Intracardiac Elasticity

Investigators:
Matthew O'Donnell, University of Michigan, Biomedical Engineering Department, Principal Investigator
Stanislav Y. Emelianov, University of Michigan, Biomedical Engineering Department,Co-Investigator
Andrei R. Skovoroda, Russian Academy of Sciences, Institute of Mathematical Problems in Biology, Co-Investigator
Graduate Students:
Xunchang Chen, University of Michigan, Biomedical Engineering Department, Graduate Research Assistant


Abstract

Strain rate images of the beating heart have been proposed to identify non-contracting regions of myocardium. Initial attempts used spatial derivatives of tissue velocity (Doppler) signals. We have proposed an alternate method based on two-dimensional speckle tracking applied to very high frame rate, real-time images. This processing can produce high resolution maps of the time derivative of the strain amplitude (i.e., square root of the strain intensity). Such images complement traditional tissue velocity images (TVI), providing a more complete description of cardiac mechanics. To test the proposed approach, strain rate images (SRI) were both simulated and measured on a thick-walled, cylindrical, tissue equivalent phantom modeling cardiac deformations. Real-time ultrasound images were captured during periodic phantom deformation, where the period was matched to the data capture rate of a commercial scanner mimicking high frame rate imaging of the heart. Simulation results show that strain rate images with spatial resolution between 1-2 mm are possible with a standard array system operating at 5 MHz. Moreover, these images are virtually free of angle dependent artifacts present in tissue velocity images and simple strain rate maps derived from these images. Measured results clearly show that phantom regions of low deformation difficult to identify on tissue velocity derived strain rate images are readily apparent with strain rate images generated from two-dimensional speckle tracking. Please see the movies - they are a click away below.

To produce images like this in the clinic, high quality two-dimensional, and ultimately three-dimensional, images must be produced at high frame rates. The propagation time of the mechanical wave in the left ventricle from base to apex during systole is, on average, about 55 msec for the adult human heart. To fully capture the details of wave propagation through the left ventricle using two-dimensional speckle tracking, this means full frame images of the ventricle must be acquired in 4-5 msec, representing frame rates of 200-250 Hz. This is certainly an achievable rate at low spatial resolution, but is at the edge of current technology for high spatial resolution images.

One possible solution to the frame rate problem for high resolution imaging is an intracardiac array. Intracardiac arrays can greatly reduce the propagation distance from the transducer to the cardiac wall, permitting high pulse repetition rates with concomitant high frame rates. Since higher operating frequencies can also be used compared to conventional echocardiography, these arrays may be able to produce very high resolution images. Coupling TVI and SRI derived from two-dimensional speckle tracking with these arrays is an exciting prospect for functional imaging of the heart.


Sample Movies



Publications




Correspondence:

Link to: http://bul.eecs.umich.edu/research/intracardiac/
Last modified by: Xunchang Chen