Table 11 -.

Comparison between intra-class correlation (ICC), true positive rate (TPR), and positive prediction value (PPV) values of different classifiers for segmentation of WMHs in different datasets using T1 data for ADC, NACC, ADNI1, and ADNI2/GO datasets. Blue color indicates the best performance in terms of SI.

Dataset	ADC			NACC			ADNI1			ADNI2/GO
Measure	ICC	TPR	PPV	ICC	TPR	PPV	ICC	TPR	PPV	ICC	TPR	PPV
Naïve Bayes	0.24	0.20	0.67	0.01	0.28	0.56	0.30	0.37	0.74	0.06	0.33	0.74
Logistic Regression	0.08	0.27	0.09	0.00	0.20	0.07	0.22	0.61	0.34	0.17	0.65	0.25
LDA	0.32	0.22	0.64	0.07	0.29	0.54	0.45	0.42	0.68	0.19	0.39	0.68
QDA	0.40	0.15	0.67	0.52	0.25	0.56	0.38	0.40	0.74	0.17	0.37	0.73
KNN	0.36	0.49	0.22	0.14	0.61	0.25	0.55	0.63	0.50	0.54	0.61	0.43
Decision Trees	0.55	0.24	0.37	0.36	0.30	0.38	0.63	0.41	0.54	0.62	0.40	0.51
Random Forests	0.56	0.45	0.31	0.54	0.55	0.36	0.60	0.59	0.56	0.65	0.57	0.50
Bagging	0.01	0.59	0.02	0.23	0.58	0.05	0.23	0.73	0.21	0.16	0.71	0.15
SVM	0.03	0.49	0.13	0.10	0.54	0.22	0.20	0.59	0.41	0.14	0.56	0.43
AdaBoost	0.25	0.49	0.21	0.10	0.61	0.30	0.52	0.60	0.56	0.51	0.59	0.51