Transferable structural, static and dynamic functional MRI features for predicting conversion from mild cognitive impairment to Alzheimer’s disease in Shanghai Memory Study and ADNI cohorts

Abstract Background Mild cognitive impairment (MCI) is a critical prodromal stage of Alzheimer’s disease (AD) with high risk of conversion. It is of great challenge to construct reliable biomarkers for predicting conversion from MCI to AD, while the underlying mechanism is still not fully explored. Inter‐dataset generalizability is a prerequisite for clinical use of biomarkers and always a shortage of neuroimaging‐based studies. Method In this study, we propose a novel framework by integrating structural MRI (sMRI) and both static and dynamic resting‐state functional MRI (fMRI) measurements to investigate the differences between MCI converters (MCI_C) and non‐converters (MCI_NC), and then utilized support vector machine (SVM) to construct the prediction models based on selected features. A total of 186 MCI patients with both MRI and three‐year outcome data were selected from two independent cohorts: Shanghai Memory Study (SMS) cohort for selection of MRI predictors and internal cross‐validation, and ADNI cohort for external validation on the generalizability of these MRI predictors. Result In comparison with MCI_NC, the MRI converters were mainly characterized by alterations of medial temporal lobe (MTL) with atrophy extending to lateral temporal and regional hyperactivity and instability, posterior parietal cortex (PPC) with atrophy and inter‐regional hypo‐connectivity and connectional instability, and occipital cortex with functional instability. All of the imaging‐based prediction models achieved an AUC above 0.7 and ACC above 70% in both SMS and ADNI cohorts. The combination of static and dynamic fMRI features resulted in overall good performance as relative to static or dynamic fMRI solely, supporting the contribution of dynamic features into the fMRI model. In both cohorts, the best imaging model was the multi‐modality MRI model which provided excellent performance with AUC above 0.85 and average ACC/sensitivity/specificity around 80%. Conclusion This inter‐cohort validation study provides a new insight into the mechanisms of MCI conversion and paves a way for eventual clinical use of MRI biomarkers.