PARAMETER EVALUATION OF THE INVERSE POWER-LAW SPECTRUM OF HEART-RATE - A QUANTITATIVE APPROACH FOR ECG ARRHYTHMIA ANALYSIS

A preliminary study was performed to ensure reliable R-wave detection and fiducial mark location. A generalized phase-sampling model for spe ctral estimation was then used, based on an interest in beat-to-beat v ariations. Heart rate spectrum characteristics of 32 presumed-normal s ubjects and 44 records from the MIT-BIH electrocardiographic database (Massachusetts Institute of Technology, Cambridge, MA) were analyzed a nd three parameters were calculated: regression line slope, intercept, and cross correlation between the spectral data and the regression li nes. Evaluation of these inverse power-law spectrum parameters provide s a potential quantitative approach for characterizing erratic fluctua tions of heart rate and is possibly used to distinguish between health y and abnormal subjects. From the 44 recordings in the MIT-BIH databas e, a V-shaped curve was found in a plot of the regression line slope v ersus cross correlation. The results for both unmedicated and medicate d patients with normal sinus rhythm cluster in the top left region of the graph. Also, 29 of the presumed-normal subjects cluster in the sam e region. Patients with premature ventricular contraction beats cluste r in the top right region of the V-shaped curve. The rest of the recor dings from the variety of arrhythmia cases in the database have low sl opes and cross correlations, so they cluster near the apex of the V-sh aped curve. Three volunteers who each had more than 32 atrial prematur e contraction and premature ventricular contraction beats also fall in this region of the graph. The results from 10 young volunteers from t he United States and 15 volunteers from seven other countries cluster into different regions. Also, presumed-normal subjects plot in differe nt locations during relaxation and under stress, as well as when they are asleep or awake. Many patients in the MIT-BIH database have regres sion line slopes greater than those from the presumed-normal subjects, leading to a conclusion that abnormal rhythm may decrease in low-freq uency fluctuations or increase in high-frequency fluctuations. Also, h eart rate frequency spectrum distributions were analyzed for three dif ferent temporal lengths of each electrocardiogram to investigate the h ypothesis of temporal self-similarity. These results do not support th is hypothesis.