Cervical cancer remains a major malignancy in women, with large-scale human papillomavirus-based screening generating a large volume of biopsies and, for advanced cases, surgical specimens, imposing an overwhelming diagnostic burden that necessitates precise histopathological evaluation. Although deep learning shows promise, general foundation models offer a broader reach but remain limited in capturing subspecialty-specific features and task adaptability. We introduce the Cervical Subspecialty Pathology (CerS-Path) diagnostic system, built upon the subspecialized dataset and developed through two synergistic pretraining stages: self-supervised learning on approximately 190 million tissue patches to build a cervical-specific feature extractor, and multimodal enhancement with 2.5 million image-text pairs, followed by diagnostic functions integration and downstream task enhancement. CerS-Path integrates eight functions—across morphological diagnosis, predictive analytics, and multimodal tasks—to cover the diagnostic workflow from biopsies to resection specimens. Comprehensive evaluations demonstrate improved performance in cervical pathology in female patients, where CerS-Path outperforms state-of-the-art foundation models across 26 of 30 diagnostic tasks. Furthermore, in multicenter retrospective and prospective validation across six independent cohorts of women, CerS-Path achieves up to 99% sensitivity for clinically relevant lesion detection with acceptable specificity across sites, highlighting its potential for screening-derived histopathology and suitability for integration into high-volume diagnostic pathways.
We extend special thanks to the teams at Ningbo Clinical Pathology Diagnosis Center, the First Affiliated Hospital of Soochow University, Affiliated Jinhua Hospital of Zhejiang University School of Medicine, Xiaogan Central Hospital of Wuhan University of Science and Technology, Hubei Maternal and Child Health Hospital, Guangdong Women and Children Hospital, and Hangzhou Adicon Clinical Laboratories for their invaluable support in data collection, annotation, and validation.
This work was supported by the National Natural Science Foundation of China (grant no. 82430062 to T.G.), the Shenzhen Engineering Research Centre (grant no. XMHT20230115004 to Y.H.H.), the National Natural Science Foundation of China Youth Program (grant no. 62401329 to D.W.S.), the Tsinghua Shenzhen International Graduate School Cross-disciplinary Research and Innovation Fund Research Plan (grant no. JC2024002 to Y.H.H.), the Ningbo Top Medical and Health Research Program (grant no. 2023010211 to M.Z.), the Natural Science Foundation of Jiangsu Province (grant no. BK20241793 to L.C.G.), and the Suzhou Science and Technology Development Program (grant no. SKY2023009 to L.C.G.). Additional support was provided by Jilin FuyuanGuan Food Group Co., Ltd. through the joint research centre. C.H. acknowledges support from the University of Oxford and the Royal Society through a University Research Fellowship (URF/R1/241734).
These authors contributed equally: Yizhi Wang, Li Chen, Qiang Huang, Tian Guan.
Source: Nature - Machine Learning — Read the full article on the original site ↗