A Binary Classification of Cardiovascular Abnormality Using Time-Frequency Features of Cardio-mechanical Signals

Chenxi Yang, Nicole D. Aranoff, Philip Green, Negar Tavassolian

Research output: Contribution to journalArticlepeer-review

13 Scopus citations

Abstract

This paper introduces a novel method of binary classification of cardiovascular abnormality using the time-frequency features of cardio-mechanical signals, namely seismocardiography (SCG) and gyrocardiography (GCG) signals. A digital signal processing framework is proposed which utilizes decision tree and support vector machine methods with features generated by continuous wavelet transform. Experimental measurements were collected from twelve patients with cardiovascular diseases as well as twelve healthy subjects to evaluate the proposed method. Results reveal an overall accuracy of more than 94% with the best performance achieved from SVM classifiers with GCG training features. This suggests that the proposed solution could be a promising method for classifying cardiovascular abnormalities.

Original languageEnglish
Pages (from-to)5438-5441
Number of pages4
JournalAnnual International Conference of the IEEE Engineering in Medicine and Biology - Proceedings
Volume2018
DOIs
StatePublished - 1 Jul 2018

Fingerprint

Dive into the research topics of 'A Binary Classification of Cardiovascular Abnormality Using Time-Frequency Features of Cardio-mechanical Signals'. Together they form a unique fingerprint.

Cite this