Table 2.
Index of Overlap between the two scanning sessions: mean per subject + standard deviation in the brackets. Naturally, methods giving the sparsest connectomes also result in the highest Index of Overlap. The group difference in Index of Overlap between task and rest is insignificant for all the methods.
| no | method | Rest | Task | Difference |
|---|---|---|---|---|
| 1 | Empirical precision | 60.15 (±2.25) | 66.09 (±2.33) | +5.94 |
| 2 | Ledoit-Wolf | 61.53 (±2.24) | 67.75 (±2.45) | +6.22 |
| 3 | Permutation testing | 93.79 (±1.08) | 94.15 (±1.66) | +0.36 |
| 5 | MM(LW,GG) | 96.03 (±0.89) | 97.53 (±4.87) | +0.38 |
| 6 | Proportional thresholding, top 5% | 97.66 (±0.54) | 97.53 (±0.65) | −0.13 |
| 7 | Proportional thresholding, top 10% | 93.47 (±0.84) | 93.21 (±0.97) | −0.26 |