Table 1.

Evaluation of particular metrics in proposed thresholding methods.

Metric / thresholding method	Minimal distance criterion	Minimal product criterion	Elbow criterion
PRE group
Mask hit	0.73 ± 0.46 (1.00)	1.00 ± 0.00 (1.00)	0.68 ± 0.48 (1.00)
Percent of overlay *	5.00 ± 4.49 (3.96)	15.13 ± 9.74 (13.06)	3.35 ± 3.38 (1.96)
Precision in EZ side	0.68 ± 0.48 (1.00)	0.73 ± 0.46 (1.00)	0.68 ± 0.48 (1.00)
TP	1.32 ± 1.04 (1.00)	1.82 ± 0.96 (2.00)	1.00 ± 0.87 (1.00)
FP	5.73 ± 1.64 (6.00)	6.64 ± 2.15 (6.00)	4.91 ± 1.74 (5.00)
PPV	0.18 ± 0.14 (0.17)	0.22 ± 0.11 (0.19)	0.16 ± 0.14 (0.17)
POST group
Mask hit	0.57 ± 0.53 (1.00)	0.71 ± 0.49 (1.00)	0.43 ± 0.53 (0.00)
Percent of overlay *	17.82 ± 20.03 (12.19)	31.98 ± 22.82 (35.74)	14.38 ± 18.31 (9.72)
Precision in EZ side	0.86 ± 0.38 (1.00)	0.86 ± 0.38 (1.00)	0.86 ± 0.38 (1.00)
TP	0.57 ± 0.53 (1.00)	0.71 ± 0.49 (1.00)	0.43 ± 0.53 (0.00)
FP	6.57 ± 1.62 (7.00)	6.29 ± 1.89 (6.00)	4.86 ± 2.04 (4.00)
PPV	0.07 ± 0.07 (0.11)	0.12 ± 0.08 (0.14)	0.07 ± 0.09 (0.00)

The values are mean ± standard deviation (median). The optimal values are marked in bold. In TP itself, there might not be a clear optimal value in all cases. Therefore we did not mark any TP value in bold. Used abbreviations: TP—true positive, FP—false positive, PPV—positive predictive value, EZ—epileptogenic zone, POST—postoperative group, PRE—preoperative group, *percent of overlay of the resection mask/ EZ hypothesis mask by the most precise cluster.