Key result
Supervised learning-based noise detection achieves ~94% accuracy in classifying noisy ECGs, outperforming the EMD+NLM method.
Why the study?
Noisy electrocardiograms frequently cause false alarms in distributed computing settings such as the medical Internet of things, necessitating methods to detect noise to improve arrhythmia detection.
Does a supervised learning-based noise detection algorithm improve the signal-to-noise ratio of filtered ECG signals compared to standard filtering of raw ECG signals?
Does a supervised learning-based noise detection algorithm improve the signal-to-noise ratio of filtered ECG signals compared to standard filtering of raw ECG signals?
A supervised learning approach to detect noise in ECG signals may improve the performance of standard filters in medical IoT settings.
Supports integration of ML noise detection in ECG IoT systems; leaves open impact on filtered SNR versus standard methods.
A significant purpose in detecting pervasive computing approaches is to improve the specificity and sensitivity of arrhythmia detection using electrocardiograms. Because ECG signals frequently propagate over distributed computer settings such as the medical Internet of things, noise is prevalent (medical IoT). In these distributed and widespread computing-aided methodologies, noisy electrocardiograms are common to false alarms. The fundamental goal of the machine learning-based arrhythmia detection algorithms discussed in this publication is to detect noise scope in electrocardiograms. The suggested approach determines whether or not the provided electrocardiograms are influenced by noise. In this regard, the approach takes advantage of the electrocardiogram's temporal and spectral characteristics. The performance of the suggested technique was evaluated using multifold cross-validation. In addition, a comparative study was performed comparing the average of peak Signal to Noise Ratio, respective standard deviation and filter sequence length obtained from filtering noise by FIR and IIR filters from raw ECG signals and the SLND-selected noisy ECG signals.
No takes yet. Share an insight, caveat, or question.
Supraja et al. (2023) studied ECG noise detection (n=14,423). Supervised Learning-Based Noise Detection (SLND) vs. Empirical Mode Decomposition with Non-Local Mean (EMD+NLM) was evaluated on Prediction accuracy for noise detection. The Supervised Learning-Based Noise Detection (SLND) method achieved a prediction accuracy of 94.41% in classifying noisy versus qualified ECGs, outperforming the contemporary EMD+NLM method (92.71%).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: