Artificial Intelligence Assisted Topographic Mapping System for Endoscopic Submucosal Dissection Specimens

Area of Science:

• Gastroenterology
• Pathology
• Artificial Intelligence

Background:

• Endoscopic submucosal dissection (ESD) is a key minimally invasive surgery for early gastrointestinal cancers.
• Accurate pathological diagnosis of ESD specimens is crucial for effective treatment.
• Traditional topographic mapping of ESD specimens is labor-intensive and prone to inaccuracies.

Purpose of the Study:

• To develop and evaluate an artificial intelligence (AI)-assisted topographic mapping system for ESD specimens.
• To automate and improve the accuracy and efficiency of the mapping process.

Main Methods:

• A topographic mapping system was developed utilizing computer vision techniques.
• The system was tested on 23 ESD cases from Peking Union Medical College Hospital.
• Reconstruction maps were generated using both the AI system and traditional methods for comparison.

Main Results:

• The AI system significantly reduced mapping time, saving between 34 to 3,336 seconds per case.
• No significant differences were observed in specimen shape, size, or tumor area between the AI system and traditional methods.
• The AI system demonstrated high accuracy in topographic mapping.

Conclusions:

• An AI-assisted system can effectively automate topographic mapping of ESD specimens.
• This technology offers a faster and more accurate approach to pathological analysis of ESD specimens.
• The developed system has the potential to enhance diagnostic precision and workflow efficiency in gastrointestinal pathology.