Robust Detection of Heart Beat in Multimodal Data

The project aims at creating and exploring ways to robust the ECG signal.

The robust methods for locating heart beats in continuous long-term data from bedside monitors and similar devices that record not only ECG but usually other physiologic signals as well, including pulsatile signals that directly reflect cardiac activity, and other signals that may have few or no observable markers of heart beats using neural network.

The project aims at creating and exploring ways to robust the ECG signal made available raw from any ECG device or if possible to recreate the same from the available physiological signals if contaminated with noise. The robusted signal or the recreated signal is made to the near proximity of the original signal so that the error is very less so that the signal created doesn’t diverge out to the maximum acceptable limit to extract the features from the signal.

Matlab Code for used for : the Robust Detection of Heart Beat in Multimodal Data. The first part is for CREATING DATABASE The second part is for NEURAL NETWORK TRAINING AND TESTING in have created a feed forward backprop network with 3 hidden layer of 12,20 &8 resp neuron.

The Physionet ATM Waveform Database data are used by making use of all relevant information. Data for this project are 1-minute (or occasionally shorter) excerpts (“records”) of longer multi-parameter recordings of human adults, including patients with a wide range of problems as well as healthy volunteers. Each record contains four to eight signals; the first is an ECG signal in each case, but the others are a variety of simultaneously recorded physiologic signals that may be useful for robust beat detection. We took 10 training records, for evaluation of project.