MICA：通过多层级图像-概念对齐实现可解释的皮肤病变诊断
MICA: Towards Explainable Skin Lesion Diagnosis via Multi-Level Image-Concept Alignment
- Hong Kong University of Science and Technology（香港科技大学）
- HKUST Shenzhen-Hong Kong Collaborative Innovation Research Institute（港科大深港协同创新研究院）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
针对现有概念方法仅从单一视角对齐医学图像与概念的不足，本文提出多层级图像-概念对齐框架MICA，在图像、令牌和概念层面进行语义对齐，并支持干预及文本与视觉解释，在三个皮肤数据集上兼顾了可解释性、诊断性能和标签效率。
AI中文摘要：
黑盒深度学习方法在医学图像分析领域已展现出巨大潜力。然而，医学领域固有的严格可信性要求推动了可解释人工智能（XAI）的研究，其中基于概念的方法尤其受到关注。现有的基于概念的方法主要从单一视角（例如全局层面）应用概念标注，忽视了医学图像中子区域与概念之间细腻的语义关系。这导致宝贵的医学信息未被充分利用，并可能使模型在采用概念瓶颈等固有可解释架构时，难以在可解释性与性能之间取得良好平衡。为缓解这些不足，我们提出了一种多模态可解释疾病诊断框架，该框架在图像层面、令牌层面和概念层面等多个层级上，对医学图像与临床相关概念进行精细的语义对齐。此外，我们的方法支持模型干预，并以人类可解释概念的形式提供文本和视觉解释。在三个皮肤图像数据集上的实验结果表明，我们的方法在保持模型可解释性的同时，在概念检测和疾病诊断方面取得了高性能和标签效率。
英文摘要：
Black-box deep learning approaches have showcased significant potential in the realm of medical image analysis. However, the stringent trustworthiness requirements intrinsic to the medical field have catalyzed research into the utilization of Explainable Artificial Intelligence (XAI), with a particular focus on concept-based methods. Existing concept-based methods predominantly apply concept annotations from a single perspective (e.g., global level), neglecting the nuanced semantic relationships between sub-regions and concepts embedded within medical images. This leads to underutilization of the valuable medical information and may cause models to fall short in harmoniously balancing interpretability and performance when employing inherently interpretable architectures such as Concept Bottlenecks. To mitigate these shortcomings, we propose a multi-modal explainable disease diagnosis framework that meticulously aligns medical images and clinical-related concepts semantically at multiple strata, encompassing the image level, token level, and concept level. Moreover, our method allows for model intervention and offers both textual and visual explanations in terms of human-interpretable concepts. Experimental results on three skin image datasets demonstrate that our method, while preserving model interpretability, attains high performance and label efficiency for concept detection and disease diagnosis.
