Implementation of Keyhole Method for Dynamic Magnetic Resonance Imaging
No Thumbnail Available
Helsinki University of Technology | Diplomityö
AbstractThe purpose of this work was to implement a keyhole dynamic imaging method on the Philips Panorama open mid-field and low-field magnetic resonance imaging (MRI) systems. The keyhole method is an established method to reduce imaging time and increase temporal resolution with contrast enhanced (CE) dynamic imaging sequences. In keyhole imaging, a high resolution image is acquired at the start of imaging, where after only a restricted part of the image data is updated. The partial data is then appended with the reference data to produce an apparent high-resolution image. This is possible because the data describing contrast changes in larger areas, without sharp boundaries, are localized in a limited part of the phase and frequency-encoded data. Keyhole imaging has recently been developed for use with a contrast enhanced timing robust angiography (CENTRA) encoding scheme. The keyhole method was implemented for normal linear encoding orders as well as for CENTRA as part of this thesis. The functionality of the keyhole method was tested by staging a CE dynamic imaging session. A grape fruit was imaged before and during injection of a contrast agent into a cavity in the fruit. A keyhole percentage of 25% was used, which reduced the imaging time to a fourth of the original at the same time increasing the temporal resolution by a factor of four. This produced nice high-resolution images, without any visible artefacts as expected. The signal to noise ratio (SNR) and imaging time reduction were measured for a few different sequences and compared with the reference images to control that no difference in SNR could be seen and that the imaging time was correctly reduced. The robustness of the reconstruction was tested with a very fast dynamic sequence and a small keyhole size. The tests passed without problems. Keyhole imaging was also tested with a patient without contrast agents to control the reproducibility of patient anatomies when only a part of the image data was updated. The images showed that when the patient did not move during the scan, the anatomies were reproduced without problems. On the other hand, if the patient moved during the scan, the images were partially or entirely spoiled because of motion artefacts. Still, the edges of the reference image could be visualized as expected.
magnetic resonance imaging, magnetresonansbildtagning, keyhole imaging, keyhole metod, restricted k-space, begränsat k-rum, contrast enhanced timing robust angiography, kontrastförstärkt timing robust angiografi