Record:   Prev Next
Author Frakes, David Harold
Title An adaptive control grid interpolation technique for the three-dimensional reconstruction of MRI data
book jacket
Descript 328 p
Note Source: Dissertation Abstracts International, Volume: 64-06, Section: B, page: 2775
Director: Ajit P. Yoganathan
Thesis (Ph.D.)--Georgia Institute of Technology, 2003
The accurate representation of two-dimensional magnetic resonance (MR) images as three-dimensional information is an important issue for medial applications aimed at visualization, analysis, and diagnosis. Computer-based reconstruction methods applied after data acquisition have the potential to provide such three-dimensional information with high quality and at little or no additional cost. Toward this end a novel technique for the three-dimensional reconstruction of MR data has been developed
The new methodology presented in this thesis approaches the reconstruction problem from a different perspective, using motion estimation to approximate data not captured by scanning. This is accomplished with adaptive control grid interpolation (ACGI), a hybrid motion estimation scheme that displays advantageous characteristics of both block matching and optical flow. The mechanics of this technique make it well suited for the reconstruction of both structural and velocity data
ACGI reconstruction is presented here in the context of palliative surgical procedures performed on children with single ventricle congenital heart defects. Specific ACGI implementations designed to facilitate visualization and power loss estimation are developed as components of a comprehensive surgical planning and evaluation tool. Additional applications for tracking and video enhancement are also explored
School code: 0078
Host Item Dissertation Abstracts International 64-06B
Subject Engineering, Biomedical
Health Sciences, Radiology
Alt Author Georgia Institute of Technology
Record:   Prev Next