Development of Real-time DSP Algorithms for Biomedical Signal Analysis(III)

In order to achieve the goals of biomedical signal monitoring, analysis and diagnosis, this project aims to the development of innovative, effective and reliable algorithms that accommodate for the real-time processing of the digitized biomedical signals received from the front end. The Electrocardiogram (ECG) signal will be the model biomedical signal adopted for demonstrating all the investigations throughout this project. The research covers five subjects as follows: (A) ECG QRS detection, ST, QT segment measurements, (B) Arrhythmia analysis and detection (e.g., Premature Ventricular Contraction (PVC) detection), (C) ECG data reduction/compression, (D) Heart Rate (HR) conversion for monitoring, and (E) Heart Rate Variability (HRV) time- and frequency-domain analysis. All these DSP-based algorithms are devised for real-time applications and will be verified using an ARM-based embedded system. Finally, both the developed software and hardware IP’s will be integrated as a System-on-Chip (SoC). We expect the SoC may achieve the functions listed as follows: (1) suspicious arrhythmia detection, storage and transmission for further diagnosis, (2) ECG features such as RR, PR, QT, ST intervals computation, (3) qualitative and quantitative HRV analysis for diagnostic decision support.

Project ID:PB9308-5075
External Project ID:NSC93-2220-E182-003