A comparison of methods for the classification of atrialfibrillation from intra-atrial electrograms

摘要

Reliable classification of atrial fibrillation from intra-atrialsignals is a central task in the development of implantable atrialdefibrillators. Methods that satisfactorily discriminate normal sinusrhythms, atrial tachycardia and atrial fibrillation include atrialperiod (AP) amplitude probability density function (APDF), correlationwaveform analysis (CWA) and power spectrum density (PSD). Neverthelesstheir sensitivity in discriminating among and classifying AF types hasbeen seldom addressed. We compared parameters obtained by the followingmethods for the classification of the AF rhythm, according to Wells'types: AP and its coefficient of variation (CV), number of points in thebaseline (NO) and Shannon entropy (ENTR) from APDF, correlationcoefficient from CWA and indexes obtained from PSD. To evaluate andcompare the proposed methods ANOVA and Student's test were used. Clusteranalysis was also used to define the best set of parameters to be usedand to test the overall performances. We used data from intra-atrialrecordings from chronic AF patients and from subjects with electricallyinduced AF. The best discriminating parameters were AP, CV, NO and CWA.Cluster analysis using this set satisfactorily discriminated AF1 and AF3electrograms. Type II AF is still difficult to identify