Radial Basis Function Artificial Neural Network for the Investigation of Thyroid Cytological Lesions

Table 7

Performance indices for the patient classification subsystems (numeric and percentages classifiers) for the training and test sets and for both sets combined.

	Numeric classifier			Percentages classifier
	Training set	Test set	Both sets	Training set	Test set	Both sets
Sensitivity (%)	92.41	91.25	91.82	94.94	95.00	94.97
Specificity (%)	94.48	93.71	94.10	95.86	95.10	95.49
PPV (%)	90.12	89.02	89.57	92.59	91.57	92.07
NPV (%)	95.80	95.04	95.42	97.20	97.14	97.17
FPR (%)	5.52	6.29	5.90	4.14	4.90	4.51
FNR (%)	7.59	8.75	8.18	5.06	5.00	5.03
OA (%)	93.75	92.83	93.29	95.54	95.07	95.30
PLR	16.75	14.50	15.56	22.94	19.41	21.04
NLR	0.08	0.09	0.09	0.05	0.05	0.05
Odds ratio	208.35	155.27	179.03	434.38	369.14	399.28