Development and validation of an interpretable machine learning model for predicting chronic atrophic gastritis in elderly patients
By
Wenjing Fan
Jinyu Wang
Lu Li
Chao Tian
Zhiwei Yang
Guangchao Zhang
Deyu Xu
Xingtang Yang
July 2, 2026
Clinical Scorecard: Creation and assessment of an interpretable machine learning model for forecasting chronic atrophic gastritis in older adults
At a Glance
Category Detail
Condition Chronic Atrophic Gastritis Key Mechanisms Machine learning models predicting CAG using demographic, lifestyle, and clinical variables. Target Population Elderly patients aged 60 and above. Care Setting Primary care settings for non-invasive screening.
Key Highlights
Developed interpretable ML models for predicting CAG in elderly patients. Identified eight key predictive variables including H. pylori infection and smoking status. Achieved optimal model performance with AUC of 0.826 in internal validation. Utilized SHAP analysis for interpreting model outputs. Addressed barriers to endoscopy in elderly populations. Guideline-Based Recommendations
Diagnosis
Use machine learning models for risk stratification of CAG in elderly patients. Management
Implement early screening and targeted interventions based on ML predictions. Monitoring & Follow-up
Regularly assess predictive factors identified by the model. Risks
Consider the invasiveness and accessibility issues of endoscopy in elderly patients. Patient & Prescribing Data
Elderly individuals aged 60 years and older.
Focus on lifestyle modifications and monitoring of identified risk factors.
Clinical Best Practices
Incorporate ML models into routine screening protocols for CAG. Utilize SHAP for understanding individual risk factors in patient management. Related Resources & Content
