证据先于准确性：用于阿尔茨海默病分类的MRI-PET融合网络及因果区域验证
Evidence Before Accuracy: A MRI-PET Fusion Network for Alzheimer Disease Classification with Causal Regional Validation
查看机构详情
- Payame Noor University（帕亚梅努尔大学）
机构由 AI 辅助整理，请以论文原文为准。
浏览论文内容
中文总结 AI 辅助
提出MRI-PET融合网络用于AD分类，通过因果区域验证证明模型依赖海马（MRI）和后扣带回（PET）的神经生物学证据，而非数据集伪影。
中文摘要 AI 辅助
用于阿尔茨海默病（AD）分类的深度学习模型通常报告接近完美的判别能力，但很少有模型被证明依赖于AD相关的神经生物学，而非数据集伪影、受试者级泄漏或非脑部图像内容。我们提出了一种融合网络，结合T1 MRI和FDG PET，在轴向、冠状和矢状平面上进行训练，训练数据来自ADNI，包含554对受试者。该融合模型达到了AUC 0.962、准确率0.909和F1 0.891，与最近的3D CNN和多模态Transformer系统相比，在显著更低的成本下具有竞争力。我们首先量化了模态、平面和切片几何的重要性。仅基于验证集的搜索，在切片中心和邻域间距上进行，使MRI的AUC变化0.180，PET的AUC变化0.078，为MRI选择窄间距，为PET选择宽间距，所选冠状中心分别落在海马体和后扣带回上。然而，贡献在于围绕该数字构建的证据层。捷径控制将模型性能降至AUC 0.622（轮廓）、0.608（外部）和0.500（空白），标签置换零模型得到0.456。前向感兴趣区域（ROI）消融显示，在MRI中掩蔽内侧颞叶皮层和在PET中掩蔽后默认模式网络（DMN）产生最大的AD logit变化，而面积匹配的对照与零模型无法区分。反向ROI消融显示，仅内侧颞叶在MRI中保留了89.2%的超机会判别能力，仅后DMN在PET中保留了79.0%。对归因方法的定量比较显示，在预先定义的AD区域内，遮挡敏感性达到3.5-5.0倍的富集，而对照组为0.10-0.43倍。消融和归因独立地建立了生物学上正确的双重分离：海马证据由MRI承载，后扣带回证据由PET承载。
英文摘要
Deep learning models for Alzheimer disease (AD) classification routinely report near-perfect discrimination, yet few are shown to rest on AD-relevant neurobiology rather than on dataset artifacts, subject-level leakage, or non-brain image content. We present a fusion network combining T1 MRI and FDG PET across axial, coronal, and sagittal planes, trained on ADNI consists of 554 paired subjects. The fusion model reaches AUC 0.962, accuracy 0.909, and F1 0.891, competitive with recent 3D CNN and multimodal transformer systems at substantially lower cost. We first quantify how much modality, plane and slice geometry matter. A validation-only search over slice centres and neighbour spacings moves AUC by 0.180 for MRI and 0.078 for PET, selecting narrow spacing for MRI and wide spacing for PET, with the chosen coronal centres falling on the hippocampal body and on the posterior cingulate respectively. The contribution, however, is the evidence layer built around that number. Shortcut controls collapse the model to AUC 0.622 (silhouette), 0.608 (exterior), and 0.500 (blank), and a label-permutation null yields 0.456. Forward region-of-interest (ROI) ablation shows that masking medial temporal cortex in MRI and the posterior default-mode network (DMN) in PET produces the largest shift in the AD logit, while area-matched controls remain indistinguishable from that null. Reverse ROI ablation shows that the medial temporal lobe alone retains 89.2% of above-chance discrimination in MRI and the posterior DMN alone retains 79.0% in PET. A quantitative comparison of attribution methods shows occlusion sensitivity reaching 3.5-5.0* enrichment inside a priori AD regions against 0.10-0.43* in controls. Ablation and attribution independently establish a biologically correct double dissociation: hippocampal evidence is carried by MRI, posterior cingulate evidence by PET.
