Department of radiology, AP-HP, Hôpital Saint-Antoine, Paris, France; Institut du Cerveau et de la Moelle épinière, ICM, Inserm, U 1127, CNRS, UMR 7225, Sorbonne Université, 75013 Paris, France; Inria, Aramis-project team, Paris, France.
Department of radiology, AP-HP, Hôpital Saint-Antoine, Paris, France.
J Neuroradiol. 2021 Nov;48(6):412-418. doi: 10.1016/j.neurad.2020.04.004. Epub 2020 May 12.
Many artificial intelligence tools are currently being developed to assist diagnosis of dementia from magnetic resonance imaging (MRI). However, these tools have so far been difficult to integrate in the clinical routine workflow. In this work, we propose a new simple way to use them and assess their utility for improving diagnostic accuracy.
We studied 34 patients with early-onset Alzheimer's disease (EOAD), 49 with late-onset AD (LOAD), 39 with frontotemporal dementia (FTD) and 24 with depression from the pre-existing cohort CLIN-AD. Support vector machine (SVM) automatic classifiers using 3D T1 MRI were trained to distinguish: LOAD vs. Depression, FTD vs. LOAD, EOAD vs. Depression, EOAD vs. FTD. We extracted SVM weight maps, which are tridimensional representations of discriminant atrophy patterns used by the classifier to take its decisions and we printed posters of these maps. Four radiologists (2 senior neuroradiologists and 2 unspecialized junior radiologists) performed a visual classification of the 4 diagnostic pairs using 3D T1 MRI. Classifications were performed twice: first with standard radiological reading and then using SVM weight maps as a guide.
Diagnostic performance was significantly improved by the use of the weight maps for the two junior radiologists in the case of FTD vs. EOAD. Improvement was over 10 points of diagnostic accuracy.
This tool can improve the diagnostic accuracy of junior radiologists and could be integrated in the clinical routine workflow.
目前正在开发许多人工智能工具来辅助磁共振成像（MRI）诊断痴呆症。然而，这些工具迄今难以整合到临床常规工作流程中。在这项工作中，我们提出了一种新的简单方法来使用它们，并评估它们在提高诊断准确性方面的效用。
我们研究了来自预存队列 CLIN-AD 的 34 名早发性阿尔茨海默病（EOAD）患者、49 名晚发性 AD（LOAD）患者、39 名额颞叶痴呆（FTD）患者和 24 名抑郁症患者。使用 3D T1 MRI 训练支持向量机（SVM）自动分类器，以区分：LOAD 与抑郁、FTD 与 LOAD、EOAD 与抑郁、EOAD 与 FTD。我们提取了 SVM 权重图，这是分类器用于做出决策的判别性萎缩模式的三维表示，并打印了这些地图的海报。4 名放射科医生（2 名高级神经放射科医生和 2 名非专业初级放射科医生）使用 3D T1 MRI 对 4 种诊断对进行了视觉分类。分类进行了两次：第一次是使用标准放射学阅读，然后使用 SVM 权重图作为指导。
对于两名初级放射科医生来说，使用权重图显著提高了 FTD 与 EOAD 的诊断性能。诊断准确性提高了 10 分以上。
该工具可以提高初级放射科医生的诊断准确性，并可以集成到临床常规工作流程中。
