Endoscopic Image-Based Prediction of Esophageal Stenosis after ESD Using Mucosal Defect Metrics

Area of Science:

• Gastroenterology
• Surgical Oncology
• Medical Imaging

Background:

• Esophageal stricture is a significant complication following endoscopic submucosal dissection (ESD).
• Current criteria for predicting stenosis after ESD are not quantitative.
• Identifying predictive factors for esophageal stenosis post-ESD is crucial.

Purpose of the Study:

• To identify predictive factors for esophageal stenosis after ESD.
• To explore quantitative indicators for predicting postoperative stenosis.
• To develop a predictive model for esophageal stenosis post-ESD.

Main Methods:

• Retrospective analysis of endoscopic data from patients undergoing esophageal ESD.
• Measurement of resected mucosal defect area and perimeter using Image Pro Plus 6.0.
• Logistic regression, ROC curve analysis, and nomogram model construction for prediction.

Main Results:

• Median defect area: 527.74 mm², median perimeter: 91.75 mm, median circumferential ratio: 33%.
• Defect area, perimeter, and circumferential ratio were independent risk factors for stenosis (P < 0.05).
• Prediction models showed good internal validation (C-statistics: 0.747-0.875).

Conclusions:

• Quantitative analysis of resected mucosal defect area and perimeter accurately predicts esophageal stenosis after ESD.
• Image analysis technology provides a reliable tool for assessing stenosis risk.
• This quantification aids in managing patients undergoing esophageal ESD.