Comparative Study of Signal-Averaged Electrocardiogram between Time Domain Analysis and Spectral Turbulence Analysis by 24-hr Holter Monitoring

Lee, Chang Won; Jang, Duk Whan; Baek, In Seok; Kim, Jang Won; Bae, Soon Chul; Kim, Cheol Woo; Kweon, Jae Koo; Choi, Rak Kyeong; Rhee, Moo Yong; Lee, Hong Soon; Yoo, Soo Woong

Abstract

OBJECTIVES
Signal-averaged electrocardiography (SAECG) has been found to be a useful noninvasive technique for identifying patients at risk for life-threatening ventricular tachycardia. Delayed and fragmented activation of abnormal myocardial tissues causes the occurrence of high frequency low amplitude (HFLA) electocardiographic signals or late potentials. Generally, there are two methods in analyzing signal-averaged electrocardiography. Late potentials in the time domain analysis do not provide sufficient diagnostic power with regard to life-threatening Ventricular tachycardia. Buckingham et al. (1989) reported a time-domain sensitivity of 62%, a specificity of 75%. Spectral turbulence analysis (STA) of the signal-averaged ECG is the most recent frequency domain technique to improve the time domain sensitivity and specificity. So, We designed the study to compare the efficacy of Time Domain Analysis and Spectral Turbulence Analysis among five groups (Normal control, QRS widening, Postmyocardial infarction, Frequent VPC's with group beats, Nonsustained ventricular tachycardia).
METHODS
88 patients were selected from the patients who had been admitted between January 1994 and October l994, at National Medical Center. Patients were divided into five groups, which were respectively, Group A: Normal control group (n=33), Group B: QRS widening group (n=14), Group C: Postmyocardial infarction group (n=10), Group D: Frequent VPC's with group beats (n=22), Group E: Nonsustained VT group (n=9). We compared Spectral Turbulence Analysis and Time Domain Analysis of Signal-Averaged Electrocardiogram by 24 hours-Holter monitoring.
RESULTS
1) In normal control group(Group A), 9.1%(3 patients) were positive by Time Domain Analysis, but, all were negative by Spectral Turbulence Analysis. 2) In QRS widening group (Group B), 71.4%(10 patients) were positive by Time Domain Analysis, but, all were negative by Spectral Turbulence Analysis. 3) In postmyocardial infarction group (Group C), 309o were positive by Time Domain Analysis, and 10% were positive by Spectral Turbulence Analysis. 4) In frequent VPC's group (Group D), 22.7% (5 patients) were positive by Time Domain Analysis, and, 4.5%(1 patient) was positive by Spectral Tur-bulence Analysis. 5) In Nonsustained VT group (Group E), 33.3% (3 patients) were positive by Time Domain Analysis, and 11.1% (1 patient) was positive by Spectral Turbulence Analysis.
CONCLUSIONS
In Time Domain Analysis, abnormal results were presented at Group R (QRS widening group) by 71.4%, which was markedly higher than other groups. But, in Spectral Turbulence Analysis, abnormal results were not presented at Group A and Group B. In Group A and Group B, Spectral Turbulence Analysis shows less false positive results than Time Domain Analysis.