Fast and Ultrafast Magnetic Resonance Diffusion-Weighted Imaging in the Presence of Physiologic Motion

Roland Bammer

Publikation: StudienabschlussarbeitDissertation


Magnetic resonance (MR) diffusion-weighted imaging (DWI) allows to study proton self-diffusion and has a major impact on the diagnosis of acute cerebral stroke. However, DWI pulse sequences are inherently motion sensitive, resulting in severe artifacts. Consequently, the influence of physiologic motion on different DWI techniques was studied along with pulse sequence-specific artifacts. Theoretical analysis as well as in-vivo results showed that interleaved EPI (IEPI) with navigator echo correction is an excellent tool for fast high-resolution imaging of acute stroke, which is also applicable on conventional MR systems. In contrast to single-shot EPI, the images are less degraded by off-resonance effects and blurring. For arbitrary diffusion-weighting directions, the navigator echo must be replaced by either a navigator k-space map or by a less motion-sensitive sequence such as single-shot EPI. In this context, a novel scanning method is presented that relies on parallel acquisition of MR signals by means of an array of different receiver coils along with simultaneous k-space undersampling. Subsequent image reconstruction is carried out by using the coil sensitivity of the individual receiver coils. This method combines the advantages of both single-and multi-shot EPI. These advantages are: ultrafast data acquisition, remedies against artifacts from bulk motion, and reduced image blurring and geometric distortions.
QualifikationDoktor der Technik
Gradverleihende Hochschule
  • Technische Universität Graz (90000)
Betreuer/-in / Berater/-in
  • Wach, Paul, Betreuer
  • Stollberger, Rudolf, Betreuer
PublikationsstatusVeröffentlicht - 25 Jan 2000
Extern publiziertJa


Untersuchen Sie die Forschungsthemen von „Fast and Ultrafast Magnetic Resonance Diffusion-Weighted Imaging in the Presence of Physiologic Motion“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren