|
|
Image reconstruction from non-Cartesian k-space data acquired by ultra-short echo-time and fast MR imaging methods
Pšorn, Tomáš ; Latta,, Peter (oponent) ; Starčuk, Zenon (vedoucí práce)
The aim of this thesis is to handle theoretically the technique of MRI image reconstruction from non-Cartesian data such as radial, spiral, etc. Since later progress of this thesis will include actual measuring of non-Cartesian data, which is prone to the gradient infidelity, methods for measuring a performance of the gradient system will also be discussed. In practical part I will present the way the 4.7 T and 9.4~T MRI systems of the Institute of Scientific Instruments of the ASCR, v. v. i., works and use the existing pulse sequences to measure some model data.
|
|
|
MR zobrazování fluoru-19
Meščánková, Veronika ; Latta,, Peter (oponent) ; Starčuk, Zenon (vedoucí práce)
Táto práca sa zaoberá MR zobrazovaním fluóru-19. Jej cieľom je pripraviť fantómy a biologické vzorky pre testovanie citlivosti detekcie 19F, optimalizovať metódy zobrazenia 19F na predklinickom MR systéme Bruker Biospec 94/30, preskúmať praktické možnosti zobrazenia 19F na pripravených vzorkách a porovnať citlivosť s protónovými obrazmi. Na záver je nutné zhodnotiť dosiahnuté výsledky a možné prínosy metódy.
|
|
|
Implementation of Dixon Methods for Preclinical MR Imaging at High Fields
Kořínek, Radim ; Latta,, Peter (oponent) ; Puková,, Andrea Šprláková - (oponent) ; Bartušek, Karel (vedoucí práce)
Preclinical magnetic resonance (MR) imaging in small animals is a very popular procedure that requires a higher sensitivity, given the small size of the subjects. A higher sensitivity can be reached when an MR imaging system with a high magnetic field is used (e.g., 4.7 T or higher). The benefits of such sensitivity include, for example, a higher resolution, an improved signal-to-noise ratio (SNR), an increased chemical shift, and a longer T1 longitudinal relaxation time. On the other hand, a high field causes stronger static magnetic field deformation along the borders between tissues with different susceptibilities, and it also results in the shortening of the T2 transversal relaxation. Adipose tissue is significantly contained in the human (or mammal) body and is primarily used to store energy in the form of fat. This tissue can be classified into white and brown subsets. Brown adipose tissue is found mainly in new-born children, and a certain (yet very small) amount of such tissue can be traced also in adults. White adipose tissue then ensures the storage of fat as a source of energy. Furthermore, white adipose tissue produces adipokines, hormones, and many other substances important for metabolism. Generally, fat can be regarded as a biomarker in the case of specific diseases (obesity, steatosis – fatty liver disease, and others). Thus, the quantification of fat is a precondition for correct diagnosis. MR imaging comprises a special group of methods for water-fat separation; these methods are referred to as Dixon methods and utilize the principle of chemical shift. In this thesis, a new T2 – weighted sequence for Dixon acquisition is introduced (Chapter 5.3). The proposed sequence is a very time-effective three-point (3PD) method. The newly proposed sequence of fast triple spin echo Dixon (FTSED) is derived from the original fast spin echo sequence (FSE). Such modification of the original FSE sequence leads to a novel FTSED sequence, where three images are acquired simultaneously without any increase of the total acquisition time. The discussed sequence was successfully implemented on a 9.4 T MR imaging system at the Institute of Scientific Instruments, ASCR Brno. The acquired data were calculated through the use of the IDEAL (iterative decomposition of water and fat with echo asymmetry and least-squares estimation) algorithm. The results of the computation are water and fat images, and the fat fraction (FF) can be calculated from these. The sequence was successfully tested in a rat. The successful FTSED implementation on a 9.4 T MR imaging system enables this method to be used in low-field MR imaging systems.
|
|
|
DEEP LEARNING FOR SINGLE-VOXEL AND MULTIDIMENSIONAL MR-SPECTROSCOPIC SIGNAL QUANTIFICATION, AND ITS COMPARISON WITH NONLINEAR LEAST-SQUARES FITTING
Shamaei, Amir Mohammad ; Latta,, Peter (oponent) ; Kozubek, Michal (oponent) ; Jiřík, Radovan (vedoucí práce)
Preprocessing, analysis, and quantification of Magnetic resonance spectroscopy (MRS) signals are required for obtaining the metabolite concentrations of the tissue under investigation. However, a fast, accurate, and efficient post-acquisition workflow (preprocessing, analysis, and quantification) of MRS is challenging. This thesis introduces novel deep learning (DL)-based approaches for preprocessing, analysis, and quantification of MRS data. The proposed methods achieved the objectives of robust data preprocessing, fast and efficient MR spectra quantification, in-vivo concentration quantification, and the uncertainty estimation of quantification. The results showed that the proposed approaches significantly improved the speed of MRS signal preprocessing and quantification in a self-supervised manner. Our proposed methods showed comparable results with the traditional methods in terms of accuracy. Furthermore, a standard data format was introduced to facilitate data sharing among research groups for artificial intelligence applications. The findings of this study suggest that the proposed DL-based approaches have the potential to improve the accuracy and efficiency of MRS for medical diagnosis. The dissertation is structured into four parts: an introduction, a review of state-of-the-art research, a summary of the aims and objectives, and a collection of publications that showcase the author's contribution to the field of DL applications in MRS.
|
|
|
Image reconstruction from non-Cartesian k-space data acquired by ultra-short echo-time and fast MR imaging methods
Pšorn, Tomáš ; Latta,, Peter (oponent) ; Starčuk, Zenon (vedoucí práce)
The aim of this thesis is to handle theoretically the technique of MRI image reconstruction from non-Cartesian data such as radial, spiral, etc. Since later progress of this thesis will include actual measuring of non-Cartesian data, which is prone to the gradient infidelity, methods for measuring a performance of the gradient system will also be discussed. In practical part I will present the way the 4.7 T and 9.4~T MRI systems of the Institute of Scientific Instruments of the ASCR, v. v. i., works and use the existing pulse sequences to measure some model data.
|
|
|
MR zobrazování fluoru-19
Meščánková, Veronika ; Latta,, Peter (oponent) ; Starčuk, Zenon (vedoucí práce)
Táto práca sa zaoberá MR zobrazovaním fluóru-19. Jej cieľom je pripraviť fantómy a biologické vzorky pre testovanie citlivosti detekcie 19F, optimalizovať metódy zobrazenia 19F na predklinickom MR systéme Bruker Biospec 94/30, preskúmať praktické možnosti zobrazenia 19F na pripravených vzorkách a porovnať citlivosť s protónovými obrazmi. Na záver je nutné zhodnotiť dosiahnuté výsledky a možné prínosy metódy.
|
|
|
Implementation of Dixon Methods for Preclinical MR Imaging at High Fields
Kořínek, Radim ; Latta,, Peter (oponent) ; Puková,, Andrea Šprláková - (oponent) ; Bartušek, Karel (vedoucí práce)
Preclinical magnetic resonance (MR) imaging in small animals is a very popular procedure that requires a higher sensitivity, given the small size of the subjects. A higher sensitivity can be reached when an MR imaging system with a high magnetic field is used (e.g., 4.7 T or higher). The benefits of such sensitivity include, for example, a higher resolution, an improved signal-to-noise ratio (SNR), an increased chemical shift, and a longer T1 longitudinal relaxation time. On the other hand, a high field causes stronger static magnetic field deformation along the borders between tissues with different susceptibilities, and it also results in the shortening of the T2 transversal relaxation. Adipose tissue is significantly contained in the human (or mammal) body and is primarily used to store energy in the form of fat. This tissue can be classified into white and brown subsets. Brown adipose tissue is found mainly in new-born children, and a certain (yet very small) amount of such tissue can be traced also in adults. White adipose tissue then ensures the storage of fat as a source of energy. Furthermore, white adipose tissue produces adipokines, hormones, and many other substances important for metabolism. Generally, fat can be regarded as a biomarker in the case of specific diseases (obesity, steatosis – fatty liver disease, and others). Thus, the quantification of fat is a precondition for correct diagnosis. MR imaging comprises a special group of methods for water-fat separation; these methods are referred to as Dixon methods and utilize the principle of chemical shift. In this thesis, a new T2 – weighted sequence for Dixon acquisition is introduced (Chapter 5.3). The proposed sequence is a very time-effective three-point (3PD) method. The newly proposed sequence of fast triple spin echo Dixon (FTSED) is derived from the original fast spin echo sequence (FSE). Such modification of the original FSE sequence leads to a novel FTSED sequence, where three images are acquired simultaneously without any increase of the total acquisition time. The discussed sequence was successfully implemented on a 9.4 T MR imaging system at the Institute of Scientific Instruments, ASCR Brno. The acquired data were calculated through the use of the IDEAL (iterative decomposition of water and fat with echo asymmetry and least-squares estimation) algorithm. The results of the computation are water and fat images, and the fat fraction (FF) can be calculated from these. The sequence was successfully tested in a rat. The successful FTSED implementation on a 9.4 T MR imaging system enables this method to be used in low-field MR imaging systems.
|
host ::
RSS.