Meeting Banner
Abstract #2918

Compressed Sensing for Accelerated MR Thermometry in MRI-Controlled Transurethral Ultrasound Therapy

Patrick Leonard1, 2, Rajiv Chopra, 12, Adrian Nachman3, 4

1Medical Biophysics, University of Toronto, Toronto, Ontario, Canada; 2Sunnybrook Research Institute, Toronto, Ontario, Canada; 3Department of Electrical Engineering, Universisty of Toronto; 4Department of Mathematics, Universisty of Toronto

In the current implementation of MRI-controlled transurethral ultrasound therapy the 3D spatial heating pattern is approximated by a series of 2D slices acquired every 5 7 seconds. Knowledge of the true 3D temperature distribution could allow for more sophisticated temperature control algorithms and improve the monitoring of nearby tissues for safety. In this work we examine the use of compressed sensing as a means of accelerating the data acquisition time, where preserving the quantitative accuracy of the temperature measurements is critical. Our results show that our treatment remains safe and effective at up to 32-fold acceleration rates.

Keywords

ability accelerated acceleration accuracy accurate acquisition adversely affected aliasing allow alone applicator approximately artifacts basis blurring boundary broad certain coagulating coagulation compatible compressed comprised compromising control controlled controller critical currently degradation degraded delivery depictions depicts described developed device distribution effective electrical enable energy ensuring equivalent error evaluated example expected experiment extracted feedback field frequency fully generated good heating impact impacted implemented important importantly improvement indicate introduced invasive inversion long magnitude maps mathematics measured measures methodology minimal minimally mistreated noise noted operation output overall pattern phantom phys positioning power produced prostate quantitative radial real reconstructed reconstruction reduced reducing reduction relatively remain remains removal report response reveals robustness rotated rotation sampled sampling scanner scheme sensing significantly similarly simulated simulation simulations slices software space spatial spokes stability statistics superimposed system table target technology temperature tendency therapy threshold tissue tissues towards treat treated treatment trend ultrasound uncertainty underestimation variable volume volumes warrant wavelet whereby zero