The Rhythm of the Heart
A mathematical analysis of heartbeat patterns can be used to diagnose certain heart disorders with flawless accuracy. Previous mathematical methods for diagnosing heart disease were mostly not reliable enough to be used clinically. However, the new technique was 100% successful in a test to identify patients with heart defects. It is based on detecting deviations in the time interval between individual heartbeats. Electrical engineer Malvin Teich and his colleagues at Boston University analyzed approximately 20-hour recordings of the heartbeats of 15 patients with heart defects and a control group of 12 he althy participants. The team applied a mathematical transformation to find the so-called wavelet coefficient on data sets ranging from two to a thousand heartbeat intervals. The researchers found, as reported in the February 16, 1998 Physical Review Letters, that the wavelet coefficients of datasets containing 16 to 32 beats were clearly different in diseased and he althy hearts, and there was no overlap between the two groups. Interestingly, all of the he althy hearts showed greater beat-to-beat variability than the diseased ones.
The technique is so simple and "foolproof" that it can be used "instantly" in hospitals to diagnose heart conditions, says Teich. Nevertheless, he notes, the method still needs to be tested on a larger number of patients. He also hopes to understand physiologically why it is precisely in the window of 16 to 32 heartbeat intervals that the normally functioning heart can be distinguished so clearly from the diseased heart.
A success of 100% is unusual and extremely remarkable, explains Ming Zhou Ding, a biophysicist at Florida Atlantic University in Boca Raton who uses wavelet analysis to study brain rhythms. Teich and his colleagues attribute their success to the technique's ability to record both short- and long-term abnormalities in heart rhythms. This was not possible with previous methods that focused on either consecutive heartbeats or long-term changes.
