Computer-aided prediction model for axillary lymph node metastasis in breast cancer using tumor morphological and textural features on ultrasound

•A computer-aided prediction (CAP) system contained automatic tumor segmentation and feature quantification on breast ultrasound was developed to evaluate the axillary lymph node (ALN) status in the preoperative staging with breast cancer.•The proposed CAP model based on morphological and textural features of primary tumor would provide a promising diagnostic suggestion for recognizing the breast cancers as a case of ALN metastasis.•In view of there has been a decrease in the patients presented with palpable ALN at diagnosis of breast cancer, a non-invasive examination like our CAP model may help physicians in the choice of the optimal axillary approach.