Abstract
dc:descriptionA new step-by-step comprehensive MR physics simulator (MRISIMUL) of the Bloch equations ispresented in this work. The aim was to develop a magnetic resonance imaging (MRI) simulator thatmakes no assumptions with respect to the underlying pulse sequence and also allows for complexlarge-scale analysis on a single-node computer without requiring simplifications of the MRI model.We hypothesized that such a simulation platform could be developed with parallel acceleration of theexecutable core within the graphic processing unit (GPU) environment. MRISIMUL integrates realisticaspects of the MRI experiment from signal generation to image formation and solves the entirecomplex problem for densely spaced isochromats and for a densely spaced time axis. The simulationplatform was developed in MATLAB whereas the computationally demanding core services weredeveloped in CUDA-C. The MRISIMUL simulator imaged three different computer models: a userdefinedphantom, a human brain model and a human heart model whereas three different motionmodels were introduced as well: cardiac motion, respiratory motion and flow.The high computational power of GPU-based simulations was compared against other computerconfigurations. A speedup of about 228 times was achieved when compared to serially executed Ccodeon the CPU whereas a speedup between 31 to 115 times was achieved when compared to theOpenMP parallel executed C-code on the CPU, depending on the number of threads used inmultithreading (2–8 threads). Moreover, simulations of motion with MRISIMUL on single-node andmulti-node multi-GPU systems were also examined demonstrating an almost linear scalableperformance with the increasing number of available GPU cards, in both single-node and multi-nodemulti-GPU computer systems.MRISIMUL is the first MR physics simulator that allows for computationally intense realistic simulationson multi-GPU computer systems. The high performance of MRISIMUL can bring the computationalpower of a supercomputer or a large computer cluster to a single GPU personal computer. Theincorporation of realistic aspects of MR experiments, such as motion, may benefit the design andoptimization of existing or new MR pulse sequences, protocols and algorithms.
Degree
thesis:*- Grantor dc:publisher
- University of Thessaly (UTH)
- Year dc:date
- 2010
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Xanthis, Christos
- Ξάνθης, Χρήστος
Subjects
dc:subject × 21- Προσομοιωτής
- Μαγνητική τομογραφία
- Συστήματα υπολογισμού υψηλών επιδόσεων
- Ανατομικά μοντέλα κίνησης
- Simulation
- Magnetic resonance imaging
- MRI
- High performance computing
- Parallel coputing
- CUDA
- Gpu computing
- GPGPU
- Gpu card
- Anatomical motion models
- Matlab
- Φυσικές Επιστήμες
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Φυσική
- Natural Sciences
- Computer and Information Sciences
- Physical Sciences
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/37245