui

AU - He, Zitong

AU - Zhang, Wenjuan

AU - Ma, Ke

AU - Zhu, Lianghui

AU - Bai, Yutong

AU - Zhou, Zongwei

AU - Yullie, Alan

AU - Bai, Xiang

AU - Wang, Mingwei

AU - Yang, Dehua

AU - Chen, Yanyan

AU - Chen, Guoan

AU - Lasenby, Joan

AU - Cheng, Chao

AU - Wu, Jia

AU - Zhang, Jianjun

AU - Wang, Xinggang

AU - Chen, Yaobing

AU - Wang, Guoping

AU - Xia, Tian

PY - 2023/10/20

Y1 - 2023/10/20

N2 - 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.

AB - 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.

KW - Cancer

KW - Histology

KW - Machine learning

KW - Pathology

UR - http://www.scopus.com/inward/record.url?scp=85171801382&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85171801382&partnerID=8YFLogxK

U2 - 10.1016/j.isci.2023.107243

DO - 10.1016/j.isci.2023.107243

M3 - Article

C2 - 37767002

AN - SCOPUS:85171801382

SN - 2589-0042

VL - 26

JO - iScience

JF - iScience

IS - 10

M1 - 107243

ER -

Focalizing regions of biomarker relevance facilitates biomarker prediction on histopathological images

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this