Table 4. Classifier performance.
Model | Sensitivity[95% CI] | Specificity[95% CI] | Best CVA[95% CI] | Noise/Not Noise | ||
Noise | M1 | All Noise | 0.890.86 0.93 | 0.830.78 0.87 | 0.860.83 0.88 | 343 (295) |
M2 | Eyeballs | 0.560.31 0.80 | 1.01.0 1.0 | 0.980.96 0.99 | 16 (622) | |
M3 | Head Motion | 0.250.05 0.49 | 0.990.99 1.0 | 0.980.96 0.99 | 12 (626) | |
M4 | Ventricles | 0.20.05 0.34 | 0.990.99 1.0 | 0.960.94 0.97 | 30 (608) |
Performance metrics (sensitivity, specificity, best cross validation accuracy (CVA), area under the curve (AUC)), and proportion of noise components in data for comprehensive noise (All Noise, M1) and three noise subtypes (M2)(M3)(M4), built with Data A and tested with ten -fold cross validation on Data B (data from the same institution, same scanner, different subject population).