A Comparative Study between Evaluation Methods for Quality Control Procedures for Determining the Accuracy of PET/CT Registration

다국어 초록 (Multilingual Abstract)

The Accuracy of registration between positron emission tomography (PET) and computed tomography (CT) images is one of the important factors for reliable diagnosis in PET/CT examinations.
Although quality control (QC) for checking alignment of PET and CT images should be performed periodically, the procedures have not been fully established. The aim of this study is to determine optimal quality control (QC) procedures that can be performed at the user level to ensure the accuracy of PET/CT registration. Two phantoms were used to carry out this study: the American college of Radiology (ACR)-approved PET phantom and National Electrical Manufacturers Association (NEMA) International Electrotechnical Commission (IEC) body phantom, containing fillable spheres. All PET/CT images were acquired on a Biograph TruePoint 40 PET/CT scanner using routine protocols. To measure registration error, the spatial coordinates of the estimated centers of the target slice (spheres) was calculated independently for the PET and the CT images in two ways. We compared the images from the ACR-approved PET phantom to that from the NEMA IEC body phantom. Also, we measured the total time required from phantom preparation to image analysis. The first analysis method showed a total difference of 0.636 ± 0.11 mm for the largest hot sphere and 0.198 ± 0.09 mm for the largest cold sphere in the case of the ACR-approved PET phantom. In the NEMA IEC body phantom, the total difference was 3.720 ± 0.97 mm for the largest hot sphere and 4.800 ± 0.85 mm for the largest cold sphere. The second analysis method showed that the differences in the x location at the line profile of the lesion on PET and CT were (1.33 , 1.33) mm for a bone lesion, (−1.26, −1.33) mm for an air lesion and (−1.67, −1.60) mm for a hot sphere lesion for the ACR-approved PET phantom. For the NEMA IEC body phantom, the differences in the x location at the line profile of the lesion on PET and CT were (−1.33, 4.00) mm for the air lesion and (1.33, −1.29) mm for a hot sphere lesion. These registration errors from this study were reasonable compared to the errors reported in previous studies. Meanwhile, the total time required from phantom preparation was 67.72 ± 4.50 min for the ACR-approved PET phantom and 96.78 ± 8.50 min for the NEMA IEC body phantom. When the registration errors and the lead times are considered, the method using the ACR-approved PET phantom was more practical and useful than the method using the NEMA IEC body phantom.

참고문헌 (Reference)

1 R. Rakheja, 201 : 1120-, 2013

2 A. H. Ng, 67 : 1864-, 2009

3 R. Shekhar, 46 : 1488-, 2005

4 K. J. Nichols, 13 : e25-, 2006

5 W. V. Vogela, 27 : 515-, 2006

6 Y. Watanabe, 109 : 21-, 2008

7 A. Rodr´ıguez-Ruano, 3832-, 2008

8 P. Zanzonico, 49 : 1114-, 2008

9 Y. S. Lee, 28 : 340-, 2014

10 C. Cohade, 30 : 721-, 2003

11 T. Chang, 39 : 5891-, 2012

12 M. C. Ba˜nos-Capilla, 34 : 1911-, 2007

13 G. Akamatsu, 53 : 1716-, 2012

14 E. B. Sokole, 37 : 662-, 2010

15 L.Sing, 71 : S38-, 2008

16 A. Isambert, 12 : 800-, 2008

17 M. Sharpe, 71 : S33-, 2008

18 G. El Fakhri, "Quality assurance for PET and PET/CT systems"

19 ACR, "PET phantom instructions for evaluation of PET image quality"

20 General Electric Company, "Optima 560, Discovery 600, 690 Elite PET/CT Service Manual"

21 CPI Innovations, "ECAT LSO PET/CT 16 with PICO 3D Operator’s Guide"