Supplementary Materials: AI Diagnostic Models for Sickle Cell Severity Stratification - Datasets, Code, Checklists

This repository supports the manuscript: "Diagnostic Accuracy of Artificial Intelligence Models for Stratifying Disease Severity in Sickle Cell Disorders..." **Contents**: 1. PRISMA-DTA/PROBAST-AI/CLAIM Checklists 2. Complete database search strategies (MEDLINE, Embase, etc.) 3. R scripts for meta-analysis (metafor, meta4diag) 4. Python code for SHAP analysis and LightGBM deployment 5. Ethnicity-specific threshold calibration tables 6. FHIR API templates for EHR integration **Validation**: - Reproduces all figures/tables in the manuscript - CPU-optimized models for low-resource settings (Raspberry Pi compatible)