Cardiac disease classification using heart rate signals

Area of Science:

• Cardiology
• Biomedical Engineering
• Data Science

Background:

• Heart rate and Heart Rate Variability (HRV) are crucial indicators of cardiovascular health.
• HRV analysis is increasingly vital in cardiology for identifying cardiac abnormalities.

Purpose of the Study:

• To evaluate the efficacy of linear and nonlinear HRV measures for classifying cardiac diseases.
• To compare the performance of Random Forests, Logistic Model Tree, and Multilayer Perceptron Neural Network classifiers.

Main Methods:

• Utilized linear (time and frequency domain) and nonlinear HRV parameters.
• Employed Random Forests, Logistic Model Tree, and Multilayer Perceptron Neural Network for classification.
• Used data from standard ECG databases available on Physionet.

Main Results:

• Classification accuracy was assessed using linear parameters, nonlinear parameters, and their combination.
• The combination of linear and nonlinear HRV measures yielded superior performance in cardiac disease classification.
• Results demonstrated that combined measures are more effective than individual linear or nonlinear measures.

Conclusions:

• The integration of linear and nonlinear HRV analysis provides a more robust approach to diagnosing cardiac diseases.
• This study's findings support the use of combined HRV metrics for enhanced cardiovascular risk assessment.
• The employed classification methods show promise for clinical application in cardiology.