Table 2 Tenfold LOOCV detection performance for a balanced and unbalanced dataset (mean ± standard deviation %).

From: STIED: a deep learning model for the spatiotemporal detection of focal interictal epileptiform discharges with MEG

Model\metric

Temporal

Spatial

Spatiotemporal

Accuracy

Specificity

Sensitivity

Accuracy

Specificity

Sensitivity

Accuracy

Specificity

Sensitivity

Balanced

90 ± 5

95 ± 6

84 ± 13

85 ± 9

96 ± 3

75 ± 19

89 ± 8

93 ± 4

85 ± 14

Unbalanced

92 ± 4

93 ± 5

84 ± 13

95 ± 3

96 ± 2

75 ± 19

95 ± 2

95 ± 2

85 ± 14

Back to article page