Table 4 Detailed accuracy metrics for the human annotator with respect to the ground truth (GT) in operative reports. All scores were computed by averaging the metrics across training and test cohorts
Outcome variable | Accuracy | Precision | Recall | F-Score |
|---|---|---|---|---|
Laterality | 0.99 | 0.99 | 1.00 | 1.00 |
Surgical Indication | 0.97 | 0.97 | 1.00 | 0.99 |
youPre-Operative Biopsy | 0.98 | 0.98 | 1.00 | 0.99 |
Pre-Operative Diagnosis | 0.93 | 0.93 | 1.00 | 0.97 |
Neoadjuvant Treatment | 0.98 | 0.98 | 1.00 | 0.99 |
Breast Procedure | 0.97 | 0.97 | 1.00 | 0.99 |
Immediate Reconstruction | 0.89 | 0.70 | 1.00 | 0.80 |
Immediate Reconstruction Type | 0.60 | 0.66 | 0.99 | 0.78 |
Wire Localization | 0.90 | 0.76 | 0.99 | 0.84 |
Breast Incision Type | 0.81 | 0.89 | 0.84 | 0.85 |
Axillary Surgery | 0.96 | 0.96 | 1.00 | 0.98 |
Overall | 0.91 | 0.89 | 0.98 | 0.92 |