用小波边缘图精准定位结肠息肉弱边界，提升分割准确率。
MEGANet-W: A Wavelet-Driven Edge-Guided Attention Framework for Weak Boundary Polyp Detection
- 通过小波变换生成无参数边缘图，引导网络关注边界信息。
- 在5个数据集上平均提升mIoU 2.3%、mDice 1.2%，无额外参数。
- 适合需要高精度边界的医学图像分割任务，如早期癌症检测。
结直肠息肉分割对早期结直肠癌检测至关重要，但弱对比度和模糊边界严重限制了自动分割的准确性。现有深度模型或模糊细粒度边缘细节，或依赖手工设计滤波器，在不同成像条件下表现不佳。本文提出MEGANet-W，一种基于小波驱动的边缘引导注意力网络，将方向性、无参数的Haar小波边缘图注入解码器每个阶段，以重新校准语义特征。其核心创新包括两层Haar小波头实现多方向边缘提取，以及融合小波线索与边界/输入分支的波形边缘引导注意力（W-EGA）模块。在五个公开息肉数据集上，MEGANet-W始终优于现有方法，mIoU最高提升2.3%，mDice提升1.2%，且不引入额外可学习参数。该方法显著提升了复杂情况下的分割可靠性，为需要精确边界检测的医学图像分割任务提供了稳健解决方案。
原文摘要 · Abstract (English)
Colorectal polyp segmentation is critical for early detection of colorectal cancer, yet weak and low contrast boundaries significantly limit automated accuracy. Existing deep models either blur fine edge details or rely on handcrafted filters that perform poorly under variable imaging conditions. We propose MEGANet-W, a Wavelet Driven Edge Guided Attention Network that injects directional, parameter free Haar wavelet edge maps into each decoder stage to recalibrate semantic features. The key novelties of MEGANet-W include a two-level Haar wavelet head for multi-orientation edge extraction; and Wavelet Edge Guided Attention (W-EGA) modules that fuse wavelet cues with boundary and input branches. On five public polyp datasets, MEGANet-W consistently outperforms existing methods, improving mIoU by up to 2.3% and mDice by 1.2%, while introducing no additional learnable parameters. This approach improves reliability in difficult cases and offers a robust solution for medical image segmentation tasks requiring precise boundary detection.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。