Room: AAPM ePoster Library
Purpose: Hybrid Positron emission tomography / magnetic resonance imaging (PET/MRI) holds potential to reduce total radiation dose and acquire simultaneous anatomical or functional MRI. However, some MRI sequences, such as echo-planar imaging (EPI), produce a considerable amount of heat and stress on PET detector components. We investigated if heating from an EPI intense MRI protocol would effects normalized PET counts.
Methods: PET/MRI data was acquired prospectively on a Siemens Biograph mMR using a Siemens Ge-68 phantom in a Siemens Head/Neck 64 coil. The protocol spanned approximately 77 minutes acquiring: one SAG T1 IRSPGR; three EP 2D BOLD MOCO (approx. 10 min each); one TRA 2D ASL (approx. 10 min); and three EP 2D BOLD MOCO (approx. 10 min each). Data was collected under two conditions: (1) cold start, when the PET/MRI gradients were cool; and (2) warm start, when the PET/MRI gradients were warm from previous acquisitions. PET count data was normalized to a reference volume, defined as the volume of data acquired during the first five minutes of the protocol. Attenuation corrected (AC) and non-attenuation corrected (NAC) data were investigated. Decay correction over the acquisition period was not applied.
Results: During the cold start acquisition, the average initial temperature was 22C +/-1.12C and rose to 32.31C +/-9.31C as shown in Figure 1a. During the warm start acquisition, the average initial temperature was 33.83C +/-10.15C and rose to 38.67C +/-8.24C as shown in Figure 1b. For counts comparison, during the first half of the acquisition the normalized counts increased for both cold and warm starts with R-squared of 0.43 and 0.61 respectively.
Conclusion: EPI sequences produce a considerable amount of heat that may be deposited in PET detectors of hybrid PET/MRI systems. Between cold and warm start, considerable variation in gradient temperature is observed.
Not Applicable / None Entered.
Not Applicable / None Entered.