NCT07381192

An Artificial Intelligence System for Multimodal, Multi-class Diagnosing Solid Pancreatic Lesions Based on Endoscopic Ultrasound

• Status: Recruiting
• Phase: —
• Study type: Observational
• Enrollment: 383 (estimated)
• Sponsor: Qilu Hospital of Shandong University · Academic / Other
• Sex: All
• Age: 18 Years
• Healthy volunteers: Accepted

Summary

The aim of this study is to validate an artificial intelligence system named iEUS-SPL(intelligent endoscopic ultrasound system-solid pancreatic lesion) for detecting and multimodal, multi-class diagnosing solid pancreatic lesions during endoscopic ultrasound(EUS) examination.

Detailed description

This is an observational study with a prospective, cohort design. We have developed an artificial intelligence system named iEUS-SPL for multimodal, multi-class diagnosing solid pancreatic lesions using endoscopic ultrasound images, endoscopic ultrasound features, clinical data and imaging features from retrospectively collected patients who underwent EUS. The lesion detection rate and diagnostic performance of iEUS-SPL in identifying solid pancreatic lesions will be evaluated in real-time EUS scanning videos over prospective enrolled cases.

Conditions

• Endoscopic Ultrasound (EUS)
• Solid Pancreatic Lesion

Interventions

Type	Name	Description
DEVICE	iEUS-SPL(intelligent endoscopic ultrasound system-pancreatic solid lesion)	The iEUS-SPL will automaticly detect solid pancreatic lesions and integrate the patients' endoscopic ultrasound images, endoscopic ultrasound features, clinical data and imaging features to perform a five-category classification for the lesions, categorizing them as pancreatic cancer, pancreatic neuroendocrine tumor, solid pseudopapillary tumor, autoimmune pancreatitis and chronic pancreatitis.

Timeline

• Start date: 2025-09-01
• Primary completion: 2028-06-30
• Completion: 2028-06-30
• First posted: 2026-02-02
• Last updated: 2026-02-02

Locations

1 site across 1 country: China