发表机构
Seoul National University; Kyung Hee University; OUTTA; Chung-Ang University; Seoul National University School of Medicine; Samsung Changwon Hospital; Samsung Medical Center; NVIDIA AI Technology Center（首尔国立大学; 庆熙大学; OUTTA; Chung-Ang 大学; 首尔国立大学医学院; 三星昌原医院; 三星医疗中心; NVIDIA AI 技术中心）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对胆管癌PNI术前MRI预测难题，传统方法有局限。本文将PNI预测设为扩散分类问题，用基于Transformer的表示实现去噪网络，并引入自适应路由提高效率，实验取得了0.731的AUC及257.57 GFLOPs的结果。
AI 中文摘要
神经周围侵犯（PNI）是胆管癌的关键预后因素。然而，由于细微的成像特征超出肿瘤边界延伸到周围区域，从磁共振成像（MRI）进行术前预测仍然具有挑战性。传统卷积神经网络在捕捉远距离空间依赖性方面有限。基于Transformer的架构通过聚合空间分布的上下文线索改进了体积MRI的全局建模，但在肿瘤周围区域捕捉细微和噪声敏感模式仍具挑战。基于扩散的分类器通过利用基于去噪的类评分提供了一种替代方案。但这些方法由于基于Transformer的建模和迭代去噪过程的结合而引入了大量计算开销。为应对这些挑战，我们将PNI预测公式化为基于扩散的分类问题，并使用基于Transformer的表示实现去噪网络。为提高计算效率，我们引入了跨注意力头、空间令牌和MLP宽度的自适应路由。实验结果表明，该方法在257.57 GFLOPs下实现了0.731的AUC。
英文摘要
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.
