This is a retrospective observational study comparing the accuracy and safety of robotic-assisted versus endoscopic-assisted sentinel lymph node biopsy (SLNB) in patients with breast cancer. The study aims to evaluate the detection accuracy, false-negative rate, and perioperative safety profile between the two minimally invasive techniques, as well as to assess postoperative complications, pain, upper limb function, quality of life, and long-term oncological outcomes including axillary recurrence and survival.

Data will be extracted from routine clinical records. The study will compare the robotic group and the endoscopic group with respect to the following outcome measures: 1.Accuracy-related outcomes: sentinel lymph node detection rate, number of sentinel lymph nodes identified, and false-negative rate, sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV). All accuracy metrics are calculated with reference to final axillary lymph node status confirmed by completion ALND or long-term clinical follow-up; 2.Safety-related outcomes: perioperative parameters including operative time, estimated blood loss, length of hospital stay, and duration of postoperative drainage; postoperative complications including but not limited to seroma, hematoma, wound infection, and lymphedema; postoperative pain assessed by the Verbal Rating Scale (VRS); upper limb function assessed by the Disabilities of the Arm, Shoulder and Hand (DASH) score; and quality of life assessed by the Functional Assessment of Cancer Therapy - Breast + 4 (FACT B+4) questionnaire; 3.Oncological outcomes: including but not limited to axillary recurrence, locoregional recurrence, distant metastasis, disease-free survival (DFS), and overall survival (OS), with a minimum follow-up period to be determined based on data availability. This study has been approved by the Research Ethics Committee of Guangzhou Medical University Affiliated Women and Children's Medical Center. A waiver of informed consent has been granted due to the retrospective nature of the study. All patient data will be anonymized and kept confidential. Statistical analyses will be performed using R, SPSS and GraphPad Prism, with a P value < 0.05 considered statistically significant.