Abstract
1525
Objectives: Data driven gating (DDG) can improve PET quantification and alleviate many issues with patient motion. However, misregistration between DDG PET and CT may occur due to the distinct temporal resolutions of PET and CT, and can be mitigated by DDG CT. Here, the effects of misregistration and respiratory motion on PET quantification and lesion segmentation were assessed in order to explore the potential influence on radiation therapy planning and tumor response assessment.
Methods: A low-dose cine CT was acquired in the region of misregistration to enable both average CT (ACT) and DDG CT. Four different sets of PET data were computed and compared: 1) baseline PET/CT, 2) PET with ACT attenuation correction (AC), 3) DDG PET with AC using baseline CT, and 4) DDG PET/CT with AC using DDG CT. For DDG PET, end-expiration (EE) data was derived from 50% of the total PET data at 30% from end-inspiration. For DDG CT, EE phase CT data was extracted from cine CT data. A total of 93 lesions from 16 patients were contoured using a threshold based on SUVmax. Changes in SUVmax, lesion volume, centroid-to-centroid distance, and DICE coefficients were tracked and compared.
Results: Relative to baseline PET/CT, mean changes in SUVmax ±σ for all 93 lesions were 32±41%, 26±22%, and 85±66%, respectively, for PET/ACT, DDG PET, and DDG PET/CT. Of the 93 lesions tracked, 39%, 42%, and 78% maintained changes in SUVmax >30% for the same categories listed above. Mean changes in lesion volume were 28±75%, ˗33±30%, and -28±42%. Centroid-to-centroid distance averages were 2.7±3.1 mm, 5.0±3.8 mm, and 5.5±4.3 mm. DICE coefficients relative to the baseline PET/CT contours had mean±σ of 0.77±0.19, 0.61±0.20, and 0.58±0.19. Compared to baseline PET/CT, average total lesion glycolysis changes for the 16 patients were 30±30%, -20±19%, and 8±36%.
Conclusions: Both accurate AC (improved registration) and motion correction (DDG) affected nearly all aspects of lesion quantification and segmentation. AC and DDG could cause opposite changes in quantification or segmentation, indicating these two corrections can at times work in distinct ways. As misregistration becomes more prominent, the effect of motion correction with DDG PET becomes less important. Tumor volume becomes smaller with DDG PET (p < 0.001). Quantification can increase over 30% for 39, 42 and 78% of the lesions with PET/ACT, DDG PET and DDG PET/CT, respectively. These results impress upon the necessity of ensuring both misregistration and motion correction are accounted for together to optimize the clinical utility of PET/CT. Acknowledgements: This research was supported in part by NIH grants R21-CA222749-01A1 and R03-EB030280-01, and a ROSI grant from the UT M.D. Anderson Cancer Center, Division of Radiation Oncology. This research was conducted at the M.D. Anderson Cancer Center for Advanced Biomedical Imaging in-part with equipment support from General Electric Healthcare.