arXiv：cs.AI· Abdullah Al Shafi, Md Kawsar Mahmud Khan Zunayed, Safin Ahmmed, Sk Imran Hossain, Engelbert Mephu Nguifo·· 6 小时前AI 评分37
面向乳腺超声分割与分类的自适应双向任务交互方法
Adaptive Bidirectional Task Interaction for Joint Segmentation and Classification of Breast Ultrasound
AI 导读
该方法在解码阶段恢复分割与分类分支间的信息交换，并通过任务交互模块（TIM）与自适应交互加权（AIW）按图像动态融合特征。在 BUSI 上达到 74.19% IoU 和 90.60% 准确率，在 BUSI-WHU 上达到 86.40% IoU 和 95.00% 准确率，优于共享编码器多任务、Transformer 分割及解码器交互基线。
正文
Abstract:Joint lesion segmentation and tissue classification in breast ultrasound are usually trained with a shared encoder, so the two branches stop exchanging information once their decoders separate. That is exactly where boundary detail and semantic evidence are most complementary. The proposed method restores this exchange during decoding and, because its value differs between images, lets the network decide per image how much to keep. A Task Interaction Module (TIM) at each of four decoder levels passes pooled boundary context into the classification representation and modulates decoder channels with class-conditioned priors. An Adaptive Interaction Weighting (AIW) unit then blends interacted and original features with a coefficient computed for each image and level. On BUSI the model reaches 74.19% IoU and 90.60% accuracy, and on BUSI-WHU 86.40% IoU and 95.00% accuracy, ahead of encoder-sharing multi-task, transformer segmentation and decoder-interaction baselines evaluated under the same protocol. The ablation shows that multi-scale context and cross-task exchange are not independent: applied separately they contribute 4.00 points of IoU in total, applied together 6.76. Adding the adaptive blend to task interaction alone raises AUC from 94.41% to 97.31%, indicating that the blend acts primarily on the classification branch. Code: this https URL.
| Comments: | 10 pages, 2 figures, 2 tables |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2603.01295 [cs.CV] |
| (or arXiv:2603.01295v2 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2603.01295 arXiv-issued DOI via DataCite |
Submission history
From: Abdullah Al Shafi [view email]
[v1]
Sun, 1 Mar 2026 22:02:06 UTC (1,752 KB)
[v2]
Tue, 6 Oct 2026 13:32:02 UTC (1,842 KB)
来源：arXiv：cs.AI · arxiv.org