Session S21.1

Semi-Automated Segmentation and Registration of Triggered 3D Echocardiographic Images as a Basis for Volumetric Analysis of Myocardial Perfusion

F Veronesi*, V Mor-Avi, E Toledo, C Corsi, KA Collins,
G Lammertin, C Lamberti, RM Lang, EG Caiani

Università di Bologna
Bologna, Italy

Despite the potential of real-time 3D echocardiography (RT3DE) to assess myocardial perfusion, there is no quantification method available for perfusion analysis from RT3DE images. Such method would require 3D regions of interest (ROI) to be defined and adjusted frame-by-frame to compensate for cardiac translation and deformation. Our aims were to develop and test a technique for automated identification of 3D myocardial ROI suitable for translation-free quantification of myocardial videointensity over time, MVI(t), from contrast-enhanced RT3DE images.
Methods: 12 ECG-triggered transthoracic RT3DE (Philips) datasets obtained in pigs during transition from no contrast to steady-state enhancement (Definity) were analyzed using custom software. Initially, left ventricular (LV) endo- and epicardial surfaces were semi-automatically detected using level-set techniques in one frame to define a 3D myocardial ROI. Then, a two-step registration procedure was performed on each consecutive frame. First, rigid 3D transformation was performed by shifting and rotating each frame throughout the image sequence for optimal match between the position and orientation of the left ventricle in the current and the reference frames. Then, the displacement field representing motion between the two images was computed using an algorithm based on optical flow techniques. Using this displacement field, elastic 3D transformation was performed and each frame was warped by forcing the LV boundaries into their position in the reference frame. To assess the effectiveness of registration, MVI(t) in the 3D ROI was quantified from the registered and non-registered datasets. For each MVI(t) curve, we computed % variability during steady-state enhancement (100•SD/mean) and goodness of fit (r²) to the indicator dilution equation MVI(t)=A•(1-exp(-ßt)).
Results: Analysis of myocardial contrast throughout contrast inflow was feasible in all datasets. 3D registration improved MVI(t) curves in terms of both % variability (2.8±1.8% to 1.5±0.9%; p<0.05) and goodness of fit (r² from 0.79±0.2 to 0.90±0.1; p<0.05).
Conclusion: This is the first study to describe a new technique for semi-automated volumetric quantification of myocardial contrast from RT3DE images that includes registration and thus provides the basis for 3D measurement of myocardial perfusion.

(Abstract Control Number: 274)