A Tele-surveillance System with Automatic ECG Interpretation based on Support Vector Machine and Rule-based Processing
Date Submitted: Mar 4, 2015
Open Peer Review Period: Mar 4, 2015 - Apr 29, 2015
Background: Telehealthcare is a global trend affecting clinical practice around the world. To mitigate the loading of health professionals and provide a ubiquitous healthcare, a comprehensive surveillance system with value-added services based on information technologies must be established. Objective: We conducted this study to describe our proposed tele-surveillance system designed for monitoring and classifying ECG signals and evaluate the performance of ECG classification. Methods: We established a tele-surveillance system with an automatic ECG interpretation mechanism. The system included: (1) automatic ECG signal transmission via telecommunication; (2) ECG signal processing, including noise elimination, peak estimation, and feature extraction; (3) automatic ECG interpretation based on the Support Vector Machine classifier and rule-based processing; (4) displaying ECG signals and their analyzing results. We analyzed 213,420 ECG signals that were diagnosed by cardiologists as the gold standard to verify the classification performance. Results: In the clinical ECG database from the Telehealthcare Center of NTUH, the experimental results showed that the ECG classifier yielded a specificity value of 96.661% for normal rhythm detection, a sensitivity value of 98.502% for disease recognition, and an accuracy value of 81.168% for noise detection, respectively. For the detection performance of specific diseases, the recognition model mainly generated the sensitivity values of 92.697% in Atrial Fibrillation, 89.104% in Pacemaker rhythm, 88.600% in Atrial Premature Contraction, 72.978% in T-wave Inversion, 62.213% in Atrial Flutter, and 62.569% in First-degree Atrio-Ventricular Block, respectively. Conclusions: Via connected telehealthcare devices, the tele-surveillance system, and the automatic ECG interpretation system, the mechanism is intentionally designed for continuous decision-making support and is reliable enough to reduce the need of face-to-face diagnosis. With this value-added service, the system could widely assist physicians and other health professionals for decision-making in clinical practice. The system will be very helpful for the patient who suffers from the cardiac disease but is inconvenient to go to the hospital very often.