ECAT- Electrocardiogram Analysis Tool -
For Assessment of Risk for Multiple Life-Threatening Cardiac Ailments

The research article presents a pioneering Non-Invasive methodology attempting to detect multiple cardiac ailments. A flexible software analysis tool has been developed which given an input rhythm strip of conventionally available high-resolution surface electrocardiogram (ECG) performs the wavelet analysis to detect the presence of cardiac abnormalities. ECAT has the ability to convert the acquired ECG data, pre-process the ECG signal, wavelet transform the processed signal and presents the results graphically in 3D. The design of the tool addresses the important issue of providing a user interface for clinical experts to choose the options to view the Wavelet Transformed ECG. This article specifically deals with the results of the analysis done to detect presence of Ventricular Arrhythmias and Tachycardia - Wide spread cardiac ailments.

Ventricular Tachycardia (V-Tach) – The signature signal for inducible patients is believed to have potentials known as late potentials in the magnitude of 40mV (typically) and of frequency of 40 – 200 Hz in the ST segment of the cardiac waveform.

Electrocardiogram (ECG) data is acquired using a high-resolution instrument, capable of resolving signal amplitudes as low as 2-3 mV with a sampling rate of 1000 samples per second.  The acquired data is in BARD-Binary format. The data is converted from the Bard-Binary format to a format suitable for the preferred development package – MATLAB®. The converted data is run through several Signal Preprocessing stages preparing the data for wavelet analysis. The choice of wavelet transform analysis makes this research effort novel in the realm to bio medical signal processing. The wavelet transform analysis has the advantage of a joint time and frequency domain representation compared to the traditional time or frequency domain representation using Fourier analyses. This is best suited for transient signals like the ECG/EEG signals where one is interested in the time localization of the frequency components. In the case of V-Tach detection, the need is to localize a 40mV anomaly in the 40-200Hz range at in time range of 200-300ms in a cardiac cycle that lasts for 600ms. The results using ECAT to assess risk for the presence of V-Tach are as shown:

For patients with the risk for V-Tach the ECAT tool has localized the presence of high amplitude coefficients in the time range of 250-300 ms and in the frequency range 40-200 Hz (corresponding to 10-50 in wavelet scale coefficients). The software has been used successfully to detect the presence of V-Tach as shown in this article; the initial results have been hugely promising. However, there large amount of data is required for more accurate testing.

|