Using Endoscopic Ultrasound-Guided Fine-Needle Aspiration and Rapid Cytology Staining with Whole Slide Scanning and AI to Diagnose Solid Pancreatic Lesions A Prospective Multicenter Study

What this Trial Is About

Researchers are investigating whether a self-developed whole slide scanning and artificial intelligence diagnostic system called the Zhiying Shunxi ROSE-AI can quickly and accurately diagnose solid pancreatic lesions SPLs. The study aims to find out if this system, using optical imaging technology on stained pancreatic puncture smears, can help distinguish between different types of solid pancreatic diseases such as pancreatic ductal adenocarcinoma, pancreatic neuroendocrine tumors, and benign lesions. The diagnoses made by the ROSE-AI system will be compared with the confirmed pathological diagnoses to evaluate its accuracy. The study involves obtaining pancreatic puncture samples stained with Diff-Quik as part of routine clinical care. These samples are then scanned using the ROSE-AI system at a specialized center to capture RGB images. An artificial intelligence algorithm uses these images to assist in differentiating the types of solid pancreatic lesions. After imaging, the specimens are returned to the original centers. This observational study does not involve treatment interventions but focuses on diagnostic assessment. Participants will have undergone imaging studies such as ultrasound, CT, or MRI to identify or suspect solid pancreatic lesions. Their stained smears will be scanned and analyzed by the AI system. Researchers will measure the diagnostic accuracy of the system over an average follow-up period of two years. The study requires informed consent and cooperation throughout, with monitoring to ensure safety and data completeness. Participation involves no direct treatment changes but includes diagnostic evaluation through imaging and AI analysis.

Endoscopic Ultrasound-guided Fine-needle Aspiration of Solid Pancreatic Lesions With Rapid Staining of Cytological Smears Followed by Whole Slide Scanning and Artificial Intelligence Diagnosis: A Prospective, Multicenter Study.