|
3D Heart Simulation Image
Geometry and fiber structure obtained from diffusion tensor MRI (DTMRI) were converted to a computational domain over which partial differential equations governing the spread of current in the myocardium were solved. DTMRI data, giving fiber orientations at tens of thousands of locations within the ventricles, was used to set the three principle directions of the conductivity tensor. Conductivities in these directions were then scaled to give conduction velocities of 70 cm/sec in the longitudinal direction, and 25 cm/sec in the transverse direction. Transverse isotropy was assumed. An explicit Euler method (time step 2 uSec) was used to integrate through time, and the spatial derivatives were approximated with finite differences. The spatial step sizes were set to that of the anatomical model (156 x 312 x 469 m m). A Fitz-Hugh Nagumo model by Aliev and Panfilov was used to describe ionic currents. An epicardial region in the shape of an ellipse, with its long axis oriented approximately circumferentially was simulated by applying a super-threshold current pulse. The left panel shows epicardial potential 5 msec, and the right panel shows response 35 msec after the pulse. The epicardial surface has been rendered partially transparent, and all surfaces are color-coded by voltage (inset). The orange ribbon represents the 10mv iso-surface wave-front. Due to the faster conduction velocity along the fibers, which are oriented approximately longitudinally on the epicardium, the wave-front is seen to acquire a more longitudinal orientation as time progresses (right panel).

 |

|
|