Machine learning can reliably predict malignancy of breast lesions based on clinical and ultrasonographic features.

Abstract

Abstract Purpose To establish a reliable machine learning model to predict malignancy in breast lesions identified by ultrasound and optimize the negative predictive value to minimize unnecessary biopsies. Methods We included clinical and ultrasonographic attributes from 1526 breast lesions classified as BI-RADS 3, 4a, 4b, 4c, 5 and 6 that underwent ultrasound guided breast biopsy in four institutions. We selected the most informative attributes to train nine machine learning models, ensemble models and models with tuned threshold to make inferences about the diagnosis of BI-RADS 4a and 4b lesions (validation dataset). We tested the performance of the final model with 403 new suspicious lesions. Results The most informative attributes were shape, margin, orientation and size of the lesions, the resistance index of the internal vessel, the age of the patient and the presence of a palpable lump. The highest mean NPV was achieved with KNN (97.9%). Making ensembles didn´t improve the performance. Tuning the threshold did improve the performance of the models and we chose the XGBoost with the tuned threshold as the final one. The tested performance of the final model was: NPV 98.1%, FN 1.9%, VPP 77.1%, FP 22.9%. Applying this final model, we would have missed 2 of the 231 malignant lesions of the test dataset (0.8%). Conclusion Machine learning can help physicians predict malignancy in suspicious breast lesions identified by the US. Our final model would be able to avoid 60.4% of the biopsies in benign lesions missing less than 1% of the cancer cases.