HyFormer-Net: A Synergistic CNN-Transformer with Interpretable Multi-Scale Fusion for Breast Lesion Segmentation and Classification in Ultrasound Images Journal: Computerized Medical Imaging and Graphics
Mohammad Amanour Rahman · 2025
Read HyFormer-Net: A Synergistic CNN-Transformer with Interpretable Multi-Scale Fusion for Breast Lesion Segmentation and Classification in Ultrasound Images Journal: Computerized Medical Imaging and Graphics by Mohammad Amanour Rahman (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-01013.txt.
Description
Mohammad Amanour Rahman proposes HyFormer-Net, a neural network for breast ultrasound that finds lesions and classifies them at the same time, while showing its reasons. Ultrasound diagnosis is hampered by speckle noise, operator dependence, and fuzzy lesion borders. Earlier deep learning systems often did one task. They missed global context or local detail, and they acted as black boxes.
The design joins a convolutional branch for local features with a transformer branch for global context, fused across scales, with an attention gated decoder. Interpretability comes from attention checks and from Grad-CAM heat maps for the classification side. On the BUSI data set it beat standard baselines on overlap and accuracy, with strong recall on malignant cases, and an ensemble version performed better still.
Catalog record
| CREATOR | Mohammad Amanour Rahman |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 60 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-01013