通过可解释对抗分解学习解耦PET阿尔茨海默病病理异常并提升临床相关性
本研究为回顾性研究,基于多个队列的Aβ和tau PET数据开发可解释对抗分解学习框架,生成体素级病理图与ADAD评分;外部测试Aβ的AUC为0.94、tau为0.98,病理归因图与专家排序相关(ρ=0.79和0.63),ADAD评分与认知结局及海马萎缩存在独立关联。
要点
- ADL生成体素级病理归因图,与专家排序的Spearman ρ:Aβ为0.79、tau为0.63。
- 外部测试区分临床AD与认知正常者的AUC:Aβ为0.94(95%CI 0.89-0.98)、tau为0.98(95%CI 0.95-1.00)。
- ADAD评分在校正后与认知结局及海马萎缩存在独立基线与纵向关联。
核心结论
- 病理生理约束的对抗分解学习提供可解释、个体化病理图及AI衍生的ADAD评分,使PET病理异常与多模态临床指标更紧密关联。
研究设计与样本
- 回顾性研究,纳入阿尔茨海默病神经影像学倡议、澳大利亚影像生物标志物与生活方式研究、全球阿尔茨海默协会交互网络及作者中心数据;Aβ PET 7457次扫描来自3595例患者(中位年龄71.4岁,IQR 65.7-77.0岁,女性1637例),tau PET 1894次扫描来自1127例患者(中位年龄72.0岁,IQR 66.9-78.5岁,女性545例)。
诊断标准与关键数值
- 外部测试区分临床AD与认知正常者的AUC:Aβ为0.94(95%CI 0.89, 0.98),tau为0.98(95%CI 0.95, 1.00);ADL病理归因图与专家排序的Spearman ρ:Aβ为0.79、tau为0.63。
临床应用与阅片要点
- 原文未据此给出推荐。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
Background Template-based PET metrics quantify Alzheimer disease (AD) amyloid-β (Aβ) and tau burden but compress whole-brain data into a single scalar, overlooking disease heterogeneity and sometimes causing imaging-clinical discordance. Artificial intelligence (AI) approaches capture richer patterns but often lack biologic interpretability. Purpose To develop and validate an interpretable deep-learning framework that separates AD-specific abnormalities from physiologic uptake using pathophysiologic constraints, generating a clinically meaningful AI biomarker. Materials and Methods In this retrospective study, Aβ and tau PET scans from the Alzheimer's Disease Neuroimaging Initiative, Australian Imaging Biomarkers and Lifestyle study, Global Alzheimer's Association Interactive Network, and the authors' center were analyzed. An adversarial decomposition learning (ADL) network generated voxel-level pathologic maps and an AD adversarial decomposition (ADAD) score. Discriminatory performance for clinical AD versus cognitively normal individuals was evaluated using the area under the curve (AUC). Clinical relevance was assessed with cognitive, hippocampal volume, cerebrospinal fluid (CSF), and neuropathologic measures using longitudinal mixed-effects models and Spearman correlations. Results The study included 7457 Aβ PET scans from 3595 patients (median age, 71.4 years; IQR, 65.7-77.0 years; 1637 female patients) and 1894 tau PET scans from 1127 patients (median age, 72.0 years; IQR, 66.9-78.5 years; 545 female patients). External testing AUCs were 0.94 (95% CI: 0.89, 0.98) for Aβ and 0.98 (95% CI: 0.95, 1.00) for tau. ADL generated interpretable pathologic attribution maps that correlated with expert rankings (Aβ and tau, Spearman ρ = 0.79 and 0.63, respectively). Although Centiloid and CenTauRz showed numerically higher correlations with postmortem neuropathologic structure and stronger associations with CSF biomarkers, the ADAD score demonstrated independent baseline and longitudinal associations with cognitive outcomes and hippocampal atrophy after adjustment. Conclusion Pathophysiologic-constrained ADL provided interpretable, personalized pathologic maps and an AI-derived ADAD score that more closely linked PET pathologic abnormalities with multimodal clinical measures. © RSNA, 2026 Supplemental material is available for this article.
引用
- Tang C, Sun X, Tang A, et al. Decoupling Alzheimer Disease Pathologic Abnormalities at PET with Improved Clinical Relevance by Interpretable Adversarial Decomposition Learning. Radiology. 2026. PMID: 41944723. doi:10.1148/radiol.252321
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。