Table 3.

Results of the classification between single AD and Nold subjects based on composite EEG markers of lagged linear connectivity.

Linear Lagged Connectivity (LLC)	Sensitivity (%)	Specificity (%)	Accuracy (%)	AUROC
Intra-hemispheric left parietal limbic	70	74	71.8	0.74
Intra-hemispheric left occipital limbic	73.3	64	69.1	0.71
Intra-hemispheric right temporal limbic	79.2	55	68.2	0.71
Intra-hemispheric right central occipital	80.8	54	68.6	0.70
Intra-hemispheric right parietal occipital	71.7	72	71.8	0.73
Intra-hemispheric right parietal temporal	67.5	71	69.1	0.71
Intra-hemispheric right parietal limbic	68.3	71	69.5	0.72
Intra-hemispheric right occipital temporal	70.8	72	71.3	0.74
Intra-hemispheric right occipital limbic	66.7	76	70.9	0.73
Intra-hemispheric right temporal limbic	76.7	64	70.9	0.73

Specifically, those composite EEG markers were obtained by computing the ratio between the theta lagged linear connectivity and the alpha 1 linear lagged connectivity, exept the intra-hemispheric right temporal limbic linear lagget connectivity, computed with the ratio between the delta lagged linear connectivity and the alpha 1 linear lagged connectivity. The classification rate is computed by the analysis of area under the receiver operating characteristic curve (AUROC). The table reports the classification indexes for the composite EEG markers having an AUROC higher than 0.70 (i.e., 70%).