Throb: system-on-chip based arrhythmia screener with self interpretation / Huey Woan Lim ...[et al.]

• Main Authors: Huey, Woan Lim, Mohd Sani, Mohd Syafiq Affendi, Hashim, Amin, Yuan, Wen Hau
• Format: Article
• Language: English
• Published: UiTM Press 2015
• Subjects: Medical technology; Computer applications to medicine. Medical informatics; Neural Networks (Computer). Artificial intelligence
• Online Access: https://ir.uitm.edu.my/id/eprint/62982/1/62982.pdf; https://ir.uitm.edu.my/id/eprint/62982/; https://jeesr.uitm.edu.my/v1/

This paper presents an in-house design of System-onChip (SoC) based arrhythmia screener, so-called Throb, with selfarrhythmia classification using electrocardiograms (ECG) as the input signal. It is a light-weight, cost effective and equips with intuitive touch screen graphical user interfaces (GUI) design. It is able to provide early screening of arrhythmias for the public, especially for the small clinics and general hospital in rural area where the specialists or cardiologists are not sufficient to the population. Throb applies knowledge-based classification to identify Premature Ventricular Contraction (PVC), Ventricular Fibrillation (VF), Second Degree Heart Block, and Atrial Fibrillation (AF). The verification input is based on offline ECG dataset obtained from MIT BIH online arrhythmia database. The complete system is implemented on Terasic Video Embedded Evaluation Kit with Multitouch (VEEK-MT) which utilizes the Altera Cyclone IV FPGA chip and capacitive touch screen. This system is also equipped with the ECG acquisition unit to obtain the ECG from the patient as input signal. Result shows that this system is user friendly, and the arrhythmia classification accuracy of PVC is 88.56%, VF is 96.30%, 2nd degree heart block is 85.71% and AF is 86.17%, respectively.