doi:10.1016/j.ultrasmedbio.2005.04.016 ● Original Contribution DOPPLER ECHOCARDIOGRAPHY FLOW-VELOCITY IMAGE ANALYSIS FOR PATIENTS WITH ATRIAL FIBRILLATION HAYIT GREENSPAN,* ORON SHECHNER,* MICKEY SCHEINOWITZ,* † and MICHA S. FEINBERG ‡ *Department of Biomedical Engineering, † Neufeld Cardiac Research Institute and ‡ The Heart Institute, Sheba Medical Center, Tel Hashomer, Tel-Aviv University, Tel-Aviv, Israel (Received 21 August 2004; revised 14 April 2005; in final form 21 April 2005) Abstract—Currently, Doppler echocardiography analysis is performed manually. An automated method that analyzes the Doppler signal can potentially improve accuracy and result in a powerful tool for noninvasive evaluation of cardiac hemodynamics, especially for patients with atrial fibrillation, where multiple samples are needed to obtain an accurate averaged measurement. The aim of this study was to develop an automated method for Doppler analysis based on image processing and computer vision algorithms. Images were obtained from the mitral valve and the tricuspid valve Doppler tracings from 45 patients, 20 with normal sinus rhythm and 25 with atrial fibrillation. The proposed algorithm automatically detects the maximal velocity envelope of the spectral Doppler ultrasound tracings. Averaged values for the time velocity integral, peak mitral inflow velocity and peak tricuspid regurgitation velocity were calculated for multiple beats available in a single screen frame. Measure- ments extracted automatically from the maximal velocity envelope were compared to measurements obtained manually by two expert technicians. High linear correlation (r) was found between the automatically- and the manually-extracted parameters (0.95 < r < 0.99). A smaller variation was found in most cases between the manually-calculated average beat and the automated average beat (bias value between 3.8% and 5.2%) than between the manually-calculated average beat and the selection of a representative beat (bias value between 6.2% and 2.6%). The newly-developed automated method offers a new, accurate and reliable clinical tool, particu- larly for the assessment of patients with irregular heart rate. (E-mail: hayit@eng.tau.ac.il) © 2005 World Federation for Ultrasound in Medicine & Biology. Key Words: Doppler echocardiography, Blood flow velocity, Maximal velocity envelope, Image processing, Edge detection, Automated system, Atrial fibrillation. INTRODUCTION Cardiac Doppler blood flow tracings are a valuable tool for the noninvasive assessment of cardiac function (Om- men and Nishimura 2003; Troughton et al. 2003). The Doppler signals are subject to great variation between subjects, changes related to flow and heart rate and within subjects with irregular heart rates. Manual tracing of the Doppler signals is used to extract the maximal velocity envelope (MVE) and corresponding parameters. This type of approach is time-consuming and sometimes has limited reproducibility (Galderisi et al. 1992). These shortcomings worsen when dealing with signals taken from patients with irregular heart rates, such as in atrial fibrillation (AF). In these patients, subjectively defined “representative” beats are used to extract specific param- eters; thus, increasing bias in the analysis process. The aim of this study was to develop an automated method for Doppler analysis based on image processing and computer vision algorithms that would reduce those lim- itations. Studies that deal with the automatic detection of Doppler MVEs are mostly based on some form of noise reduction and an edge-following algorithm, followed by parametric model estimation. An overview of several such works has been reported (Doherty et al. 2002). Beat-by-beat analysis of Doppler tracings for the case of the brachial artery was conducted by Tschirren et al. (2000, 2001). Noise reduction as an initial preprocessing stage was used, followed by advanced edge-detection methodologies to extract the MVE and a final fitting of the extracted waveform to a partial Fourier series model. A more simplified edge-detection methodology, along with averaging of the extracted MVEs, was used by Hall and Kovacs (1994) as a means for noise removal, as well Address correspondence to: Dr. H. Greenspan, Department of Biomedical Engineering, Tel-Aviv University, Ramat-Aviv 69978 Is- rael. E-mail: hayit@eng.tau.ac.il Ultrasound in Med. & Biol., Vol. 31, No. 8, pp. 1031–1040, 2005 Copyright © 2005 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 0301-5629/05/$–see front matter 1031