CNN-Based Real-Time System for Atrial Fibrillation Detection Using ECG Signal

Automated detection of atrial fibrillation (AF) is essential for the early diagnosis and treatment of cardiac conditions. This chapter proposes a Convolutional Neural Network (CNN)-based approach for automated AF classification. ECG records from standard databases are utilized for the simulation study. A CNN model is developed and trained using these ECG datasets, and its effectiveness is evaluated under both intra-patient and inter-patient conditions. The proposed method achieves an average accuracy of 99.88% for intra-patient ECG records and 94.17% for inter-patient records. Real-time performance analysis is performed by implementing the scheme on a Raspberry Pi system. A comprehensive analysis of the results confirms that the scheme effectively detects AF from ECG records on a limited resource-based embedded platform.