EE 5345/7345

Medical Signal Analysis

Instructor: Carlos Davila

Dept. of Electrical Engineering, Southern Methodist University

Module 2: Digital Filtering of EKG and EEG Data

Homework Problems:

Work these review problems in digital signal processing.

Project:

The project component of this module will consist in designing a series of filters for removing the noise present in the following EKG and EEG signals. The EKG is a standard bipolar lead recording which has been corrupted by additive noise sampled at 100 samples/sec, the EEG has been sampled at 200 samples/sec. Begin by performing a frequency analysis of the EKG and EEG data using the averaged periodogram to establish the relative frequency composition of the two signals and to determine the nature of the noise (each signal has a different type of noise). Also look at each signal and the time domain, and try computing a spectrogram of the EKG signal and describe how the frequencies in the signal change with time. Then design a filter which removes the noise in each signal. Design filters using the following methods:

IIR: bilinear transform

FIR: windowing

FIR: Parks-McClelland

You may use Matlab tools for doing this design. The most important part of this project is that you indicate what your design criteria are. Submit a plot of the noisy signal and of the filtered signal. Discuss the relative performance of each of the filters in meeting your goal of removing the noise in the EKG and EEG signals.