Predicting Long-Term Clinical Outcomes in Chinese Breast Cancer Patients Receiving Neoadjuvant Chemotherapy

At present, the majority of studies on neoadjuvant chemotherapy (NAC) in patients with breast cancer (BC) use pathological complete response (pCR) as a surrogate marker for patient prognosis, with significant improvements in pCR indicating better long-term survival. However, there is still a lack of non-invasive tools for accurately predicting the prognosis and pCR of BC patients undergoing NAC. Recent research has introduced emerging artificial intelligence machine learning (ML) and deep learning (DL) algorithms such as Bayesian methods, K-nearest neighbors (KNN), decision trees, support vector machines (SVM), XGBoost, ResNet, convolutional neural networks, and Transformer models, which have brought new avenues of exploration for cancer researchers.

The integration of AI with imaging, pathology, genomics, and other multi-omics has non-invasively improved preoperative diagnosis of breast cancer and, when combined with clinical factors, can assess postoperative survival. Moreover, current research data is limited, and reliable predictive models require extensive data for training. Therefore, establishing a multi-center database is essential.