Alzheimer's disease resemblance atrophy index in the identification of Alzheimer's disease with mild dementia. (1st February 2022)

• Record Type: Journal Article
• Authors: Fan, Xiang; Zhao, Lei; Liu, Wanting; Cai, Yuan; Luo, Yishan; Shi, Lin; Mok, Vincent C.T.
• Is Part Of: Alzheimer's & dementia. Volume 17(2021)Supplement 4
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-02-01
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.052705
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: Alzheimer's disease (AD) resemblance atrophy index (AD‐RAI) is an MRI‐based biomarker that was developed using a machine learning technology that combines multi‐region MRI features to reflect the characteristic atrophy pattern of AD. It was validated to be accurate in predicting conversion from mild cognitive impairment to dementia (1) and in identifying AD patients with moderate dementia. We aimed to compare the performance of AD‐RAI with various MRI‐based quantitative medial temporal lobe atrophy (MTA) measures in identifying AD with mild dementia. Method: We reviewed 3489 subjects in 3 datasets, i.e., ADNI (n=1529), AIBL (n=862), and OASIS 3 (n=1098). We selected subjects aged 65 to 85 years with clinical dementia rating scale of 1. Among potential eligible AD subjects with mild dementia who also had amyloid data based on either cerebrospinal fluid analysis or positron emission tomography (n=73), we further excluded those who were amyloid negative (n=5). A total of 121 AD subjects with mild dementia and 691 age‐matched elderly normal controls (NC) were eventually recruited. An automated computational tool, AccuBrain, was used to generate the following imaging biomarkers: hippocampal volume (HV), hippocampal fraction (HF), quantitative MTA (QMTA) and AD‐RAI. We compared the following performance metrics in identifying AD with mild dementia for each of the imaging biomarkers: area under curve (AUC) of receiver operating characteristic (ROC) curve, sensitivity, specificity and balanced accuracy. Result: Regarding the ROC analysis, the AD‐RAI performed the best in identifying mild AD from NC (AUC=0.92, sensitivity=0.88 and balanced accuracy=0.83), followed by QMTA and HV/ HF (Table 1 and Figure 1). Conclusion: AD‐RAI outperformed other MRI‐based quantitative MTA measures in identifying AD with mild dementia. Since majority of AD patients when they first presented to memory clinic are at the stage of mild dementia, AD‐RAI may aid accurate diagnosis of AD in clinical practice. Reference: 1) Zhao L, Luo Y, Lew D, Liu W, Au L, Mok V, et al. Risk estimation before progression to mild cognitive impairment and Alzheimer's disease: an AD resemblance atrophy index. Aging (Albany NY). 2019;11(16):6217‐36.