» BEATS DOING IT YOURSELF
Adaptive Filters
Applied to Heart ECG


Navigation

Home
Abstract
ECG Filtering
Linear
Parsing
Introduction
Theory
Methods
Results
Future
Heart Rate Variability
Conclusion
Team
Code

Future Research

     is recommended for this algorithm: Although this algorithm is effective, it is not perfect, and there are many places where it could be improved. Since the algorithm is quite effective for short- or low-noise ECG data, all that is needed is an accurate trigger which determines when the SNR is too close to 1 for a sustained period of time. We implemented this filter, but we did not have sufficient time to optimize it, and remains quite prone to error. However, the trigger being developed was not effective, and since the exact same trigger was implemented into the linear filtering method as a boxcar filter, it seems quite likely that this modified version of the parsing algorithm could be made quite accurate.

In addition, it might worthwhile to implement this filter as a matrix-based filter, for currently it has been developed as a real-time parsing algorithm that reads data values one line at a time.







Template © 2002 Devil Design