Key result
RAMS-AEF segments ECG waveforms with ~95% accuracy.
Why the study?
Precise ECG waveform segmentation from image-based recordings is important in diagnosing cardiac abnormalities, prompting the development of a dynamic region-aware multi-threshold segmentation method with adaptive energy functions to improve upon conventional fixed-threshold techniques.
Does the Region-Aware Multi-Threshold Segmentation integrated with Adaptive Energy Functions (RAMS-AEF) method improve ECG waveform segmentation accuracy compared to existing methods in image-based ECG recordings of cardiac patients?
Does the Region-Aware Multi-Threshold Segmentation integrated with Adaptive Energy Functions (RAMS-AEF) method improve ECG waveform segmentation accuracy compared to existing methods in image-based ECG recordings of cardiac patients?
Answers are grounded in this paper and cite their sources.
The proposed RAMS-AEF algorithm achieves high accuracy (94.8%) and F1-score (94.3%) in isolating P waves, QRS complexes, and T waves from image-based ECG recordings.
Precise ECG waveform segmentation from image-based recordings is important in the diagnosis of cardiac abnormalities. This paper presents a new Region-Aware Multi-Threshold Segmentation (RAMS) method integrated with Adaptive Energy Functions (AEF) tailored for isolating ECG waveforms. In contrast to conventional fixed-threshold techniques, the new model dynamically adapts threshold levels between regional waveform segments in accordance with shape complexity, local contrast, and frequency information. The adaptive energy functions are sensitive to gradient and intensity changes, successfully separating P waves, QRS complexes, and T waves from complex background images. Testing was conducted on the ECG Images Dataset of Cardiac Patients, consisting of 11 148 images from four classes: myocardial infarction (2880), abnormal heartbeat (2796), history of infarction (2064), and normal persons (3408). The RAMS-AEF model used here, proposed for this work, recorded a segmentation accuracy of 94.8%, precision of 93.5%, recall of 95.2% and F1-score of 94.3%, a performance much better compared to existing methods.
No takes yet. Share an insight, caveat, or question.
Vasanth et al. (2026) studied Cardiac abnormalities (myocardial infarction, abnormal heartbeat, history of infarction, and normal controls) (n=11,148). Region-Aware Multi-Threshold Segmentation integrated with Adaptive Energy Functions (RAMS-AEF) vs. Conventional fixed-threshold techniques / existing methods was evaluated on ECG waveform segmentation accuracy. The RAMS-AEF model achieved 94.8% segmentation accuracy, 93.5% precision, 95.2% recall, and a 94.3% F1-score for ECG waveform isolation across 11,148 cardiac ECG images.
Synapse has enriched 10 related papers. Consider them for comparative context: