A Comparative Study of ECG Sampling Rates in Predicting Cardiac Arrhythmias Using a Deep Learning Approach

Cardiac arrhythmias are a group of conditions that have caused great concern among healthcare professionals because they have a high incidence and prevalence worldwide, and their early diagnosis and effective treatment are essential to alleviate the negative impacts they can have on people’s quality of life. In recent years, many studies have suggested Deep Learning models for diagnosing cardiac arrhythmias by analysing data from electrocardiograms (ECG). The heart’s electrical activity can be measured using different ECG sampling rates, and the chosen rate can impact the model’s prediction accuracy and associated computational costs. This study presents a comparative analysis of various sampling rates for predicting cardiac arrhythmias using a Deep Learning model and a large public dataset. The findings can serve as a guideline for doctors and researchers to select the most appropriate sampling rates to achieve greater efficiency when using the ECG and better outcomes in predicting cardiac arrhythmias.