Development of Efficient Dynamic Magnetic Resonance Imaging Methods with Application to Breast Cancer Detection and Diagnosis
Abstract
The goal of this predoctoral fellowship research project is to improve the temporal and spatial resolutions in dynamic contrast-enhanced magnetic resonance imaging of the breast. Specifically, we investigated the use of non-Fourier encoding for collecting the reduced encoding dynamic data sets. The conclusion from our study was that the current wavelet and SVD encoding methods do not achieve the desired goal of exploiting the desirable truncation properties of these basis functions and, at times, can introduce dangerous artifacts. For Fourier-encoded data, the Reduced-encoding Imaging by Generalized-series Reconstruction (RIGR) method was shown to be a better way to extrapolate the unmeasured dynamic data using a priori constraints than Fourier-keyhole. Next, we further optimized the basis functions of the generalized-series (GS) model used by the RIGR algorithm. The Two-reference RIGR (TRIGR) method resulted from suppressing background information in the GS basis functions, enabling them to better represent the regions of dynamic change. Explicit edge constraints derived from the reference image were then used with contrast information from the dynamic data to inject dynamic information into the GS basis functions. Finally, the problem of motion in dynamic imaging was addressed using a similarity norm to accurately detect the motion in spite of contrast changes and the low-resolution nature of the dynamic data.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1997
- Accession Number
- ADA333252
Entities
People
- Jill M. Hanson
- Paul Lauterbur
Organizations
- University of Illinois Urbana–Champaign