Going beyond the borders : field-of-view extension in MR/PET hybrid imaging

The recent combination of Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET) is of emerging interest in clinical routine. On the one hand, MRI is a widely-used diagnostic tool in medical practice. The excellent soft-tissue contrast offers reliable anatomical information. On the other hand, PET is a key imaging technique in nuclear medicine. It shows the functional metabolism thus providing information on the biochemical and physiological processes of the patient. The recently developed MR/PET hybrid system combines these advantages in one whole-body system.
The integration of the PET detector in an MR system is a new challenge. In particular, new concepts for the attenuation correction of the PET data are
required. This attenuation correction can be performed based on the MR data. However, the MR field-of-view (FoV) is limited by intrinsic physical restrictions such as B0 inhomogeneities and gradient nonlinearities. It has been reported that the PET quantification is biased due to the limitation of the MR-based FoV. Thus, new acquisition techniques are required for imaging beyond the FoV limitation. This was addressed in this thesis.
A novel method was developed to compensate the distortions arising from the gradient nonlinearity and the B0 inhomogeneity by using an optimal gradient amplitude. This imaging technique was implemented in a spin-echo-based sequence featuring multi-slice acquisition, bipolar readout and continuous table movement. Thereby, an extension of the FoV was achieved. The optimal sequence parameters were determined inline and applied automatically. No further user interaction or hardware modifications were required. The additional acquisition time is easily tolerable for clinical routine.
In a patient study the improvement achieved by the proposed method was verified. The reported bias in the PET reconstruction was reduced significantly. Therefore, the proposed technique improves the MR-based attenuation correction of the PET emission data. Furthermore, the achieved FoV extension might be of interest for diverse MR applications such as image-guided therapy or measuring of large patients. In summary a method has been developed that improves the PET quantification in MR/PET hybrid imaging.