IJBEM, Vol. 5, No. 1, 2003

Reproducibility of Atrial Electromagnetic Signal Analysis with Multichannel Magnetocardiography

Mika Lehto^a, Raija Koskinen^a, Heikki Väänänen^b, Juha Rantonen^a, Juha Montonen^c, Markku Mäkijärvi^a, Lasse Lehtonen^d, Liisa-Maria Voipio-Pulkki^e, Lauri Toivonen^a

^aDivision of Cardiology, Helsinki University Central Hospital, Helsinki, Finland
^bLaboratory of Biomedical Engineering, Helsinki University of Technology, Helsinki, Finland
^cBioMag Laboratory, Helsinki University Central Hospital, Helsinki, Finland,
^dLaboratory Services, Helsinki University Central Hospital, Helsinki, Finland,
^eDeparment of Medicine, Helsinki University Central Hospital, Helsinki, Finland

Correspondence: M Lehto, Division of Cardiology, Helsinki University Central Hospital, P.O. Box 340 00290 Helsinki, Finland.
E-mail: mika.lehto@fimnet.fi, phone +358 9 4711, fax +358 9 47174574

Abstract. The aim of this study was to explore the reproducibility of atrial electromagnetic signals with magnetocardiography (MCG) and signal averaged ECG (SAECG). Nine healthy controls and nine patients with paroxysmal lone atrial fibrillation (AF) underwent two MCG and SAECG recordings. The onset and offset of averaged and high-pass filtered (40 Hz) atrial complexes were automatically set. Mean duration of atrial P-waves in included channels (Pd), dispersion in Pd (PDi = SD/Pd x 100), area of P-wave and root mean squares (RMS) of the last 20 and 60 ms of atrial P-waves were measured. Reproducibility was studied as mean difference between the measurements and as coefficient of variation (CV=SD/mean value of the recordings x 100). The mean difference of Pd was 2.9 ms and CV 1.9 % by MCG and 4.7 ms and 3.1 % by SAECG. Between the groups, CV of Pd by MCG was slightly higher in patients (2.6% vs. 1.3 % in controls, p=0.02). Reproducibility of other measurements was less satisfactory and did not differ between the groups. Duration of the atrial signal is reproducibly recorded with MCG in paroxysmal lone AF patients and in healthy controls. Reproduciblity of other more specified atrial parameters seems to be lower.

Keywords: Magnetocardiography; Signal Averaged ECG; Signal Processing; Atrial Fibrillation