ECGLight: Compute-Light Framework For Paper ECG Digitization and Myocardial Infarction Screening
The framework converts smartphone photos or scans of paper ECGs into calibrated 12-lead digital signals using a lightweight, on-device pipeline running on CPU-only resources. It uses YOLO models for segmentation and layout detection, followed by classification for Myocardial Infarction (MI) screening. This allows legacy paper records to be used for decision support in areas lacking high connectivity or powerful computers.