Details

Autor(en) / Beteiligte

• SLOCUM, JANET ELIZABETH

Titel

COMPUTER DETECTION OF ATRIO-VENTRICULAR DISSOCIATION AND ATRIAL FIBRILLATION (ELECTROCARDIOGRAM)

Ort / Verlag

ProQuest Dissertations Publishing

Erscheinungsjahr

1986

Link zum Volltext

• ProQuest

Quelle

ProQuest Dissertations & Theses A&I

Beschreibungen/Notizen

• The detection of atrial activity is the most significant limiting factor in the diagnostic abilities of currently available arrhythmia interpretation systems. Current arrhythmia interpretation systems for use on the surface electrocardiogram attempt to detect atrial activity only in the isoelectric region between the QRS complex and the preceding T wave and only when atrial activity is in the form of P waves. The aim of this research is to produce algorithms to detect atrial activity when little isoelectric region is present, the P waves occur outside of the isoelectric region or atrial activity is not in the form of P waves. A-V dissociation and atrial fibrillation are two arrhythmias in which atrial activity is not synchronous with ventricular activity. A remainder electrocardiogram was generated by subtracting a mean beat from each beat of the original electrocardiogram and thereby isolating the asynchronous atrial activity. A-V dissociated P waves were detected by autocorrelating the remainder electrocardiogram and searching the autocorrelation function for a peak at a cycle length different from the ventricular. Atrial fibrillatory waves were detected by taking the power spectrum of the remainder electrocardiogram and measuring the amount of power in the peak of the 5-9 Hertz region. This region was found to contain large amounts of power during atrial fibrillation in intra-atrial leads. Three algorithms were generated for detection of atrial activity. An algorithm for detection of atrial fibrillation in intra-atrial leads was generated which can be used in implantable anti-tachycardia devices. Two algorithms for detection of atrial activity on the surface electrocardiogram were generated, one for detection of A-V dissociation and one for detection of atrial fibrillation, which can be added as modules to currently existing arrhythmia interpretation systems.