Blinded analysis of an exercise ECG database using high frequency QRS analysis

Noam Omer, Yair Granot, Mika Kähönen, Rami Lehtinen, Tuomo Nieminen, Kjell Nikus, Terho Lehtimäki, Jari Viik, Shimon Abboud

Research output: Contribution to journalConference articlepeer-review

Abstract

High frequency QRS (HFQRS) analysis was shown to be more accurate than ST changes in detecting stress induced ischemia in designed clinical studies. Since it utilizes energy extracted from the frequency band of 150-250 Hz, HFQRS analysis is performed with high end electrocardiographs which are equipped with adequate hardware. In this study, we aim to examine whether it is possible to perform the HFQRS analysis using ECG data acquired by a 500 Hz commercial electrocardiograph and also to assess the clinical performance of such analysis. One hundred and thirty two ECG records of bicycle exercise tests were obtained from the FINCAVAS database. Fifteen records with a wide QRS duration and 28 patients who have not reached their target heart rate were excluded. HFQRS and computerized ST-segment analyses were performed for the remaining 89 records in a blinded fashion. Three records were excluded due to excessive high frequency noise. Accordingly, the group of records with a definite HFQRS interpretation included 57 patients without stenosis and 29 patients who had >75% stenosis. Angiography, was used as gold standard. The clinical performance of both methods were assessed. The HFQRS has statistically significant higher sensitivity of 86% comparing to the 41% of the ST (p<0.005) with a statistically insignificant difference in specificity of 68% vs. 67% for the HFQRS and the ST respectively (p=0.84). This analysis demonstrated that HFQRS analysis may be applied for ECG data acquired by standard 500 Hz electrocardiographs and demonstrates its potential in diagnosing stress induced ischemia.

Original languageEnglish
Pages (from-to)1-4
Number of pages4
JournalComputing in Cardiology
Volume44
DOIs
StatePublished - 2017
Event44th Computing in Cardiology Conference, CinC 2017 - Rennes, France
Duration: 24 Sep 201727 Sep 2017

Fingerprint

Dive into the research topics of 'Blinded analysis of an exercise ECG database using high frequency QRS analysis'. Together they form a unique fingerprint.

Cite this