Classification of cardiac arrhythmias using competitive networks

Area of Science:

• Biomedical Engineering
• Artificial Intelligence in Healthcare
• Cardiology

Background:

• Continuous vital signal data from patient sensors presents monitoring challenges.
• Traditional methods require specialized equipment and programs for real-time data processing.
• Biomedical devices generate vast, complex data streams requiring advanced analytical approaches.

Purpose of the Study:

• To analyze data from biomedical devices, specifically electrocardiogram (ECG) signals.
• To develop and apply a neural classifier for automated ECG analysis.
• To detect the presence of cardiac arrhythmia in ECG data.

Main Methods:

• Utilized Kohonen competitive neural networks as the primary classification tool.
• Processed continuous data sequences from ECG recordings.
• Implemented a classification system to differentiate between normal and abnormal cardiac rhythms.

Main Results:

• Successfully classified ECG signals using the Kohonen neural network.
• Demonstrated the capability to detect cardiac alterations indicative of arrhythmia.
• Validated the neural network's effectiveness in identifying ECG signal abnormalities.

Conclusions:

• Kohonen competitive neural networks are effective for analyzing ECG data.
• Automated detection of cardiac arrhythmia from ECG signals is feasible.
• This approach offers a viable method for monitoring heart rhythm normality.