Advancements in the Utilization of Deep Learning Techniques based on Histopathological Imaging for Predicting the Efficacy of Neoadjuvant Chemotherapy in Breast Cancer

Neoadjuvant chemotherapy (NAC), alternatively termed preoperative or induction chemotherapy, involves the administration of systemic cytotoxic drugs before local treatment for non-distant metastatic tumors. Since its inception in the early 1970s, NAC has become the standard treatment protocol for locally advanced breast cancer, primarily due to its ability to reduce tumor burden. Currently, NAC is increasingly employed in the treatment of early-stage breast cancer to enhance breast conservation rates. While the majority of breast cancer patients benefit from NAC, a minority experience no response during treatment or even disease progression, thus losing the opportunity for durable local control of the disease. The early identification of these non-responders facilitates personalized treatment strategies and avoids delays in therapy. Consequently, the quest for biomarkers capable of predicting early responses to NAC has become a focal point of research. With the advent of deep learning technologies in the medical field, AI-based diagnostics and prognostic predictions using imaging and pathology have evolved. This article provides an overview of the latest research on the prediction of neoadjuvant chemotherapy efficacy in breast cancer.