A large model for non-invasive and personalized management of breast cancer from multiparametric MRI

Abstract Breast Magnetic Resonance Imaging (MRI) demonstrates the highest sensitivity for breast cancer detection among imaging modalities and is standard practice for high-risk women. Interpreting the multi-sequence MRI is time-consuming and prone to subjective variation. We develop a large mixture...

Full description

Saved in:
Bibliographic Details
Main Authors: Luyang Luo, Mingxiang Wu, Mei Li, Yi Xin, Qiong Wang, Varut Vardhanabhuti, Winnie CW Chu, Zhenhui Li, Juan Zhou, Pranav Rajpurkar, Hao Chen
Format: Article
Language:English
Published: Nature Portfolio 2025-04-01
Series:Nature Communications
Online Access:https://doi.org/10.1038/s41467-025-58798-z
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849699144288960512
author Luyang Luo
Mingxiang Wu
Mei Li
Yi Xin
Qiong Wang
Varut Vardhanabhuti
Winnie CW Chu
Zhenhui Li
Juan Zhou
Pranav Rajpurkar
Hao Chen
author_facet Luyang Luo
Mingxiang Wu
Mei Li
Yi Xin
Qiong Wang
Varut Vardhanabhuti
Winnie CW Chu
Zhenhui Li
Juan Zhou
Pranav Rajpurkar
Hao Chen
author_sort Luyang Luo
collection DOAJ
description Abstract Breast Magnetic Resonance Imaging (MRI) demonstrates the highest sensitivity for breast cancer detection among imaging modalities and is standard practice for high-risk women. Interpreting the multi-sequence MRI is time-consuming and prone to subjective variation. We develop a large mixture-of-modality-experts model (MOME) that integrates multiparametric MRI information within a unified structure, leveraging breast MRI scans from 5205 female patients in China for model development and validation. MOME matches four senior radiologists’ performance in identifying breast cancer and outperforms a junior radiologist. The model is able to reduce unnecessary biopsies in Breast Imaging-Reporting and Data System (BI-RADS) 4 patients, classify triple-negative breast cancer, and predict pathological complete response to neoadjuvant chemotherapy. MOME further supports inference with missing modalities and provides decision explanations by highlighting lesions and measuring modality contributions. To summarize, MOME exemplifies an accurate and robust multimodal model for noninvasive, personalized management of breast cancer patients via multiparametric MRI.
format Article
id doaj-art-8b1317e31b8046b881fd3eb0c16631da
institution DOAJ
issn 2041-1723
language English
publishDate 2025-04-01
publisher Nature Portfolio
record_format Article
series Nature Communications
spelling doaj-art-8b1317e31b8046b881fd3eb0c16631da2025-08-20T03:18:42ZengNature PortfolioNature Communications2041-17232025-04-0116111410.1038/s41467-025-58798-zA large model for non-invasive and personalized management of breast cancer from multiparametric MRILuyang Luo0Mingxiang Wu1Mei Li2Yi Xin3Qiong Wang4Varut Vardhanabhuti5Winnie CW Chu6Zhenhui Li7Juan Zhou8Pranav Rajpurkar9Hao Chen10Department of Computer Science and Technology, The Hong Kong University of Science and TechnologyDepartment of Radiology, Shenzhen People’s HospitalDepartment of Radiology, PLA Middle Military Command General HospitalDepartment of Computer Science and Technology, The Hong Kong University of Science and TechnologyShenzhen Institute of Advanced Technology, Chinese Academy of SciencesDepartment of Diagnostic Radiology, Li Ka Shing Faculty of Medicine, The University of Hong KongDepartment of Imaging and Interventional Radiology, The Chinese University of Hong KongDepartment of Radiology, the Third Affiliated Hospital of Kunming Medical University, Yunnan Cancer Hospital, Yunnan Cancer CenterDepartment of Radiology, 5th Medical Center of Chinese PLA General HospitalDepartment of Biomedical Informatics, Harvard UniversityDepartment of Computer Science and Technology, The Hong Kong University of Science and TechnologyAbstract Breast Magnetic Resonance Imaging (MRI) demonstrates the highest sensitivity for breast cancer detection among imaging modalities and is standard practice for high-risk women. Interpreting the multi-sequence MRI is time-consuming and prone to subjective variation. We develop a large mixture-of-modality-experts model (MOME) that integrates multiparametric MRI information within a unified structure, leveraging breast MRI scans from 5205 female patients in China for model development and validation. MOME matches four senior radiologists’ performance in identifying breast cancer and outperforms a junior radiologist. The model is able to reduce unnecessary biopsies in Breast Imaging-Reporting and Data System (BI-RADS) 4 patients, classify triple-negative breast cancer, and predict pathological complete response to neoadjuvant chemotherapy. MOME further supports inference with missing modalities and provides decision explanations by highlighting lesions and measuring modality contributions. To summarize, MOME exemplifies an accurate and robust multimodal model for noninvasive, personalized management of breast cancer patients via multiparametric MRI.https://doi.org/10.1038/s41467-025-58798-z
spellingShingle Luyang Luo
Mingxiang Wu
Mei Li
Yi Xin
Qiong Wang
Varut Vardhanabhuti
Winnie CW Chu
Zhenhui Li
Juan Zhou
Pranav Rajpurkar
Hao Chen
A large model for non-invasive and personalized management of breast cancer from multiparametric MRI
Nature Communications
title A large model for non-invasive and personalized management of breast cancer from multiparametric MRI
title_full A large model for non-invasive and personalized management of breast cancer from multiparametric MRI
title_fullStr A large model for non-invasive and personalized management of breast cancer from multiparametric MRI
title_full_unstemmed A large model for non-invasive and personalized management of breast cancer from multiparametric MRI
title_short A large model for non-invasive and personalized management of breast cancer from multiparametric MRI
title_sort large model for non invasive and personalized management of breast cancer from multiparametric mri
url https://doi.org/10.1038/s41467-025-58798-z
work_keys_str_mv AT luyangluo alargemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT mingxiangwu alargemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT meili alargemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT yixin alargemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT qiongwang alargemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT varutvardhanabhuti alargemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT winniecwchu alargemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT zhenhuili alargemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT juanzhou alargemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT pranavrajpurkar alargemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT haochen alargemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT luyangluo largemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT mingxiangwu largemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT meili largemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT yixin largemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT qiongwang largemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT varutvardhanabhuti largemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT winniecwchu largemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT zhenhuili largemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT juanzhou largemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT pranavrajpurkar largemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri
AT haochen largemodelfornoninvasiveandpersonalizedmanagementofbreastcancerfrommultiparametricmri