Model statistics for brain imaging only, brain and non-imaging measure models and non-imaging measures only model
| Brain only full model | Brain and non-imaging measures full model | Non-imaging measures full model | |
| Accuracy | 0.70 | 0.83 | 0.66 |
| 95% CI | 58 to 81 | 0.72 to 0.90 | 0.54 to 0.77 |
| Sensitivity | 0.93 | 0.88 | 0.68 |
| Specificity | 0.40 | 0.77 | 0.20 |
| AUC | 0.79 | 0.87 | 0.70 |
| P value | 0.02* | <0.001** | 0.09 |
| Confusion matrices | Reference | Reference | Reference | ||||
| Yes | No | Yes | No | Yes | No | ||
| Prediction | Yes | 38 | 18 | 36 | 7 | 28 | 11 |
| No | 3 | 12 | 5 | 23 | 13 | 19 | |
All models contained drug ID as an additional term.
P value is expressed as the result of a one-tailed binomial test of model accuracy compared with the null information rate.
*p<0.05
AUC, area under the curve.