Digital processing of equine exercise electrocardiograms.

This research outlines the successful development and application of a system that digitally records and analyzes the electrocardiogram (ECG) of a horse during exercise, processing data susceptible to disruption from exercise artifact effectively and providing accurate ECG readings.

Research Methodology

• The study made use of a sophisticated system involving a commercial ECG transmitter. This transmitter captured a Z lead (base-apex) ECG from an exercising horse, telemetering the information for further analysis.
• The data thus received was remodulated at an audio frequency and stored conveniently on an audio cassette recorder facilitating data capture over longer periods of up to 10 minutes of exercise.
• The data was then digitized using Macintosh II computer, a crucial part of the process and central to the success of the entire venture. It allowed for further digital manipulation and analysis of the cardiac signals.

Analysis Process

• For analyzing the rhythm, a computer program was developed to identify the various waves of the ECG. This was achieved using a refined version of the integrated-squared-derivative (ISD) technique.
• The refined ISD technique included frequency-selective digital filtering. Once completed, the differentiation was followed by integration of the squared derivative.
• The ECG waves were identified using an adaptive thresholding system. This system was applied on both the ISD and the derivative of the ECG to secure more precise results.
• This technique was successful in correctly identifying the Q, R, and the end of T in ECGs that were severely contaminated by exercise artifacts – an impressive feat considering the susceptibility of ECGs to disruption through physical activity.

Results and Data Export

• The outcome was the accurate location of the absolute time of occurrence of the Q, R, and end of T waves within the ECG data.
• This information was then stored in a text file, ready to be exported to other analysis programs like spreadsheets and statistical packages. Thus, facilitating not only an effortless data capture and extraction but also easy interoperability with numerous other data processing applications.