Charles A. Mistretta, Ph.D.
Professor
Contact Information
- Office: J3 119, 600 Highland Ave., Madison WI 53726
- Phone: (608) 265-9685
- Email:
camistre@wisc.edu
Work and Education
-
Education
- B.S., Engineering Physics, University of Illinois.
- Ph.D., High Energy Physics, Harvard, 1968.
-
Responsibilities:
- Co-Director of the Diagnostic CT Lab.
- Research Professor in the UW MR Angio Group.
- Instructor of graduate course in Diagnostic Radiology.
- Vice Chairman for Research - Dept. of Medical Physics
-
Current Work & Research: Vastly undersampled
data acquisition to achieve order of magnitude increases in the product
of spatial resolution, coverage and acquisition time in MR and X-ray CT.
- Curriculum Vitae:
[PDF]
Awards
- Tau Beta Pi, Phi Kappa Phi with Highest Honors
University of Illinois.
- James Picker Advanced Fellow
Academic Radiology.
- Laufman-Greatbatch Prize (for DSA)
Association for the Advancement of Medical Instrumentation.
- J. Allyn Taylor International Prize in Medicine (for distinguished lifetime achievement)
Sponsored by Robarts Research Institute.
- John R. Cameron Professor
University of Wisconsin Medical Physics Department.
- AAPM Fellow
American Association of Physicists in Medicine.
- AIMBE Fellow
American Institute of Medical and Biological Engineering.
|
|
The images obtained with MRI depend on how we travel through k-space during data acquisition. The most common k-space trajectory is a Cartesian trajectory in which spatial resolution is proportional to the number of phase encoding lines and therefore proportional to imaging time. This imposes a serious speed penalty for high resolution imaging. If we return to the radial acquisition technique first used by Lauterbur to reconstruct MRI images, each excitation produces a projection passing through the center of k-space. Although a fully sampled radial acquisition takes 1.5 times longer than the Cartesian acquisition, if one undersamples by obtaining fewer projections than demanded by the Nyquist theorem, full resolution is almost immediately attained at the expense of steak artifacts. In some applications where the data space is sparse, such as in phase contrast angiography, a technique called vastly undersampled Isotropic projection imaging (VIPR) has been used to increase the product of scan volume, scan speed and spatial resolution by a factor of 30-60 relative to commercially available 3D phase contrast imaging. When the undersampling is done in 3D, artifacts disperse favorably and ultimately become limiting only through decreased SNR . Vastly undersampled acquisition also appears to be a feasible approach for cone beam X-ray CT where simulations suggest that with a total of 300 focal spot positions an entire cardiac volume can be imaged with a potentially robust gated technique providing a temporal resolution of about 15 ms.