Table 5.
The LOOCV results using top 10 MRI features and top 10 network features.
| Diagnostic pair | Top 10 MRI features | Top 10 network features | ||||||
|---|---|---|---|---|---|---|---|---|
| ACC (%) | SEN (%) | SPE (%) | AUC | ACC (%) | SEN (%) | SPE (%) | AUC | |
| MCInc vs. MCIc_mixed | 72.33 | 68.42 | 75.90 | 0.7865 | 64.78 | 61.84 | 67.47 | 0.6974 |
| MCInc vs. MCIc_m6 | 75.00 | 63.93 | 83.13 | 0.8002 | 61.81 | 49.18 | 71.08 | 0.6006 |
| MCInc vs. MCIc_m12 | 73.29 | 63.49 | 80.72 | 0.7885 | 70.55 | 61.90 | 77.11 | 0.7481 |
| MCInc vs. MCIc_m18 | 78.40 | 45.24 | 95.18 | 0.7321 | 66.40 | 33.33 | 83.13 | 0.6410 |
ACC, accuracy; SEN, sensitivity; SPE, specificity; AUC, area under the curve.