Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

In this paper, we apply a sparse Bayesian learning algorithm called the Relevance Vector Machine (RVM) which was used to classify the 1126 ischaemic ST events and 1126 non-supply ischaemic ST events in the Long Term ST Database as supply or non-supply ST episodes. A Genetic Algorithm (GA) method was used to identify which of the extracted features used as input to the RVM were the most important with respect to the model's performance. The GA indicated that 9 of the 35 extracted features were the most relevant. The 9 features that were selected are heart rate variability, slope of the ST segment, energy in the QRS complex and Mahalanobis distance of the first five Karhunen Loève Transform of the QRS complex and ST segment for differentiation between supply and non-supply ischaemic ST episodes. The classification accuracy achieved using the 35 features was 80.1% on the test set. When using the 9 most relevant features determined from the GA, the classification accuracy rose to 87.4%. © 2011 CCAL.

Type

Journal article

Journal

Computing in Cardiology

Publication Date

01/12/2011

Volume

38

Pages

633 - 636