Detection of arrhythmia using weightage-based supervised learning system for COVID-19

Area of Science:

• Cardiology
• Infectious Diseases
• Artificial Intelligence in Medicine

Background:

• COVID-19 infection can cause severe cardiovascular complications, leading to mortality.
• Cardiovascular problems are a primary cause of death in COVID-19 patients and are key prognostic indicators.
• Detecting cardiac abnormalities like arrhythmia from electrocardiogram (ECG) signals is crucial for assessing cardiovascular health.

Purpose of the Study:

• To develop an automated system for identifying cardiovascular abnormalities from ECG signals in COVID-19 patients.
• To select the most suitable supervised learning model for accurate arrhythmia detection based on user-defined requirements.
• To improve the efficiency and accuracy of cardiovascular disorder detection in the context of COVID-19.

Main Methods:

• Utilized supervised learning algorithms for the analysis of ECG signals.
• Developed a model selection system that assigns weights based on user requirements to identify the optimal predictive model.
• Trained and tested various models to identify abnormalities in ECG waves indicative of cardiovascular issues.

Main Results:

• The automated system successfully identified abnormalities in ECG waves.
• The selected models met user-defined requirements, demonstrating high performance.
• Achieved up to 97% accuracy and 97% precision in predictive tasks for arrhythmia detection.

Conclusions:

• The proposed automated method effectively detects cardiovascular abnormalities from ECG signals in COVID-19 patients.
• The model selection system ensures the deployment of high-performing models tailored to specific needs.
• This approach offers a promising tool for early detection and improved prognosis of cardiovascular complications associated with COVID-19.