MR-based attenuation correction (MRAC) techniques are a critical component to enable quantitative PET techniques for simultaneous PET/MR imaging. This project seeks to develop robust, whole-body MRAC techniques utilizing a multi-echo flyback UTE imaging acquisition with advanced IDEAL signal modeling. These acquisitions have been demonstrated to be robust for body imaging, and allow direct reconstruction of separate water, fat, and short T2* components. Additional algorithms will be developed to segment this short T2* component data into lung and bone tissue, investigating the efficacy of methods that utilize fully-automated segmentation algorithms, as well as those that use anatomical cues. Quantitative PET data from a range of clinical examinations will be analyzed using the developed MRAC methods and existing CT-based attenuation correction (CTAC) to validate these MR-based techniques for whole-body MRAC.
April 2014 to December 2014
This project led by: Alan McMillan PhD