Multi-Features-Based Automated Breast Tumor Diagnosis Using Ultrasound Image and Support Vector Machine

Table 4

The comparison of classification using different numbers of DLF and CF (CFs—characteristic features; DLFs—deep learning features).

Classification models	Accuracy	Sensitivity	Specificity	F1-score
2 DLF + CF + SVM	0.875	0.895	0.85	0.872
4 DLF + CF + SVM	0.925	0.905	0.95	0.927
6 DLF + CF + SVM	0.9	0.9	0.9	0.9
8 DLF + CF + SVM	0.85	0.889	0.8	0.84
10 DLF + CF + SVM	0.875	0.941	0.8	0.84
12 DLF + CF + SVM	0.85	0.889	0.8	0.84
16 DLF + CF + SVM	0.875	0.895	0.85	0.872