Evaluation of an Artificial Intelligence Diagnostic Model for Detecting Lymph Node Metastasis in Cancer Patients Undergoing Tumor Surgery

What this Trial Is About

Researchers are developing an artificial intelligence AI-based diagnostic model to detect pathological lymph node metastasis LNM across various cancers. The study aims to evaluate this AI models diagnostic performance compared to routine pathological reports by pathologists, with the goal of improving the clinical workflow in real-world settings. Accurate lymph node staging after cancer surgery is crucial for treatment decisions and prognosis, but current manual evaluations can miss small metastatic lesions and are time-consuming. This observational diagnostic study involves collecting pathological slides from patients undergoing radical tumor resection and lymph node dissection. These slides are digitized into whole-slide images WSIs, which the AI model analyzes to produce detailed heatmaps and slide-level diagnoses indicating the presence or absence of LNM. Routine pathological examinations proceed independently, and any discrepancies between AI and pathology results are reviewed by senior pathologists, using additional tests if needed, to decide the final diagnosis presented in the patients pathology report. Participants provide pathological samples as part of their standard surgery, with no additional interventions required. Researchers collect diagnostic results from the AI model over an average of three years to assess sensitivity and specificity. The study monitors how well the AI detects LNM compared to traditional pathology, aiming to enhance diagnostic accuracy and efficiency in cancer care.

Development and Prospective Validation of a Digital Pathology-based Artificial Intelligence Diagnostic Model for Pan-cancer Lymphatic Metastasis