Recently, systems that integrate positron emission tomography and magnetic resonance imaging (PET/MR) have become available for clinical use. This new technology, which combines the high spatial resolution and superior soft-tissue contrast of MR with the picomolar sensitivity, quantitative capabilities, and wide array of tracers of PET, has the potential to benefit patients and provide insights that were previously unattainable in standalone systems. Simultaneous measurement of PET and MR parameters provides complementary information, allowing for a more complete assessment of disease, as well as cross validation and calibration of MR and PET measurements and techniques.
To take full advantage of such a multi-modal system, accurate quantification of the PET data is necessary. Due to the low spatial resolution of PET – which can be further reduced by external factors like patient motion – and the inherent lack of anatomic detail, accurate quantification can be challenging. The simultaneously acquired MR information provides an opportunity to optimize PET quantification and analysis. In order to fully realize the benefits provided by the simultaneously acquired MR data, the MR data cannot be treated as discrete sequences, but as the continuous flow of information. This is due to differences in the time required for data collection to generate PET and MR images.
This work describes the development and optimization of a pipeline for the reconstruction and analysis of PET data in a brain-dedicated prototype PET/MR system, the BrainPET (Siemens Healthcare). First, the performance of the BrainPET system was optimized for neurological imaging. MR-hardware interference and characteristics of the PET camera were quantified and a method for multimodal alignment was developed. To simplify and streamline the reconstruction and quantification process, a platform was designed which utilizes the functionality of a number of specialized brain imaging analysis software packages in an automated fashion. Second, MR-based methods addressing specific challenges to PET quantification were addressed. Simultaneously acquired structural MR data was used to correct the PET data for attenuation and partial volume effects. The use of MR data for motion correction was addressed and a unified algorithm which derives motion estimates from the PET data when MR data is unavailable was presented. Finally, the value of the optimized PET processing for neurological studies was evaluated in three instances: first an upper limit on the physiologic noise introduced by MR imaging on cerebral metabolism was estimated using PET and found to be minimal; next the benefit of MR-based motion correction and partial volume effect correction were estimated in a patient study; and lastly, a method to derive the PET radiotracer input function from the PET data using multiple MR sequences was presented. / Biophysics
| Identifer | oai:union.ndltd.org:harvard.edu/oai:dash.harvard.edu:1/17467219 |
| Date | 01 May 2017 |
| Creators | Chonde, Daniel B. |
| Contributors | Hooker, Jacob M., Roffman, Joshua L., Rosen, Bruce R., Hogle, James M. |
| Publisher | Harvard University |
| Source Sets | Harvard University |
| Language | English |
| Detected Language | English |
| Type | Thesis or Dissertation, text |
| Format | application/pdf |
| Rights | open |
Page generated in 0.003 seconds