Transformer optimization with meta learning on pathology images for breast cancer lymph node micrometastasis
Abstract Lymph node micro-metastasis represents the initial stage of breast cancer spread or metastasis. However, the limited size of these hidden lesions restricts dataset expansion, presenting a significant challenge for manual examination and conventional deep learning techniques. By harnessing t...
Saved in:
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-07-01
|
| Series: | npj Digital Medicine |
| Online Access: | https://doi.org/10.1038/s41746-025-01833-6 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849332417148485632 |
|---|---|
| author | Jing Huang Jingtao Wang Junhai Shi Hengli Ni Shan Xu Ping Wu Yuexiang Ren Lijuan Bian Chenhan Su Yuxuan Xu Xinyu He Xinjian Chen Jianming Li |
| author_facet | Jing Huang Jingtao Wang Junhai Shi Hengli Ni Shan Xu Ping Wu Yuexiang Ren Lijuan Bian Chenhan Su Yuxuan Xu Xinyu He Xinjian Chen Jianming Li |
| author_sort | Jing Huang |
| collection | DOAJ |
| description | Abstract Lymph node micro-metastasis represents the initial stage of breast cancer spread or metastasis. However, the limited size of these hidden lesions restricts dataset expansion, presenting a significant challenge for manual examination and conventional deep learning techniques. By harnessing the power of meta-learning on limited datasets, we developed a novel network named MetaTrans, equipped with a 34-category dataset (MT-MCD) to effectively pinpoint micro-metastases in lymph nodes from pathological images. MetaTrans demonstrated superior performance on two different multi-center datasets and excelled in the 0-shot task for intraoperative frozen section diagnosis. Beyond breast cancer, MetaTrans efficiently identifies micro-metastases in thyroid and colorectal cancers and can be directly applied to recognize images captured by digital cameras under a microscope. Across all clinical validation scenarios, our method surpasses state-of-the-art baselines, exhibiting robust cross-domain adaptation and task-specific reliability, which highlight its translational potential in diverse pathological settings. |
| format | Article |
| id | doaj-art-d2bc75a8b4ac4a2b960c03c6f3c85802 |
| institution | Kabale University |
| issn | 2398-6352 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | npj Digital Medicine |
| spelling | doaj-art-d2bc75a8b4ac4a2b960c03c6f3c858022025-08-20T03:46:12ZengNature Portfolionpj Digital Medicine2398-63522025-07-018111510.1038/s41746-025-01833-6Transformer optimization with meta learning on pathology images for breast cancer lymph node micrometastasisJing Huang0Jingtao Wang1Junhai Shi2Hengli Ni3Shan Xu4Ping Wu5Yuexiang Ren6Lijuan Bian7Chenhan Su8Yuxuan Xu9Xinyu He10Xinjian Chen11Jianming Li12Department of Pathology, Soochow Medical College, Soochow UniversitySchool of Electronics and Information Engineering, Soochow UniversityDepartment of Pathology, Soochow Medical College, Soochow UniversityDepartment of Pathology, Children’s Hospital, Soochow UniversityDepartment of Pathology and Institute of Molecular Pathology, Jiangxi Provincial Key Laboratory for Precision Pathology and Intelligent Diagnosis, The First Affiliated Hospital, Jiangxi Medical College, Nanchang UniversityDepartment of Pathology and Institute of Molecular Pathology, Jiangxi Provincial Key Laboratory for Precision Pathology and Intelligent Diagnosis, The First Affiliated Hospital, Jiangxi Medical College, Nanchang UniversityDepartment of Pathology and Institute of Molecular Pathology, Jiangxi Provincial Key Laboratory for Precision Pathology and Intelligent Diagnosis, The First Affiliated Hospital, Jiangxi Medical College, Nanchang UniversityDepartment of Pathology, Sun Yat-sen Memorial Hospital, Sun Yat-sen UniversityDepartment of Pathology, Soochow Medical College, Soochow UniversityDepartment of Pathology, Soochow Medical College, Soochow UniversityDepartment of Pathology, Soochow Medical College, Soochow UniversitySchool of Electronics and Information Engineering, Soochow UniversityDepartment of Pathology, Soochow Medical College, Soochow UniversityAbstract Lymph node micro-metastasis represents the initial stage of breast cancer spread or metastasis. However, the limited size of these hidden lesions restricts dataset expansion, presenting a significant challenge for manual examination and conventional deep learning techniques. By harnessing the power of meta-learning on limited datasets, we developed a novel network named MetaTrans, equipped with a 34-category dataset (MT-MCD) to effectively pinpoint micro-metastases in lymph nodes from pathological images. MetaTrans demonstrated superior performance on two different multi-center datasets and excelled in the 0-shot task for intraoperative frozen section diagnosis. Beyond breast cancer, MetaTrans efficiently identifies micro-metastases in thyroid and colorectal cancers and can be directly applied to recognize images captured by digital cameras under a microscope. Across all clinical validation scenarios, our method surpasses state-of-the-art baselines, exhibiting robust cross-domain adaptation and task-specific reliability, which highlight its translational potential in diverse pathological settings.https://doi.org/10.1038/s41746-025-01833-6 |
| spellingShingle | Jing Huang Jingtao Wang Junhai Shi Hengli Ni Shan Xu Ping Wu Yuexiang Ren Lijuan Bian Chenhan Su Yuxuan Xu Xinyu He Xinjian Chen Jianming Li Transformer optimization with meta learning on pathology images for breast cancer lymph node micrometastasis npj Digital Medicine |
| title | Transformer optimization with meta learning on pathology images for breast cancer lymph node micrometastasis |
| title_full | Transformer optimization with meta learning on pathology images for breast cancer lymph node micrometastasis |
| title_fullStr | Transformer optimization with meta learning on pathology images for breast cancer lymph node micrometastasis |
| title_full_unstemmed | Transformer optimization with meta learning on pathology images for breast cancer lymph node micrometastasis |
| title_short | Transformer optimization with meta learning on pathology images for breast cancer lymph node micrometastasis |
| title_sort | transformer optimization with meta learning on pathology images for breast cancer lymph node micrometastasis |
| url | https://doi.org/10.1038/s41746-025-01833-6 |
| work_keys_str_mv | AT jinghuang transformeroptimizationwithmetalearningonpathologyimagesforbreastcancerlymphnodemicrometastasis AT jingtaowang transformeroptimizationwithmetalearningonpathologyimagesforbreastcancerlymphnodemicrometastasis AT junhaishi transformeroptimizationwithmetalearningonpathologyimagesforbreastcancerlymphnodemicrometastasis AT henglini transformeroptimizationwithmetalearningonpathologyimagesforbreastcancerlymphnodemicrometastasis AT shanxu transformeroptimizationwithmetalearningonpathologyimagesforbreastcancerlymphnodemicrometastasis AT pingwu transformeroptimizationwithmetalearningonpathologyimagesforbreastcancerlymphnodemicrometastasis AT yuexiangren transformeroptimizationwithmetalearningonpathologyimagesforbreastcancerlymphnodemicrometastasis AT lijuanbian transformeroptimizationwithmetalearningonpathologyimagesforbreastcancerlymphnodemicrometastasis AT chenhansu transformeroptimizationwithmetalearningonpathologyimagesforbreastcancerlymphnodemicrometastasis AT yuxuanxu transformeroptimizationwithmetalearningonpathologyimagesforbreastcancerlymphnodemicrometastasis AT xinyuhe transformeroptimizationwithmetalearningonpathologyimagesforbreastcancerlymphnodemicrometastasis AT xinjianchen transformeroptimizationwithmetalearningonpathologyimagesforbreastcancerlymphnodemicrometastasis AT jianmingli transformeroptimizationwithmetalearningonpathologyimagesforbreastcancerlymphnodemicrometastasis |