Research Article
[Retracted] Deep Learning Mechanism for Predicting the Axillary Lymph Node Metastasis in Patients with Primary Breast Cancer
Table 2
Evaluation of radiology’s efficacy in comparison to that of three ANN models.
| | Test finding | Feed forward | Radial basis function | Kohonen self-organizing | Radiology | P measure |
| | Test dataset A () | | | Precision | 92 (88/110) | 94 (90/110) | 98 (86/98) | 88 (84/98) | 0.98 | | Sensitivity | 92 (50/60) | 90 (50/58) | 98 (48/60) | 83 (45/59) | 0.30 | | Specificity | 89 (49/59) | 95 (52/59) | 99 (48/59) | 89 (49/68) | 0.36 | | Measure positive prediction | 90 (50/60) | 95 (49/57) | 99 (48/59) | 87 (51/65) | 0.34 | | Measure negative prediction | 91 (48/58) | 90 (52/61) | 98 (48/59) | 85 (53/69) | 0.29 | | K-means | 0.72 | 0.76 | 0.68 | 0.52 | — | | F-score | 0.92 | 0.93 | 0.98 | 0.85 | — | | Test dataset B () | | | Precision | 90 (74/90) | 88 (73/83) | 93 (68/90) | 88 (65/91) | 0.98 | | Sensitivity | 95 (46/52) | 89 (43/52) | 93 (41/52) | 84 (45/65) | 0.62 | | Specificity | 83 (40/52) | 85 (40/50) | 93 (39/52) | 85 (35/50) | 0.73 | | Measure positive prediction | 86 (45/57) | 86 (42/53) | 93 (40/51) | 87 (40/55) | 0.85 | | Measure negative prediction | 91 (40/45) | 87 (40/48) | 93 (39/42) | 89 (45/60) | 0.78 | | K-means | 0.64 | 0.68 | 0.57 | 0.56 | — | | F-score | 0.92 | 0.88 | 0.94 | 0.80 | — |
|
|
