Rochester Institute of Technology
Collection and analysis of multispectral magnetic resonance imaging data using a clinical imaging system
Abstract
dc:description.abstractAs most radiologists would attest, the accuracy of pathology diagnosis is greatly increased by acquiring more information about the tissue in question. An innovative data acquisition technique has been developed that will do just that - gather more information about the tissues being studied. As each tissue type is composed of a unique arrangement of chemical components, each tissue exhibits a characteristic nuclear magnetic resonance (NMR) frequency spectrum. Unfortunately, current clinical magnetic resonance imaging (MRI) systems integrate the characteristic spectra of the tissue within a voxel to create a single signal for the voxel, thus throwing away what could be invaluable information. In an effort to regain the typically discarded information, a spatial-spatial-spectral imaging technique has been developed and demonstrated using a clinical MRI system. The technique, which was originally developed and demonstrated by Lauterbur on a NMR system, involves acquiring projections through the spatial-spectral plane of the spatial-spatial-spectral volume and reconstructing that plane through use of a back-projection procedure. This technique is made possible on clinical MRI systems through the advent of variable bandwidth imaging - a technological advance that allows for a variation in the sampling bandwidth used in the acquisition steps of MRI. Combined with specific variations of the frequency-encoding gradient, specific bandwidth values can be used with existing pulse sequences to acquire projections through the spatial-spectral plane at very particular angles. The technique has been demonstrated on a 1.5T, whole-body, clinical imaging system (Signa, GE Medical Systems, Milwaukee, WI) using chemical shift phantoms as well as in vivo tissue. In addition, the accuracy of the technique has been analyzed via point spread function (PSF) estimation and classification accuracy techniques. The impacts on these analysis metrics due to variations in acquisition and processing parameters have also been investigated.
Degree
thesis:*- Name thesis:degree_name
- Imaging Science (Ph.D.)
- Level thesis:degree_level
- Dissertation
- Discipline thesis:degree_discipline
- Chester F. Carlson Center for Imaging Science (COS)
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Servoss, Thomas G.
- Contributors dc:contributor
-
- Hornak, Joseph
Subjects
dc:subject × 2Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.rit.edu/theses/2886