Diagnostic performance of MRI radiomics for classification of Alzheimer's disease, mildcognitive impairment, and normal subjects: a systematic review and meta-analysis

Background: Alzheimer's disease (AD) is a debilitating neurodegenerative disease. Early diagnosis of AD andits precursor, mild cognitive impairment (MCI), is crucial for timely intervention and management. Radiomicsinvolves extracting quantitative features from medical images and analyzing them using advancedcomputational algorithms. These characteristics have the potential to serve as biomarkers for diseaseclassifcation, treatment response prediction, and patient stratifcation. Of note, Magnetic resonance imaging(MRI) radiomics showed a promising result for diagnosing and classifying AD, and MCI from normal subjects.Thus, we aimed to systematically evaluate the diagnostic performance of the MRI radiomics for this task.Methods and materials: A comprehensive search of the current literature was conducted using relevantkeywords in PubMed/MEDLINE, Embase, Scopus, and Web of Science databases from inception to August ۵,۲۰۲۳. Original studies discussing the diagnostic performance of MRI radiomics for the classifcation of AD,MCI, and normal subjects were included. Method quality was evaluated with the Quality Assessment ofDiagnostic Accuracy Studies (QUADAS-۲) and the Radiomics Quality Score (RQS) tools.Results: We identifed ۱۳ studies that met the inclusion criteria, involving a total of ۵۴۴۸ participants. Theoverall quality of the included studies was moderate to high. The pooled sensitivity and specifcity of MRIradiomics for diferentiating AD from normal subjects were ۰.۹۲ (۹۵% CI [۰.۸۵; ۰.۹۶]) and ۰.۹۱ (۹۵% CI [۰.۸۵;۰.۹۵]), respectively. The pooled sensitivity and specifcity of MRI radiomics for diferentiating MCI from normalsubjects were ۰.۷۴ (۹۵% CI [۰.۶۰; ۰.۸۵]) and ۰.۷۹ (۹۵% CI [۰.۷۰; ۰.۸۶]), respectively. Also, the pooledsensitivity and specifcity of MRI radiomics for diferentiating AD from MCI were ۰.۷۳ (۹۵% CI [۰.۶۴; ۰.۸۰])and ۰.۷۹ (۹۵% CI [۰.۶۴; ۰.۹۰]), respectively.Conclusion: MRI radiomics has promising diagnostic performance in diferentiating AD, MCI, and normalsubjects. It can potentially serve as a non-invasive and reliable tool for early diagnosis and classifcation of ADand MCI.