Abstract
1548
Purpose: Respiratory motion can degrade 18F-FDG PET/CT image quality and affect quantitative evaluation of lesions causing inaccurate PET indices such as standard uptake value (SUV) and tumor volume. It can also lead to misregistration in fusion PET/CT images. To address this limitation, our clinical institute has installed PET/CT DMIDR (Discovery Molecular Imaging Digital Ready, General Electric Healthcare, USA) using AI algorithm-based respiratory motion artifact correction (Motion FreeTM). In this study, we aimed to evaluate the benefit of Motion Free algorithm through Phantom test and applying it to patients in clinical practice. Materials and Methods: The various size of NEMA IEC Body Phantom (sphere size 10mm, 22mm, 37mm) and vacuum vial (single, dual) were linked to RPM Phantom to produce the respiratory motion, before and after the Motion Free application through Phantom test. In addition, total 600 patients, who underwent 18F-FDG PET/CT for cancer work-up, were sequentially divided into 6 groups with 100 patients by each breathing protocol and CT scan time in a prospective research from September 2020 to December 2020. Various breathing protocols were applied as follows: a) Patients who were performed without any breathing instruction during image acquisition, b) Patients who were instructed as breathing holding after expiration during CT scan, c) Patients who were instructed as natural breathing. The CT scan protocol was performed as two protocols of 13 seconds and 8 seconds to evaluate the misregistration of fusion PET/CT images. The misregistration evaluation was assessed by distance more than 10mm of fusion PET/CT images based on liver dome.
Results: In the phantom test, the volume of Sphere using Motion Free was indicated as smaller size and higher SUV than those of Non Motion Free. This finding seems to be consistent with static image without causing motion. There was statically significant difference in fusion PET/CT misregistration according to CT scan time and each breathing protocols (p < 0.05). In ⓒ protocol with 13-second CT scan, the frequency of misregistration occurred in 51 cases, but the distance of misregistration was shorter than using other respiratory protocols (p < 0.05). In 8 seconds CT scan, the distance of misregistration revealed significantly shorter in ⓐ protocol than ⓑ. The frequency of misregistration occurred the least in ⓒ protocol, however, the distance of misregistration could not show statistical significance compared with ⓐ protocol.
Conclusions: In our study, Motion Free application can be clinically useful for respiratory motion artifact correction during 18F-FDG PET/CT acquisition. Furthermore, the shorter time of CT scan (8 seconds) can reduce the frequency of fusion PET/CT misregistration and natural breathing instruction can also improve the distance of misregistration.
The number of frequency of misregistration in fusion PET/CT images and misregistration distance
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.