BI-RADS评分是放射科医生用于表达基于乳腺X线影像形态特征预测乳腺癌不确定性的概率报告工具。在描述肿块时存在显著的变异性，这有时会导致BI-RADS误分类。需要一个BI-RADS预测系统来支持放射科医生的最终决策。本研究利用贝叶斯深度学习模型提取的不确定性信息来预测BI-RADS评分。基于病理信息的研究结果表明，放射科医生预测的f1分数分别为42.86%、48.33%和48.28%，而模型在BI-RADS 2、3和5数据集样本中的表现f1分数分别为73.33%、59.60%和59.26%。此外，该模型能够以75.86%的准确率区分所用数据集中BI-RADS 0类别中的恶性与良性样本，并正确识别所有恶性样本为BI-RADS 5。Grad-CAM可视化显示，模型关注病变的形态特征。因此，本研究表明，具有不确定性感知的贝叶斯深度学习模型可以根据形态特征像放射科医生一样报告其对病灶恶性的不确定性。

The BI_RADS score is a probabilistic reporting tool used by radiologists to express the level of uncertainty in predicting breast cancer based on some morphological features in mammography images. There is a significant variability in describing masses which sometimes leads to BI_RADS misclassification. Using a BI_RADS prediction system is required to support the final radiologist decisions. In this study, the uncertainty information extracted by a Bayesian deep learning model is utilized to predict the BI_RADS score. The investigation results based on the pathology information demonstrate that the f1-scores of the predictions of the radiologist are 42.86%, 48.33% and 48.28%, meanwhile, the f1-scores of the model performance are 73.33%, 59.60% and 59.26% in the BI_RADS 2, 3 and 5 dataset samples, respectively. Also, the model can distinguish malignant from benign samples in the BI_RADS 0 category of the used dataset with an accuracy of 75.86% and correctly identify all malignant samples as BI_RADS 5. The Grad-CAM visualization shows the model pays attention to the morphological features of the lesions. Therefore, this study shows the uncertainty-aware Bayesian Deep Learning model can report his uncertainty about the malignancy of a lesion based on morphological features, like a radiologist.