Intracardiac Atrial Fibrillation Database

Published: March 4, 2003. Version: 1.0.0

Please include the standard citation for PhysioNet: (show more options)
Goldberger, A., Amaral, L., Glass, L., Hausdorff, J., Ivanov, P. C., Mark, R., ... & Stanley, H. E. (2000). PhysioBank, PhysioToolkit, and PhysioNet: Components of a new research resource for complex physiologic signals. Circulation [Online]. 101 (23), pp. e215–e220.

Data Description

This database consists of endocardial recordings from the right atria of 8 patients in atrial fibrillation or flutter. A decapolar catheter with 2-5-2mm spacing (7mm spacing between bipoles) was placed in four separate regions of the heart. In each region, the 5 bipolar signals were recorded along with 3 surface ECG leads. Data were digitized at 1kHz. Catheter bipoles are labeled (distal to proximal) CS12, CS34, ..., CS90. Regions are:

SVC - The distal tip of the catheter (CS12) is close to the annulus of the superior vena cava
IVC - The proximal tip of the catheter (CS90) is close to the annulus of the inferior vena cava
TVA - The distal tip (CS12) is close to the tricuspid valve annulus
AFW - The entire catheter rests against the atrial free wall
The figures in the figures subdirectory show approximate catheter placement in a model heart (click on any of the links below, for a larger view):

Anterior view of a heart model, with the interior of the right atrium (RA) and right ventricle (RV) exposed.
Catheter placement for SVC, where the distal tip is just inferior to the annulus of the superior vena cava, and the catheter is against the posterior wall of the atrium. Eight poles (black marks) of the decapolar catheter model (copper wire) are visible.
Placement for TVA, where the distal tip is on the tricuspid valve annulus and the electrodes are against the septal wall.
Placement for IVC, with the proximal catheter end just superior to the annulus of the inferior vena cava, and the electrodes against the posterior wall.
Placement for AFW, with the catheter against the atrial free wall (which is hinged open here).
The database includes a set of four records (one for each placement) for each of the eight patients. The name of each record identifies the patient (iaf1, iaf2, ..., iaf8) and the placement of the catheter (svc, ivc, tva, afw). Each record contains eight signals (intracardiac: CS12 - CS90, or ECG: I, II, V1, aVF). Each signal is sampled at 1 kHz with 14-bit resolution; the signal amplitudes are uncalibrated.

The atrial free wall (afw) records are considerably longer than those for the other regions, and include the effects of drug delivery and washout. During these records, patients iaf1, iaf2, iaf6, and iaf8 received adenosine, and patients iaf3, iaf4, iaf5, and iaf7 received Isuprel; the drug, dosage, and delivery time(s) are noted in the .hea files of the afw records. Note that these times are approximate, based on clinical notes taken during the procedure.

The .qrs annotation files were produced using sqrs, and have not been corrected; they may be useful for beginning a study of these recordings.

Acknowledgements

We wish to thank Jon Peterson and Dave Kynor of Creare, Inc. for their generous contribution of these recordings and for the notes and figures above.