用自适应路由提升扩散模型效率，精准预测胆管癌神经侵犯
Adaptive Routing for Efficient Diffusion Transformer-Based PNI Prediction
- 通过自适应路由优化注意力头、空间标记和MLP宽度，提升计算效率
- 在MRI数据上实现0.731的AUC，仅需257.57 GFLOPs
- 适合需要高精度且低算力的医学影像辅助诊断场景
胆管癌的神经周围浸润（PNI）是重要预后指标，但术前基于磁共振成像（MRI）预测仍具挑战，因其影像特征细微且超出肿瘤边界。传统卷积网络难以捕捉长程空间依赖，虽基于Transformer的架构能建模体积MRI的全局上下文，但在瘤周区域捕捉细微、易受噪声干扰的模式仍困难。扩散类分类器通过去噪评分机制可更好识别此类模式，但结合Transformer与迭代去噪过程带来巨大计算开销。为此，本文将PNI预测建模为扩散分类问题，采用Transformer表示去噪网络，并引入跨注意力头、空间标记及MLP宽度的自适应路由策略以提升效率。实验表明，该方法在保持0.731 AUC的同时，仅需257.57 GFLOPs计算量。
原文摘要 · Abstract (English)
Perineural invasion (PNI) is a critical prognostic factor in cholangiocarcinoma. However, its preoperative prediction from magnetic resonance imaging (MRI) remains challenging due to subtle imaging features that extend beyond tumor boundaries into surrounding regions. Conventional convolutional neural networks are limited in capturing long-range spatial dependencies. Transformer-based architectures improve global modeling of volumetric MRI by aggregating spatially distributed contextual cues, yet capturing subtle and noise-sensitive patterns in peritumoral regions remains challenging. Diffusion-based classifiers offer an alternative formulation by leveraging denoising-based class scoring to better capture such subtle patterns. However, these approaches introduce substantial computational overhead due to the combination of transformer-based modeling and iterative denoising processes. To address these challenges, we formulate PNI prediction as a diffusion-based classification problem and implement the denoising network using a transformer-based representation. To improve computational efficiency, we introduce adaptive routing across attention heads, spatial tokens, and MLP width. Experimental results demonstrate that the proposed approach achieves an AUC of 0.731 with 257.57 GFLOPs.
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。