Title :
Beat-to-beat analysis of QRS duration in time and wavelet domains
Author :
Laciar, E. ; Jané, R. ; Brooks, Dh
Author_Institution :
Dept. ESAII, Univ. Politecnica de Catalunya, Barcelona, Spain
Abstract :
The authors propose a temporal and wavelet analysis to assess the beat-to-beat variability of QRS duration (QRSD) in healthy subjects and in patients with Chagas´ disease who are prone to ventricular tachycardia (CH-VT). An improved temporal QRSD (TQD) is obtained from the vector magnitude of the filtered XYZ leads. The wavelet duration (WQD) is determined from the magnitude of a quadratic spline wavelet transform of the unfiltered leads. In contrast to temporal methods, no correlation was found between WQD and noise level. QRSD variability was quantified by the standard deviation of TQD and WQD for normal beats (TQDV and WQDV). CH-VT patients tested had higher values of QRSD than healthy subjects in both domains (TQD: 117.1±13.9 ms vs. 95.6±4.2 ms and WQD: 121.2±13.7 ms vs. 93.1±3.1 ms) and exhibited a higher QRSD variability (TQDV: 8.5±1.5 ms vs. 5.4±0.9 ms and WQDV: 5.5±1.2 ms vs. 2.7±0.4 ms). It is concluded that WQDV may robustly discriminate CH-VT patients from healthy subjects
Keywords :
diseases; electrocardiography; medical signal processing; splines (mathematics); time-domain analysis; wavelet transforms; 2.7 to 121.2 ms; Chagas´ disease; ECG analysis; QRS duration; beat-to-beat variability; electrodiagnostics; filtered XYZ leads; healthy subjects; quadratic spline wavelet transform; temporal analysis; unfiltered leads; vector magnitude; ventricular tachycardia; wavelet analysis; wavelet duration; Algorithm design and analysis; Cancer; DH-HEMTs; Diseases; Measurement standards; Noise level; Signal analysis; Time measurement; Wavelet analysis; Wavelet domain;
Conference_Titel :
Computers in Cardiology 2000
Conference_Location :
Cambridge, MA
Print_ISBN :
0-7803-6557-7
DOI :
10.1109/CIC.2000.898625