Table 8.
The area under the curve (AUC) for the training and test datasets for each brain modality from the logistic elastic net. No covariates were added to the model to quantify the probability of the AUC by the brain alone. The AUC and p values as well as the number of children in the analysis and features included are displayed per modality. WG = weight gain group; MID = Monetary Incentive Delay Task; SST = Stop Signal Task; rsfMRI = resting state functional magnetic resonance imaging.
| Training dataset |
Testing dataset | ||||
|---|---|---|---|---|---|
| Modality | AUC (p) | nTotal (nWG) | AUC (p) | nTotal (nWG) | # of features |
| Structure | 0.71(0.0004) | 652 (136) | 0.61 (0.02) | 157 (36) | 209 |
| rsfMRI | 0.06 (0.06) | 333 (70) | 0.45 (0.25) | 87 (13) | 7 |
| MID | 0.5 (0.5) | 430 (85) | 0.5 (0.5) | 115 (26) | 2 |
| SST | 0.5 (0.5) | 413 (80) | 0.5 (0.5) | 115 (27) | 2 |
| EN-back | 0.5 (0.5) | 410 (81) | 0.5 (0.5) | 111 (24) | -- |