Toshiki Takeshita1, Keishin Morita2, Yuji Tsutsui3, Daisuke Kidera2, Shohei Mikasa2, Akira Maebatake2, Go Akamatsu2, Kenta Miwa2, Shingo Baba3, Masayuki Sasaki4. 1. Department of Health Sciences, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan. t.takeshita.1227@gmail.com. 2. Department of Health Sciences, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan. 3. Division of Radiology, Department of Medical Technology, Kyushu University Hospital, Fukuoka, Japan. 4. Department of Health Sciences, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan. msasaki@hs.med.kyushu-u.ac.jp.
Abstract
OBJECTIVE: The purpose of this study was to investigate the influence of respiratory motion on the evaluation of the intratumoral heterogeneity of FDG uptake using cumulative SUV-volume histogram (CSH) and fractal analyses. METHODS: We used an NEMA IEC body phantom with a homogeneous hot sphere phantom (HO) and two heterogeneous hot sphere phantoms (HE1 and HE2). The background radioactivity of (18)F in the NEMA phantom was 5.3 kBq/mL. The ratio of radioactivity was 4:2:1 for the HO and the outer rims of the HE1 and HE2 phantoms, the inner cores of the HE1 and HE2 phantoms, and background, respectively. Respiratory motion was simulated using a motion table with an amplitude of 2 cm. PET/CT data were acquired using Biograph mCT in motionless and moving conditions. The PET images were analyzed by both CSH and fractal analyses. The area under the CSH (AUC-CSH) and the fractal dimension (FD) was used as quantitative metrics. RESULTS: In motionless conditions, the AUC-CSHs of the HO (0.80), HE1 (0.75) and HE2 (0.65) phantoms were different. They did not differ in moving conditions (HO, 0.63; HE1, 0.65; HE2, 0.60). The FD of the HO phantom (0.77) was smaller than the FDs of the HE1 (1.71) and HE2 (1.98) phantoms in motionless conditions; however, the FDs of the HO (1.99) and HE1 (2.19) phantoms were not different from each other and were smaller than that of the HE2 (3.73) phantom in moving conditions. CONCLUSION: Respiratory motion affected the results of the CSH and fractal analyses for the evaluation of the heterogeneity of the PET/CT images. The influence of respiratory motion was considered to vary depending on the object size.
OBJECTIVE: The purpose of this study was to investigate the influence of respiratory motion on the evaluation of the intratumoral heterogeneity of FDG uptake using cumulative SUV-volume histogram (CSH) and fractal analyses. METHODS: We used an NEMA IEC body phantom with a homogeneous hot sphere phantom (HO) and two heterogeneous hot sphere phantoms (HE1 and HE2). The background radioactivity of (18)F in the NEMA phantom was 5.3 kBq/mL. The ratio of radioactivity was 4:2:1 for the HO and the outer rims of the HE1 and HE2 phantoms, the inner cores of the HE1 and HE2 phantoms, and background, respectively. Respiratory motion was simulated using a motion table with an amplitude of 2 cm. PET/CT data were acquired using Biograph mCT in motionless and moving conditions. The PET images were analyzed by both CSH and fractal analyses. The area under the CSH (AUC-CSH) and the fractal dimension (FD) was used as quantitative metrics. RESULTS: In motionless conditions, the AUC-CSHs of the HO (0.80), HE1 (0.75) and HE2 (0.65) phantoms were different. They did not differ in moving conditions (HO, 0.63; HE1, 0.65; HE2, 0.60). The FD of the HO phantom (0.77) was smaller than the FDs of the HE1 (1.71) and HE2 (1.98) phantoms in motionless conditions; however, the FDs of the HO (1.99) and HE1 (2.19) phantoms were not different from each other and were smaller than that of the HE2 (3.73) phantom in moving conditions. CONCLUSION: Respiratory motion affected the results of the CSH and fractal analyses for the evaluation of the heterogeneity of the PET/CT images. The influence of respiratory motion was considered to vary depending on the object size.
Authors: Attila Forgacs; Piroska Kallos-Balogh; Ferenc Nagy; Aron K Krizsan; Ildiko Garai; Lajos Tron; Magnus Dahlbom; Laszlo Balkay Journal: PLoS One Date: 2019-01-25 Impact factor: 3.240