Table 2. Comparison of AD and NC classification performance between the proposed MSRNet and other methods in the ADNI cohort.
| Method | ACC (%) | SEN (%) | SPE (%) | AUC (%) |
|---|---|---|---|---|
| AGNN | 88.1 | 75.0 | 94.1 | 91.5 |
| GAT | 86.5 | 71.1 | 93.7 | 92.4 |
| GraphSAGE | 88.9 | 78.3 | 93.8 | 94.1 |
| HGNN | 74.9 | 34.2 | 94.0 | 79.3 |
| Graph U-Net | 73.8 | 38.5 | 90.3 | 77.2 |
| ConvMixer | 86.5 | 76.6 | 91.1 | 92.3 |
| 3D VGG-16 | 90.6 | 81.8 | 94.7 | 95.1 |
| 3DAN | 88.9 | 79.1 | 93.1 | 94.4 |
| Proposed MSRNet | 92.8 | 88.2 | 95.0 | 95.6 |
AD, Alzheimer’s disease; NC, normal control; ACC, accuracy; SEN, sensitivity; SPE, specificity; AUC, area under the curve; MSRNet, multispatial information representation model; ADNI, Alzheimer’s Disease Neuroimaging Initiative database; AGNN, attention-based graph neural network; GAT, graph attention network; HGNN, hypergraph neural network; 3D VGG-16, 3D visual geometry group 16; 3DAN, 3D attention network.