Multiparametric MRI and transfer learning for predicting positive margins in breast-conserving surgery: a multi-center study.

Journal: International journal of surgery (London, England)
PMID:

Abstract

This study aimed to predict positive surgical margins in breast-conserving surgery (BCS) using multiparametric MRI (mpMRI) and radiomics. A retrospective analysis was conducted on data from 444 BCS patients from three Chinese hospitals between 2019 and 2024, divided into four cohorts and five datasets. Radiomics features from preoperative mpMRI, along with clinicopathological data, were extracted and selected using statistical methods and LASSO logistic regression. Eight machine learning classifiers, integrated with a transfer learning (TL) method, were applied to enhance model generalization. The model achieved an AUC of 0.889 in the internal test set and 0.771 in the validation set. Notably, TL significantly improved performance in two external validation sets, increasing the AUC from 0.533 to 0.902 in XAH and from 0.359 to 0.855 in YNCH. These findings highlight the potential of combining mpMRI and TL to provide accurate predictions for positive surgical margins in BCS, with promising implications for broader clinical application across multiple hospitals.

Authors

  • Xue Zhao
    Fujian Key Laboratory of Precision Diagnosis and Treatment in Breast Cancer, Xiang'an Hospital of Xiamen University, School of Medicine, Xiamen University, Xiamen, China; National Institute for Data Science in Health and Medicine, Xiamen University, Xiamen, China; Department of Breast-Thyroid-Surgery and Cancer Center, Xiang'an Hospital of Xiamen University, School of Medicine, Xiamen University, Xiamen, China; Xiamen Research Center of Clinical Medicine in Breast & Thyroid Cancers, Xiang'an Hospital of Xiamen University, School of Medicine, Xiamen University, Xiamen, China; Xiamen Key Laboratory of Endocrine-Related Cancer Precision Medicine, Xiang'an Hospital of Xiamen University, School of Medicine, Xiamen University, Xiamen, China.
  • Jing-Wen Bai
    Fujian Key Laboratory of Precision Diagnosis and Treatment in Breast Cancer, Xiang'an Hospital of Xiamen University, School of Medicine, Xiamen University, 361100, Xiamen, China.
  • Sen Jiang
    Department of Radiology, Cancer Hospital of Shantou University Medical College, Shantou, China.
  • Zhen-Hui Li
    Department of Radiology, Yunnan Cancer Hospital & The Third Affiliated Hospital of Kunming Medical University & Peking University Cancer Hospital Yunnan, Kunming, China,Xiang'an Hospital of Xiamen University, School of Medicine, Xiamen University, Xiamen, China.
  • Jie-Zhou He
    Institute of Artificial Intelligence, Xiamen University, Xiamen, China and.
  • Zhi-Cheng Du
    Fujian Key Laboratory of Precision Diagnosis and Treatment in Breast Cancer & Xiamen Key Laboratory of Endocrine-Related Cancer Precision Medicine, Xiang'an Hospital of Xiamen University, School of Medicine, Xiamen University, Xiamen, China.
  • Xue-Qi Fan
    Fujian Key Laboratory of Precision Diagnosis and Treatment in Breast Cancer & Xiamen Key Laboratory of Endocrine-Related Cancer Precision Medicine, Xiang'an Hospital of Xiamen University, School of Medicine, Xiamen University, Xiamen, China.
  • Shao-Zi Li
    Department of Artificial Intelligence, Xiamen University, Xiamen, China.
  • Guo-Jun Zhang
    Changjiang Scholar's Laboratory, Shantou University Medical College, Guangdong, China. Electronic address: guoj_zhang@yahoo.com.