Breast Cancer Segmentation and Diagnosis in Mammography: Toward a Pan-European Virtual Biopsy Framework
Abstract
Background
Most mammographic AI stops at a binary cancer decision and ignores the histopathologic information that guides therapy.
Purpose
To build an end-to-end virtual-biopsy pipeline that simultaneously segments tumours and predicts pT, pN, cM and molecular subtype across scanners and countries.
Materials and Methods
A residual U-Net was trained on 8 469 public mammograms and externally tested on 200 cases from six ChAImeleon sites. From each mask we derived 1 559 radiomics, 5 120 RadImageNet embeddings and 12 clinical covariates; a 5 × 3 nested cross-validation tuned an RBF-SVM/XGBoost/MLP ensemble with a random-forest meta-learner.
Results
External Dice was 0.458; adding benign masks cut false-positive pixels by 19% but left Dice unchanged (0.455).
Related articles
Related articles are currently not available for this article.