Table 2.

Distribution of epoch-wise accuracy quantifiers for predictions.

		Favorable	Unfavorable	No surgery	p -value
		median [q1, q3]	median [q1, q3]	median [q1, q3]	fav. vs. unfav.
Events to SOZ	Precision	0.24 [0.13, 0.35]	0.21 [0.13, 0.41]	0.25 [0.075, 0.41]	0.689
	Recall	0.25 [0.13, 0.46]	0.19 [0.12, 0.38]	0.25 [0.1, 0.38]	0.296
	F1-score	0.24 [0.13, 0.38]	0.19 [0.12, 0.38]	0.22 [0.087, 0.41]	0.523
Events to RBT	Precision	0.47 [0.28, 0.62]	0.29 [0.14, 0.45]	n/a	0.019
	Recall	0.17 [0.098, 0.28]	0.13 [0.038, 0.18]	n/a	0.018
	F1-score	0.23 [0.15, 0.37]	0.19 [0.06, 0.27]	n/a	0.022
Events to Core	Precision	0.15 [0.064, 0.28]	0.046 [0.012, 0.25]	0.3 [0.12, 0.39]	0.034
	Recall	0.35 [0.22, 0.51]	0.13 [0.026, 0.48]	0.34 [0.25, 0.38]	0.013
	F1-score	0.24 [0.088, 0.35]	0.068 [0.016, 0.28]	0.27 [0.13, 0.35]	0.014
SOZ to RBT	Precision	1 [0.62, 1]	0.2 [0, 0.85]	n/a	<10−3
	Recall	0.29 [0.22, 0.45]	0.062 [0, 0.31]	n/a	0.008
	F1-score	0.44 [0.36, 0.53]	0.095 [0, 0.43]	n/a	0.004
Core to SOZ	Precision	1 [0, 1]	0 [0, 0.05]	0.5 [0, 1]	0.002
	Recall	0.25 [0, 0.5]	0 [0, 0.062]	0.29 [0, 0.8]	0.002
	F1-score	0.4 [0, 0.67]	0 [0, 0.071]	0.34 [0, 0.5]	0.001
Core to RBT	Precision	1 [1, 1]	0 [0, 0.7]	n/a	<10−5
	Recall	0.18 [0.077, 0.25]	0 [0, 0.11]	n/a	<10−4
	F1-score	0.3 [0.14, 0.4]	0 [0, 0.15]	n/a	<10−4