Table 2 Specification of machine-learning results.

Analysis	BAC (%)	Accuracy (%)	Sensitivity (%)	Specificity (%)	AUC	PPV (%)	NPV (%)	p value
ON_ROP	60.17	60.00	80.59	39.74	0.59	56.84	67.50	0.0001
ON_CHR	67.38	65.62	44.93	89.83	0.69	83.78	58.24	0.029
ON_ROP + CHR	55.89	55.51	44.85	66.92	0.62	59.22	53.13	0.221
ON_ROD	51.57	76.86	13.04	90.09	0.46	21.43	83.33	0.360
FPN_ROP	52.67	52.59	64.18	41.18	0.57	51.81	53.85	0.250
FPN_CHR	63.02	61.72	46.37	79.66	0.67	72.73	55.95	0.100
FPN_ROP + CHR	61.11	60.84	52.94	69.29	0.66	64.86	57.89	0.006
FPN_ROD	47.67	76.12	4.35	90.99	0.51	9.09	82.11	0.700
ON-FPN_ROP	56.46	56.29	79.11	33.82	0.59	54.08	62.16	0.222
ON-FPN_CHR	62.77	61.72	49.28	76.27	0.64	70.83	56.25	0.122
ON-FPN_ROP + CHR	59.69	59.32	48.53	70.86	0.64	64.07	56.25	0.016
ON-FPN_ROD	41.28	62.68	8.69	73.87	0.37	6.45	79.61	0.956

Significant p values are printed in bold type.
ON occipital network, FPN frontoparietal network, ON-FPN occipital and frontoparietal network combined, CHR clinical high risk for psychosis, ROP recent onset of psychosis, ROP + CHR combined group of ROP and CHR, ROD recent onset of depression, BAC balanced accuracy, AUC area under the curve, PPV positive predictive value, NPV negative predictive values.

Quick links

Search