Clinical Impact of Respiratory Motion Correction in Simultaneous PET/MR, Using a Joint PET/MR Predictive Motion Model.

In PET imaging, patient motion due to respiration can lead to artifacts and blurring, in addition to quantification errors. The integration of PET imaging with MRI in PET/MRI scanners provides spatially aligned complementary clinical information and allows the use of high-contrast, high-spatial-resolution MR images to monitor and correct motion-corrupted PET data. On a patient cohort, we tested the ability of our joint PET/MRI-based predictive motion model to correct respiratory motion in PET and show it can improve lesion detectability and quantitation and reduce image artifacts. Using multiple tracers and multiple organ locations, we applied our motion correction method to 42 clinical PET/MRI patient datasets containing 162 PET-avid lesions. Quantitative changes were calculated using SUV changes in avid lesions. Lesion detectability changes were explored with a study in which 2 radiologists identified lesions in uncorrected and motion-corrected images and provided confidence scores. Mean increases of 12.4% for SUV and 17.6% for SUV after motion correction were found. In the detectability study, confidence scores for detecting avid lesions increased, with a rise in mean score from 2.67 to 3.01 (of 4) after motion correction and a rise in detection rate from 74% to 84%. Of 162 confirmed lesions, 49 showed an increase in all 3 metrics-SUV, SUV, and combined reader confidence score-whereas only 2 lesions showed a decrease. We also present clinical case studies demonstrating the effect that respiratory motion correction of PET data can have on patient management, with increased numbers of detected lesions, improved lesion sharpness and localization, and reduced attenuation-based artifacts. We demonstrated significant improvements in quantification and detection of PET-avid lesions, with specific case study examples showing where motion correction has the potential to affect diagnosis or patient care.