Alzheimer's disease (AD) is a progressive neurodegenerative disorder that imposes a substantial burden on families and healthcare systems. Mild cognitive impairment (MCI), as an intermediate stage between normal aging and AD, can be further divided into progressive MCI (pMCI) and stable MCI (sMCI) based on follow-up outcomes. Unlike the marked differences observed between cognitively normal (CN) individuals and AD patients, sMCI and pMCI share highly similar characteristics, making early identification of pMCI extremely challenging. Although deep learning methods based on structural magnetic resonance imaging (sMRI) have advanced AD classification, research on predicting MCI progression remains limited due to the high similarity between sMCI and pMCI as well as the substantial cost of prospectively collecting longitudinal data. Accurate early identification of pMCI is essential for timely intervention, slowing disease progression, and reducing healthcare costs. Therefore, this study focused on the early identification of progressive MCI. To address this, we proposed a novel vision Transformer framework, the multi-scale multi-domain frequency-aware vision Transformer (MMF-ViT), which employs a multi-scale cross-domain fusion (MSCDF) module to enable deep interaction between spatial and frequency domain features, thereby enhancing the modeling of fine-grained brain structural variations. The multi-scale frequency encoder (MSFE) and multi-scale context encoder (MSCE) were designed to extract and fuse frequency and spatial information, effectively improving classification performance. Experimental results on the ADNI dataset demonstrate that MMF-ViT achieves an accuracy of 72.84% and an AUC of 72.99% for sMCI versus pMCI classification, significantly outperforming mainstream 2D and 3D models. In AD vs. CN classification, MMF-ViT also achieves an accuracy of 85.59%, highlighting its strong feature representation capability and practical potential.

As is well known, the Apolipoprotein E (APOE) ε4 allele is the most pertinent genetic hazardous element for Alzheimer's disease (AD). Mild cognitive impairment (MCI) is considered a prodromal stage of AD. How the APOE ε4 allele modulates functional connectivity of brain network in MCI group is a question worth exploring. At present, some studies have evaluated the relationship between APOE ε4 allele and static functional network connectivity (sFNC) for MCI individuals, while the relationship of dynamic FNC (dFNC) with APOE ε4 allele still remained puzzled. Thus, we aim to detect aberrant dFNC for APOE ε4 carriers in the MCI group. On the basis of the resting-state functional magnetic resonance imaging (rs-fMRI) data, seven intrinsic brain functional networks were first recognized by the group independent component analysis. Then, the technique of sliding window was employed to determine the dFNC, and two dFNC states were detected by the k-means clustering algorithm. Finally, three temporal properties of fraction time, mean dwell time as well as transition numbers in the dFNC states were investigated. The results found that the dFNC and temporal properties in APOE ε4 carriers were abnormal compared with those in APOE ε4 noncarriers. In detail, in the MCI group, compared with APOE ε4 noncarriers, carriers had 9 pairs of abnormal dFNC and had significant differences in all the three temporal properties of the two dFNC states. In addition, two pairs of dFNC were found significantly correlated with clinical measure. This detected abnormal dynamics of temporal properties and dFNC in APOE ε4 carriers were similar with that reported for AD patients in previous studies. These results may suggest that in the MCI group, APOE carriers are more at risk for AD compared to noncarriers. Our findings may offer novel insights into the mechanisms of abnormal brain reconfiguration for individuals at genetic risk for AD, which could also be regarded as biomarkers for the early identification of AD.