Permutation Entropy and Signal Energy Increase the Accuracy of Neuropathic Change Detection in Needle EMG

Table 2

Comparison of the SVM classifier with different parameters for performance measures (all the SVM classifiers were trained with a set of default parameters; see SVM under Data Processing). A leave-one-out cross-validation was performed for each result.

Performance measure	Turns, amplitude	Turns, amplitude, entropy	Turns, amplitude, energy	Turns, amplitude, entropy, energy
TP	35.13	33.18	33.63	36.10
FN	3.88	5.83	5.38	2.90
FP	9.88	7.68	7.23	4.95
TN	29.13	31.33	31.78	34.05
Sn+ (%)	0.90	0.85	0.86	0.93
Sp+ (%)	0.78	0.81	0.82	0.88
Sn− (%)	0.75	0.80	0.81	0.87
Sp− (%)	0.88	0.84	0.86	0.92
Ac (%)	0.82	0.83	0.84	0.90
MCC	0.66	0.65	0.68	0.80