Modern Backbones Improve Multi-task DETR for Mammography Classification and Lesion Localization

Dinh Tan Nguyen, Hien Quang Kha, Le-Hoang Nguyen, Minh-Toan Dinh, Xuan-Huy Nguyen, Phu. D Ho, Cao Truong Tran, Sai Ho Ling, Lan T. Ho-Pham, Liem Tan Pham, Nguyen Quoc Khanh Le

Keywords: Mammography, multi-task learning, object detection, classification, DETR, lesion localization

Registration Requirement: Yes

Abstract: Joint exam-level prediction and candidate-region localization may improve the usefulness of AI support in mammography. We study this setting using a multi-task DETR framework, where shared representations support both image-level malignancy prediction and lesion localization, and evaluate its performance on OPTIMAM and a biopsy-confirmed SGM1k cohort. Across both datasets, modern backbones consistently outperformed older ResNet-style features, with ConvNeXtV2 and DINOv3 giving the strongest overall results, whereas MambaVision was less competitive. On OPTIMAM, ConvNeXtV2 achieved the best overall performance, reaching 97.96% AUC, 99.89% sensitivity, 25.08% mAP@.5, and 74.38% recall@.25. On SGM1k, DINOv3 gave the strongest overall results, with 90.97% AUC, 86.28% sensitivity, 82.00% specificity, 27.04% mAP@.5, and 77.32% recall@.25. These findings suggest that backbone quality is a critical factor in effective multi-task mammography, with ConvNeXtV2 emerging as a particularly strong and well-matched CNN backbone for mammography in this framework.

Reproducibility: https://github.com/saigonmec/mammo2detr

Visa & Travel: Yes

Read CFP & Author Instructions: Yes

Originality Policy: Yes

Single-blind & Not Under Review Elsewhere: Yes

LLM Policy: Yes

Submission Number: 33