Reliable ECG Rhythm analysis and detection of serious cardiac trends are essential for the recognition of cardiac rhythm diseases and also for providing quick treatment. This will improve cardiac safety in many ECG applications including home-monitoring, ambulatory Holter, and cardiac drug safety.
The Holter ECG analysis program, presented in this work, provides localization, classification, and measurements for every single heartbeat with very high precision even in case of noisy ECG signals. Besides, numbers of basic and advanced cardiac rhythm analyses are performed by the program in order to detect numerous cardiac events including critical ventricular and atrial arrhythmia. Furthermore, various cardiac trends including PR interval, QT and QTc trendings are also provided. The program is designed not only to handle single-channel and multi-channel lead schemes but also to manage input ECG signals digitalized in different sample frequencies. Furthermore, it can be configured to meet the individual needs of users.
The detection of the heartbeats is validated using MIT-BIH Arrhythmia Database, AHA Series1 and AHA Series 2. The sensitivity and specificity are both greater than 99.7 % in all databases. The classification of the heartbeats is also validated the same databases mentioned above. The sensitivity and specificity are greater than 93.5 % and 96.5 % in all databases, respectively. The delineation algorithm used by the program is validated and tested according to the procedure and databases described in the standard IEC 60601-2-25:2011. Several initial validations showed very promising results in detecting atrial fibrillation and ventricular fibrillation sofar. However, extensive validation processes are carried out at present for many critical cardiac events. Results are expected in the near future.
Due to its high accuracy, the program can be used to minimize the hazard and risk to patients and to increase the overall cardiac safety.