|
|
Methods of the Spectroscopic I maging in clinical practice and experiments
Jírů, Filip ; Hájek, Milan (advisor) ; Horký, Jaroslav (referee) ; Mechl, Marek (referee)
Spectroscopic imaging (SI) is the method which enables non-invasive studying of the metabolite composition of tissues in vivo from multiple regions simultaneously. The objective of this thesis was to develop a methodology for the reliable evaluation of in vivo SI data. The thesis addresses several aspects of the evaluation of SI data as described below. The processing of SI data represents a complex issue requiring dedicated evaluation programs. Although various spectra processing programs are provided by the vendors of MR scanners, software enabling complete SI data processing and analysis is not available. The CULICH program has been developed within the framework of this thesis to enable the comprehensive processing of SI data. The program offers advanced functionality for evaluation of SI data. The initial experience with the program shows that CULICH is suitable for evaluation of SI data measured in both clinical practice and the experiments. The accuracy of calculated concentrations is of high importance for each quantification method. SVS techniques can be taken as the gold standard for quantification of MR spectra. Therefore, to asses the accuracy of the metabolite concentrations measured by SI, the comparison of results obtained by SI and SVS was performed. The direct comparison of results obtained...
|
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
Some comments on frequency selective excitation in newly proposed MRSI sequences
Starčuk jr., Zenon ; Horký, Jaroslav ; Starčuk, Zenon ; Mlynárik, V. ; Gruber, S. ; Moser, E.
In many pathologies it is desirable to compare the metabolism inside a lesion, near the lesion, and in healthy tissue. The diagnostic value of MRS is enhanced by spectroscopic imaging (MRSI), permitting the simultaneous acquisition of spatially resolved spectra. In this way, assessing metabolic information about different brain regions within a single examination is possible. As with all in vivo spectroscopy, low signal-to-noise ratio (SNR) is the fundamental limiting factor in MRSI due to very low concentrations of metabolites of interest. Proton spectroscopic imaging, however, offers additional technical challenges as compared to single-voxel techniques, especially if acquisition of short echo time (TE<30 ms) is required. Critical to the success of proton MRSI studies of the human brain is the elimination of the very intense water and lipid signals arising from outside the volume of interest (VOI). Water suppression in MRSI can be very problematic in MRSI, in which the achievable degree of water suppression is limited by B.sub.0./sub. and B.sub.1./sub. inhomogeneities and water T.sub.1./sub. variations, invariably present throughout the larger VOIs. Suppression of liquid signals (from bone marrow and subcutaneous fat) is often severely complicated due to the fact that their relaxation behavior differs substantially from that of water. One of the basic approaches used to reduce the undesired contamination if the MRSI signals of interest consists in the use of the STEAM or PRESS selective excitation of the VOI. Both STEAM and PRESS techniques suffer from some drawbacks. STEAM reduces the sensitivity of measurement, PRESS has higher RF power requirements.
|
|
|
High resolution multivoxel spectroscopy of human brain at 3 Tesla
Mlynárik, V. ; Gruber, S. ; Starčuk, Zenon ; Starčuk jr., Zenon ; Horký, Jaroslav ; Moser, E.
Localisation in in vivo NMR spectroscopy can be achieved using different concepts. Single voxel localized spectroscopy exploits three slice selective pulses defining a cube (or a parallelepiped) in their intersection. By proper combination of spoiling gradients the excited magnetization outside the cube is dephased and does not contribute to the NMR signal. Another way of localisation is combination of slice selection with phase encoding of the NMR signal in the other two dimensions by means of gradient magnetic fields. The latter method is referred to as spectroscopic imaging and provides an array of spectra corresponding to individual voxels defined by the phase encoding. Further subdivision of the voxels is possible by Hadamard encoding of the excitation pulses.
|
|
| |
|
| |
guest ::
