Automatic Detection of Paroxysmal Atrial Fibrillation

Autor: Redmond Shouldice, Jong-Il Choi, Conor Heneghan, Philip de Chazal
Rok vydání: 2012
Předmět:
Zdroj: Atrial Fibrillation: Basic Research and Clinical Applications
DOI: 10.5772/26860
Popis: The purpose of this chapter is to provide a tutorial level introduction to (a) the physiology and clinical background of paroxysmal (intermittent) atrial fibrillation (PAF), and (b) methods for detection of patterns consistent with AF using electrocardiogram (ECG) processing. The document assumes that the reader is familiar with basic signal processing concepts, but assumes no prior knowledge of AF or pattern classification. A practical implementation of an automatic AF detector is presented; a supervised linear discriminant classifier is used to estimate the likelihood of a block of inter-heartbeat intervals being PAF, with accuracies of 92%, 94%, 100% and 100% when the method was used to process the publically available Physionet (Goldberger et al., 2000) signal databases MITDB, AFDB, NSRDB and NSR2DB respectively.
Databáze: OpenAIRE