Dynamic FDG PET/CT for differentiating focal pelvic uptake in patients with gynecological cancer
This study aimed to evaluate the ability of serial whole-body dynamic PET/CT to differentiate physiological from abnormal <sup>18</sup>F-FDG uptake in the abdomen and pelvis of gynecological cancer patients. We conducted a retrospective study of 61 <sup>18</sup>F-FDG PET/CT examinations for suspected gynecological malignancies or metastases between March 2018 and January 2020. Our protocol included four-phase dynamic whole-body scans. High-uptake foci with SUVmax > 2.5 in the abdominopelvic region caudal to the renal portal were picked up and visually evaluated as "changed" (disappeared during any phase or morphological changes in more than half of the foci) or "unchanged" in motion on the serial dynamic images. Focal <sup>18</sup>F-FDG uptake was observed in 84 foci. Of the 58 foci determined pathologically or clinically to have pathological uptake, no change was observed on serial dynamic imaging in 54 foci (sensitivity, 93%). Of the 26 foci of physiological uptake, temporal changes in uptake were observed in 20 foci using dynamic imaging (specificity, 77%). The positive and negative predictive values were 90% and 83%, respectively, with an accuracy of 88%. Dynamic whole-body <sup>18</sup>F-FDG PET/CT imaging allows for differentiation between pathological and physiological uptake in the abdominopelvic region of patients with gynecological cancer.
Authors
- Yasuchiyo Toyama
- Takeshi Nii (ORCID: https://orcid.org/0000-0003-0409-8378)
- Nagara Tamaki (ORCID: https://orcid.org/0000-0001-6193-8594)
- Yoshitomo Nakai (ORCID: https://orcid.org/0000-0002-8259-2147)
- Tomoya Kotani (ORCID: https://orcid.org/0000-0003-1930-6635)
- Motoki Nishimura
- Kei Yamada (ORCID: https://orcid.org/0000-0002-8123-3935)
- Yasunori Nakamura (ORCID: https://orcid.org/0000-0001-5731-8536)
- Sachimi Yamada
Institutions
- Kyoto Prefectural University of Medicine (JP)
- Kyoto College of Medical Science (JP)
- University Hospital Kyoto Prefectural University of Medicine (JP)
- Nantan General Hospital (JP)
- Japanese Red Cross Society Kyoto Daini Hospital (JP)
Publication Details
- Journal
- Scientific Reports
- Published
- 2024-11-27
- DOI
- https://doi.org/10.1038/s41598-024-81236-x
- Citations
- 2
- Primary Topic
- Radiomics and Machine Learning in Medical Imaging
- Type
- article
- Field-Weighted Citation Impact
- 0.81
Funders
- Nihon Medi-Physics
- Japan Society for the Promotion of Science