Toward a robust lesion detection model in breast DCE-MRI: adapting foundation models to high-risk women
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912568582340608 |
|---|---|
| author | Nascimento, Gabriel A. B. do Dong, Vincent Cavalcante, Guilherme J. Nguyen, Alex Rêgo, Thaís G. do Malheiros, Yuri Filho, Telmo M. Silva Torrez, Carla R. Zeballos Gee, James C. McCarthy, Anne Marie Maidment, Andrew D. A. Barufaldi, Bruno |
| author_facet | Nascimento, Gabriel A. B. do Dong, Vincent Cavalcante, Guilherme J. Nguyen, Alex Rêgo, Thaís G. do Malheiros, Yuri Filho, Telmo M. Silva Torrez, Carla R. Zeballos Gee, James C. McCarthy, Anne Marie Maidment, Andrew D. A. Barufaldi, Bruno |
| contents | Accurate breast MRI lesion detection is critical for early cancer diagnosis, especially in high-risk populations. We present a classification pipeline that adapts a pretrained foundation model, the Medical Slice Transformer (MST), for breast lesion classification using dynamic contrast-enhanced MRI (DCE-MRI). Leveraging DINOv2-based self-supervised pretraining, MST generates robust per-slice feature embeddings, which are then used to train a Kolmogorov--Arnold Network (KAN) classifier. The KAN provides a flexible and interpretable alternative to conventional convolutional networks by enabling localized nonlinear transformations via adaptive B-spline activations. This enhances the model's ability to differentiate benign from malignant lesions in imbalanced and heterogeneous clinical datasets. Experimental results demonstrate that the MST+KAN pipeline outperforms the baseline MST classifier, achieving AUC = 0.80 \pm 0.02 while preserving interpretability through attention-based heatmaps. Our findings highlight the effectiveness of combining foundation model embeddings with advanced classification strategies for building robust and generalizable breast MRI analysis tools. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_02710 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Toward a robust lesion detection model in breast DCE-MRI: adapting foundation models to high-risk women Nascimento, Gabriel A. B. do Dong, Vincent Cavalcante, Guilherme J. Nguyen, Alex Rêgo, Thaís G. do Malheiros, Yuri Filho, Telmo M. Silva Torrez, Carla R. Zeballos Gee, James C. McCarthy, Anne Marie Maidment, Andrew D. A. Barufaldi, Bruno Medical Physics Computer Vision and Pattern Recognition Machine Learning Accurate breast MRI lesion detection is critical for early cancer diagnosis, especially in high-risk populations. We present a classification pipeline that adapts a pretrained foundation model, the Medical Slice Transformer (MST), for breast lesion classification using dynamic contrast-enhanced MRI (DCE-MRI). Leveraging DINOv2-based self-supervised pretraining, MST generates robust per-slice feature embeddings, which are then used to train a Kolmogorov--Arnold Network (KAN) classifier. The KAN provides a flexible and interpretable alternative to conventional convolutional networks by enabling localized nonlinear transformations via adaptive B-spline activations. This enhances the model's ability to differentiate benign from malignant lesions in imbalanced and heterogeneous clinical datasets. Experimental results demonstrate that the MST+KAN pipeline outperforms the baseline MST classifier, achieving AUC = 0.80 \pm 0.02 while preserving interpretability through attention-based heatmaps. Our findings highlight the effectiveness of combining foundation model embeddings with advanced classification strategies for building robust and generalizable breast MRI analysis tools. |
| title | Toward a robust lesion detection model in breast DCE-MRI: adapting foundation models to high-risk women |
| topic | Medical Physics Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2509.02710 |