Table 2.

Classification of 1 × 4096 feature vector obtained from CNN with support vector machine (SVM), K-nearest neighbors (KNN), and LSTM.

Method	Accuracy (%)	Sensitivity (%)	Specificity (%)	Youden’s Index
CNN + SVM	93.8 ± 0.6	93.6 ± 1.0	93.9 ± 1.5	0.87 ± 0.01
CNN + KNN	90.2 ± 0.6	90.7 ± 1.4	89.8 ± 2.0	0.80 ± 0.01
CNN + LSTM	95.4 ± 1.2	95.0 ± 0.8	95.7 ± 1.6	0.90 ± 0.02