Assessment of the sentinel lymph node (SLN) in patients with early stage breast cancer is vital in selecting the appropriate surgical approach. But identification of metastatic LNs within the fibro-adipose tissue of the fossa axillaris specimen remains a challenge. Recently, indocyanine green (ICG) and methylene blue are commonly used in clinical practice. ICG as a fluorescent dyes, has effectiveness in mapping SLNs during surgery. Surgeons can follow the fluorescence display to detect SLN, and simultaneously capture real-time fluorescent video images. Several other groups has been demonstrated that the usage of ICG fluorescent surgical navigation system to detect SLNs in breast cancer patients is technically feasible. But no study investigate the variability between fluorescent images of breast sentinel lymph node with and without metastasis in the existing paper. Deep learning (DL) artificial intelligence (AI) algorithms in medical imaging are rapidly expanding.
In this study, the investigators aim to develop and validate an easy-to-use artificial intelligence prediction model to intraoperatively identify the sentinel lymph node metastasis status. Furthermore, to explore whether this independent and parallel intraoperative lymph node assessment workflow can provide rapid and accurate skull base on lymph node fluorescent images analysis, meanwhile detecting occult lymph node (micro-) metastasis, using optical imaging and artificial intelligence.
