BACKGROUND: We have elaborated a new software called ACS designed for left ventricle 3-dimensional reconstruction and quantitative analysis of regional wall motion using gated blood-pool tomography (GBPS). This paper presents the principle and the validation of this software. METHODS AND RESULTS: Blood pool tomography was reconstructed by filtered back-projection. Voxels corresponding to ventricular volumes were extracted using 3-D thresholding and 3-D elliptic mask. The auriculoventricular plane was located by phase analysis. Then, the reconstruction of the 3-D volume was initiated. First, a pre-processing based on a bi-dimensional Fourier transform re-sampled the initial points representing the wall of the left ventricle. Left ventricular contraction could be visualized in three dimensions. The points corresponding to the wall of the left ventricle were reorganized in order to obtain a regular sampling that allowed connection to the points during the cardiac cycle. Elementary volumes were defined by three adjacent wall points and the centre of the whole volume. The variation of elementary volumes during the cardiac cycle was related to the wall motion. Elementary volumes were regrouped to create regions of equivalent volume. Finally, regional and global ejection fractions could be calculated in 17 different regions. The method has been validated using a mathematical deformable ellipsoidal model. It had also been tested in vivo on a set of 59 patients in comparison to equilibrium radionuclide angiography for left ventricular ejection fraction and to echocardiography for regional wall motion evaluation. CONCLUSIONS: ACS is a new software that allows reconstruction and visualization of the left ventricle in three dimensions, and it quantified 3-dimensional regional wall motion analysis.
BACKGROUND: We have elaborated a new software called ACS designed for left ventricle 3-dimensional reconstruction and quantitative analysis of regional wall motion using gated blood-pool tomography (GBPS). This paper presents the principle and the validation of this software. METHODS AND RESULTS: Blood pool tomography was reconstructed by filtered back-projection. Voxels corresponding to ventricular volumes were extracted using 3-D thresholding and 3-D elliptic mask. The auriculoventricular plane was located by phase analysis. Then, the reconstruction of the 3-D volume was initiated. First, a pre-processing based on a bi-dimensional Fourier transform re-sampled the initial points representing the wall of the left ventricle. Left ventricular contraction could be visualized in three dimensions. The points corresponding to the wall of the left ventricle were reorganized in order to obtain a regular sampling that allowed connection to the points during the cardiac cycle. Elementary volumes were defined by three adjacent wall points and the centre of the whole volume. The variation of elementary volumes during the cardiac cycle was related to the wall motion. Elementary volumes were regrouped to create regions of equivalent volume. Finally, regional and global ejection fractions could be calculated in 17 different regions. The method has been validated using a mathematical deformable ellipsoidal model. It had also been tested in vivo on a set of 59 patients in comparison to equilibrium radionuclide angiography for left ventricular ejection fraction and to echocardiography for regional wall motion evaluation. CONCLUSIONS:ACS is a new software that allows reconstruction and visualization of the left ventricle in three dimensions, and it quantified 3-dimensional regional wall motion analysis.
Authors: François Harel; Vincent Finnerty; Quam Ngo; Jean Grégoire; Paul Khairy; Bernard Thibault Journal: J Nucl Cardiol Date: 2007-06-27 Impact factor: 5.952
Authors: Marcus Hacker; Xaver Hoyer; Sandra Kupzyk; Christian La Fougere; Johann Kois; Hans-Ulrich Stempfle; Reinhold Tiling; Klaus Hahn; Stefan Störk Journal: Int J Cardiovasc Imaging Date: 2005-11-22 Impact factor: 2.357