|Year : 2016 | Volume
| Issue : 1 | Page : 131-136
Comparing SUV values of images at PET-CT console and the RT planning console using identical dataset of a study phantom
Anusheel Munshi1, Sayan Paul1, Biplab Sarkar1, Pinkey Bala1, Tharmar Ganesh1, Ishita B Sen2, Vineet Pant2, Bidhu K Mohanti1
1 Department of Radiation Oncology, Fortis Memorial Research Institute, Gurgaon, Haryana, India
2 Department of Bioimaging, Fortis Memorial Research Institute, Gurgaon, Haryana, India
|Date of Web Publication||13-Apr-2016|
Department of Radiation Oncology, Fortis Memorial Research Institute, Gurgaon - 122 002, Haryana
Source of Support: None, Conflict of Interest: None
Purpose: The use of positron emission tomography (PET) for radiotherapy planning purposes has become increasingly important in the last few years.In the current study, we compared the SUV values of images at the PET CT console to the SUV values obtained at the RT planning workstation.
Materials and Methods: The PET-CT cylindrical body phantom was filled with a uniform 18F solution of 5.3. ± 0.27 kBq/mL radioactivity concentration. PET-CT scans were performed on a16 slice Time of Flight system. On a single day, the three consecutive scans were done at three time points 15 minutes apart to generate time points image data sets titled T1, T2, and T3. SUV calculations were performed by drawing region of interest. (ROI) encompassing the entire hot spot on each slice on the PET-CT console and the iPlan workstation. Minimum SUV, Maximum SUV and the Mean SUV were recorded. Statistical analysis was done using the SPSS software. (SPSS Inc.) (Version 18).
Results: The absolute difference in average max SUV values i.e. Max (PET-CT) – Max (iPlan) for the time points T1, T2 and T3 were -0.168 (SD 0.175), -0.172 (SD 0.172) and -0.178 (SD 0.169). The difference in the minimum SUV values were -0.513 (SD 0.428), -0.311 (SD 0.358) and -0.303 (SD 0.322), respectively. Finally, the difference in the mean SUV values were -0.107 (SD 0.040), -0.096 (SD 0.067) and -0.072 (SD 0.044), respectively.
Conclusions: Our study found out that the average difference in the two systems for maximum SUV values was < 0.2 absolute units.Our study suggests good reproducibility of SUV between the two systems. The relevance of these findings would be of seminal importance in current and future SUV-based PET-CT-based contouring in treatment planning systems.
Keywords: Positron emission tomography computerized tomography console, radiotherapy planning console, study phantom, SUV values
|How to cite this article:|
Munshi A, Paul S, Sarkar B, Bala P, Ganesh T, Sen IB, Pant V, Mohanti BK. Comparing SUV values of images at PET-CT console and the RT planning console using identical dataset of a study phantom. J Can Res Ther 2016;12:131-6
|How to cite this URL:|
Munshi A, Paul S, Sarkar B, Bala P, Ganesh T, Sen IB, Pant V, Mohanti BK. Comparing SUV values of images at PET-CT console and the RT planning console using identical dataset of a study phantom. J Can Res Ther [serial online] 2016 [cited 2020 Jul 7];12:131-6. Available from: http://www.cancerjournal.net/text.asp?2016/12/1/131/154934
| > Introduction|| |
Radiation oncology has moved from two-dimensional approaches to sophisticated contouring based planning. With the availability of precision planning, accurate contouring of the target has also become a desirable end point., An obvious way to reduce geographic misses in cancer is through a more accurate definition of radiotherapy (RT) target volumes. Contrast-enhanced computerized tomography (CT) is the current standard imaging modality for planning three-dimensional conformal RT (3D-CRT). However, it suffers from relatively low sensitivity and specificity in determining the extent of primary tumor and nodal involvement., Therefore, it is desirable to augment the target definition of gross tumor volume (GTV) generated by use of CT with other imaging modalities.
The usage of functional data in radiation therapy treatment planning (RTP) process is currently the focus of significant technical, scientific, and clinical research. In this context, the use of PET in the radiotherapy planning was proposed 10 years ago by Ling et al. The use of positron emission tomography (PET) for radiotherapy planning purposes has become increasingly important in the last few years. PET may be used as any other anatomic imaging modality to define a GTV., Fluorodeoxyglucose (FDG) is the most commonly used radiotracer for diagnosis, staging, recurrent disease detection, and monitoring of tumor response to therapy.,,, Multiple studies have demonstrated the utility and benefit of PET in target delineation. In the context of lung cancer, these studies have established the ability of PET-CT to differentiate areas of atelectasis from tumor in areas of lung collapse. Further, these imaging modalities have enabled easy recognition of involved nodal tissues, especially with borderline gross nodal size. In several tumor sites, functional 18F-fluoro-deoxyglucose-positron emission tomography (FDG-PET) has been shown to have higher sensitivity and specificity than CT in the detection of primary tumors and lymphatic extension.,,, Using co-registered functional PET and anatomic CT images can help improve delineation of target volumes by providing better discrimination between malignant and benign lesions and determining lymphatic status. PET-CT-based RTP has been shown to significantly alter RT fields in patients with tumors of various sites including head and neck, lung and esophagus.,,,
Modern day RTP is typically performed on dedicated RTP systems away from the PET-CT console. Image data is transferred from the PET-CT console to the RT planning system based on Digital Information and Communication in Medicine (DICOM) standards. In case of PET-CT-based planning, a target volume is then proposed on the treatment planning system (TPS) based on the PET SUV values. For accurate definition of target volumes it is essential to ensure that there is correct calculation of the SUV value from the available DICOM header information to the treatment planning system and fidelity of the SUV Value is maintained between the PET console and the RT planning workstation.
In the current study, we compared the SUV values of images at the PET-CT console to the SUV values obtained at the RT planning workstation using the same set of images acquired from a special phantom.
| > Materials and Methods|| |
For the purpose of this study, the PET-CT cylindrical body phantom [Figure 1] was filled with a uniform 18F solution of 5.3 ± 0.27 kBq/mL radioactivity concentration. The 5.3 kBq/mL radioactivity concentration corresponded to a total injected radioactivity of 5.3 kBq/mL × 70000 mL = 371 MBq (~10 mCi). PET-CT scans were performed on a 16 slice Time of Flight system (TruFlight Select from Philips CT, Phillips Medical System, Amsterdam, The Netherlands) system. The body phantom was positioned at the center of the table. A helical CT was performed for attenuation correction using 120 Kv and 250 mAs. A single-bed PET acquisition was subsequently acquired for a 3-min bed position. Images were reconstructed at 10-mm slice thickness [Figure 2]. On a single day, the three consecutive scans were done at three time points 15 minutes apart to generate time points image data sets titled T1, T2, and T3. Images were transferred directly to iPlan planning system (version 4.5.1) (BrainLAB) through direct network transfer using DICOM enabled protocols.
|Figure 2: Comparative images of sections of phantom at PET-CT and iPlan workstation|
Click here to view
SUV calculations were performed by drawing region of interest (ROI) encompassing the entire hot spot on each slice on the PET-CT console [Figure 2]. Identical ROIs were then drawn on the iPlan radiation planning system and SUV values noted [Figure 2]. For a particular slice on the phantom image on PET-CT, the following parameters were noted on each slice of the acquired set of images: Minimum SUV, Maximum SUV and the Mean SUV. Values as stated above were noted for the same slice in PET-CT console as well as in iPlan planning system. The values were cross-checked by two independent observers for accuracy. An identical methodology for selecting the denominator area on the console screen was adopted at both places. Absolute and percentage difference between the PET-CT data set and iPlan data set were computed. Time-dependent SUV variations were evaluated. Mathematical formulation to characterize the time dependency of SUV variation independently in iPlan and PET was evaluated. The differential temporal variation of SUV between two systems was calculated and time-dependent mathematical functional form was calculated. Statistical analysis was done using the SPSS software (SPSS Inc.) (Version 18), Microcal origin (OriginLab Corporation, Northampton, MA).
| > Results|| |
The absolute difference in average max SUV values i.e. Max (PET-CT) – Max (iPlan) for the time points T1, T2 and T3 were -0.168 (SD 0.175), -0.172 (SD 0.172) and -0.178 (SD 0.169). The difference in the minimum SUV values were -0.513 (SD 0.428), -0.311 (SD 0.358) and -0.303 (SD 0.322), respectively. Finally, the difference in the mean SUV values were -0.107 (SD 0.040), -0.096 (SD 0.067) and -0.072 (SD 0.044), respectively.
[Table 1] gives the average and SD of all the recorded values of SUV at T1, T2 and T3 during PET-CT and the corresponding iPlan values.
|Table 1: Average and SD of all the recorded values of SUV at T1, T2 and T3 during PET-CT and the corresponding iPlan values|
Click here to view
The nature of variation of the difference of SUV between PET and iPlan were evaluated and mathematical relationship was established between absolute as well as relative values. The relative percentage difference between maximum SUV values was found out to be 4.174 (±4.356), 4.187 (±4.088) and 4.360 (±4.072) for T1, T2 and T3, respectively. These results are depicted in [Table 2]. Minimum SUV percentage difference was noted for T1, T2 and T3 as 50.115 (±57.039), 26.056 (±34.797) and 22.214 (±23.987), respectively. Mean SUV percentage difference values were noted as 3.460 (±1.310), 3.059 (±2.187) and 2.299 (±1.426), respectively.
|Table 2: Percentage difference between PET-CT console and iPlan planning system|
Click here to view
[Figure 3] gives the temporal representation of the maximum, minimum and mean SUV values for iPLAN and PET-CT workstations. [Figure 4] gives the temporal view of the pattern of SUV difference between the two systems. Pearson correlation coefficients were calculated for combining the data at T1, T2 and T3. The maximum and mean SUV was found to be strongly correlated between PET and iPlan with an absolute value of 0.273 and 0.352, respectively. Correlation was less for minimum SUV values with a value of 0.034 with two-tailed significance as 0.966.
|Figure 3: Temporal curves of Minimum, Maximum and Mean SUV at the PET-CT console and at the iPlan workstation|
Click here to view
|Figure 4: Absolute difference, percentage fraction and percentage difference of values between PET-CT and iPlan workstations (Ave = average, SD = Standard Deviation)|
Click here to view
| > Discussion|| |
Our study found out that the average difference in the two systems for maximum SUV values was <0.2 absolute units. The average percentage difference for maximum SUV values was around 4%. This suggests reasonable good reproducibility between the two systems. We also analysed the results in terms of percentage difference between the two systems. There was <5% difference between the two systems for both the maximum and the mean SUV. For minimum SUV values, the difference in the two systems was relatively large but this is because the absolute values of minimum SUV were small and hence even a small difference would appear large in percentage. Further, the minimum SUV values have much lesser clinical significance as compared to maximum SUV values.
Further evaluating the time dependency of the difference of SUV in PET and iPlan, we obtained a definite mathematical form (parabolic) of the time dependency. This is valid for all cases if the systems are tested individually [Figure 3] or collectively [Figure 4]. SUV difference in all versions i.e., absolute difference (AD), percentage fraction (PF) and percentage difference (PD), indicated a parabola with very low eccentricity, indicating a linear variation of the difference between maximum SUV of PET and iPlan [Figure 4]. Minimum and mean SUV in all its form, AD, PF and PD shows a parabola toward the negative and positive y-axis, respectively. The parabola along y-axis indicate the temporal variation is square of time not as a square root of time (x-axis parabola). Therefore, it can be suggested that time dependency of the SUV has a definite mathematical form, which can be used to predict the SUV difference between two system on a temporal basis.
PET offers advantages in treatment planning in radiotherapy with respect to the visualization of biological processes and tumor delineation. One way of PET quantification is by using SUVs. If a strict protocol for data acquisition and analysis are followed, SUVs represent a fairly stable parameter. The goal of the present study was to evaluate the consistency and reproducibility of the SUVs and SUV-based volume calculated in two different systems– one diagnostic (e.soft) and the other radiotherapy treatment planning (iPLAN).
A major prerequisite for routine image-guided radiation therapy planning is complete fidelity of data transfer between the simulator and the RT planning workstation. PET/CT-guided RTP mandates attention to logistical aspects, patient set-up and acquisition parameters as well as an in-depth appreciation of quality control and protocol standardization.
There are two ways to contour the gross tumor using PET-CT. The radiation oncologist can do the GTV contouring on the PET-CT console or on the radiation oncology treatment planning workstation. Increasingly, manufacturers are providing SUV handling features in RT planning workstations as well and one such example is the iPlan workstation provided by BrainLAB. At present, there is a severe paucity of studies that demonstrate the maintenance of SUV fidelity during transfer of data set from PET-CT to RT planning workstation and this formed the basis of our study.
All the efforts to improve both PET and computed tomography (CT) image quality and, consequently, lesion detectability have a common objective to increase the accuracy in functional imaging and thus of co-registration into RT planning systems. In radiotherapy, improvement in target localization permits reduction of tumor margins, consequently reducing volume of normal tissue irradiated. Furthermore, smaller treated target volumes create the possibility of dose escalation, leading to increased chances of tumor cure and control.
In a relevant study, evaluation of the SUV calculation and integration of the gated (4D) PET in the iPlan 4.0 treatment planning software (BrainLAB) was done. Phantom and patient data for different tracers were used. Two comparisons were performed for each patient: for the delineated volume of interest, the maximum value of SUV in iPlan was compared with the results from a selected software. For 10 patients lesion volumes were defined in both systems for a given SUV threshold and differences were calculated. For four patients examined with respiratory gated PET, SUVmax and volume analysis was performed in each phase of the breathing cycle in the gated and the ungated PET. Maximum differences of 6% and 10% were found for phantom and patient measurements of SUVmax. For patient data, maximal differences in delineated volume of 10% for ungated and up to 27% for gated PET were found in both systems. This study suggested that for the safe implementation of PET data and delineation algorithms in the radiotherapy planning system, one had to be aware of the differences in SUVs and volumes found in the two systems.
Another study assesses the usefulness of PET/CT images to determine the target volume in radiotherapy planning by evaluating the SUV. They collected (18) F-FDG images of acrylic spheres (10-48 mm in diameter) in a phantom. The (18) F-FDG concentration in the spheres was 10-fold higher than that of the phantom. The contours were delineated according to the SUV by the threshold and ROI methods. Comparisons of two- and three-dimensional (2D and 3D) acquisition images indicated that the sharpness and quantitative qualities of the sphere boundaries were better in the former than in the latter. In the extraction of outlines using the SUV, outlines obtained at an SUV of 40-50% of the maximum agreed well with the actual acrylic sphere size. A SUV of 40-50% of the maximum was suggested to be appropriate for GTV contouring of sphere tumors with homogenously distributed (18) F-FDG.
There have been studies assessing errors in radiobiological endpoints during PET-CT. In a study, data was acquired using a human torso phantom (comprised of a hot 18F-filled spheroidal “tumor”, 40 mm in diameter) suspended in an air-filled “lung” cylinder and surrounded by a warm 18F-filled background. Reduced count statistics and misaligned CTAC images had the most detrimental impact on the image fidelity. It was found that in both cases the images became less intense, demonstrated by smaller number of voxels at the maximum values. Based on the results of this study, it is believed that simple techniques of biologically guided radiotherapy planning for lung cancer should be feasible at intermediate contrast levels (tumor-to-background ratio of approximately 10) with the clinically achievable image quality.
The primary aim of this study was to find the correlation between the iPlan and PET-CT console. In addition, our analysis found a parabolic time dependency of SUV for both iPLAN as well was PET-CT images. Intuitively, the SUV pattern should have followed the exponential radioactive decay pattern over time. However, our obtained variations seem parabolic in nature. This is established in both iPlan and PET-CT independently, therefore can be considered as not governed by systematic or random error. This could be attributed to two reasons. First the SUV is not direct representation of activity in the system or physical decay of the activity. Secondly the time frame we used (15-min gap between scans) is much shorter than the half- life of FDG (118 min). In an actual patient the scenario is different as the activity time dependency is governed by biological decay; therefore the SUV time dependency will also change. This aspect can be the subject of a subsequent study.
There are some potential drawbacks in our study. It is essentially a phantom-based study done under ideal conditions. Measurements in actual patients are possible, but are subject to much more variables than in the phantom-based controlled environment. We have validated and correlated our PET-CT with the iPlan. However, these results cannot be extrapolated to intersystem variations of iPLAN with a PET-CT machine provided by another vendor.
| > Conclusions|| |
To summarize, this is the first study of its kind correlating the SUV values between a planning system and a simulation (PET- CT) workstation. Our study suggests good reproducibility of SUV between the two systems. The relevance of these findings would be of seminal importance in current and future SUV-based PET-CT-based contouring in treatment planning systems and the contouring workstations.
| > References|| |
Niyazi M, Landrock S, Elsner A, Manapov F, Hacker M, Belka C, et al
. Automated biological target volume delineation for radiotherapy treatment planning using FDG-PET/CT. Radiat Oncol 2013;8:180.
Munshi A, Agarwal JP. Evolution of radiation oncology: Sharp gun, but a blurred target. J Cancer Res Ther 2010;6:3-4.
Orando LA, Kulasingam SL, Matchar DB. Meta-analysis: The detection of pancreatic malignancy with positron emission tomography. Aliment Pharmacol Ther 2004;20:1063-70.
Friess H, Langhans J, Ebert M, Beger HG, Stollfuss J, Reske SN, et al
. Diagnosis of pancreatic cancer by 2[18F]-fluoro-2-deoxy-D-glucose positron emission tomography. Gut 1995;36:771-7.
Ling CC, Humm J, Larson S, Amols H, Fuks Z, Leibel S, et al
. Towards multidimensional radiotherapy (MDCRT): Biological imaging and biological conformality. Int J Radiat Oncol Biol Phys 2000;47:551-60.
Geets X, Tomsej M, Lee JA, Duprez T, Coche E, Cosnard G, et al
. Adaptive biological image-guided IMRT with anatomic and functional imaging in pharyngo-laryngeal tumours: Impact on target volume delineation and dose distribution using helical tomotherapy. Radiother Oncol 2007;85:105-15.
Schaefer A, Kremp S, Hellwig D, Rübe C, Kirsch CM, Nestle U. A contrast oriented algorithm for FDG-PET-based delineation of tumour volumes for the radiotherapy of lung cancer: Derivation from phantom measurements and validation in patient data. Eur J Nucl Med Mol Imaging 2008;35:1989-99.
Lin M, Ambati C. The management impact of clinically significant incidental lesions detected on staging FDG PET-CT in patients with non-small cell lung cancer (NSCLC): An analysis of 649 cases. Lung Cancer 2012;76:344-9.
Subedi N, Scarsbrook A, Darby M, Korde K, McShane P, Muers MF. The clinical impact of integrated FDG PET-CT on management decisions in patients with lung cancer. Lung Cancer 2009;64:301-7.
Ong SC, Schöder H, Lee NY, Patel SG, Carlson D, Fury M, et al
. Clinical utility of 18F-FDG PET/CT in assessing the neck after concurrent chemoradiotherapy for Loco regional advanced head and neck cancer. J Nucl Med 2008;49:532-40.
Quon A, Fischbein NJ, McDougall IR, Le QT, Loo BW Jr, Pinto H, et al
. Clinical role of 18F-FDG PET/CT in the management of squamous cell carcinoma of the head and neck and thyroid carcinoma. J Nucl Med 2007;48:58S-67S.
Cervino AR, Evangelista L, Alfieri R, Castoro C, Sileni VC, Pomerri F, et al
. Positron emission tomography/computed tomography and esophageal cancer in the clinical practice: How does it affect the prognosis? J Cancer Res Ther 2012;8:619-25.
Kim HW, Won KS, Zeon SK, Ahn BC, Gayed IW. Peritoneal carcinomatosis in patients with ovarian cancer: Enhanced CT versus 18F-FDG PET/CT. Clin Nucl Med 2013;38:93-7.
Antoch G, Saoudi N, Kuehl H, Dahmen G, Mueller SP, Beyer T, et al
. Accuracy of wholebody dual-modality fluorine-18-2-fluoro-2-deoxy-D-glucose positron emission tomography and computed tomography (FDG-PET/CT) for tumor staging in solid tumors: Comparison with CT and PET. J Clin Oncol 2004;22:4357-68.
Kantorová I, Lipská L, Bêlohlávek O, Visokai V, Trubaĉ M, Schneiderová M. Routine (18) F-FDG PET preoperative staging of colorectal cancer: Comparison with conventional staging and its impact on treatment decision making. J Nucl Med 2003;44:1784-8.
Räsänen JV, Sihvo EI, Knuuti MJ, Minn HR, Luostarinen ME, Laippala P, et al
. Prospective analysis of accuracy of positron emission tomography, computed tomography, and endoscopic ultrasonography in staging of adenocarcinoma of the esophagus and the esophagogastric junction. Ann Surg Oncol 2003;10:954-60.
Young CS, Young BL, Smith SM. Staging Hodgkin's disease with 18-FDG PET. Comparison with CT and surgery. Clin Positron Imaging 1998;1:161-4.
van Baardwijk A, Baumert BG, Bosmans G, van Kroonenburgh M, Stroobants S, Gregoire V, et al
. The current status of FDG-PET in tumour volume definition in radiotherapy treatment planning. Cancer Treat Rev 2006;32:245-60.
Munley MT, Marks LB, Scarfone C, Sibley GS, Patz EF Jr, Turkington TG, et al
. Multimodality nuclear medicine imaging in three-dimensional radiation treatment planning for lung cancer: Challenges and prospects. Lung Cancer 1999;23:105-14.
Leong TE, Yuen K. A prospective study to evaluate the impact of coregistered PET/CT images on radiotherapy treatment planning for esophageal cancer. Int J Radiat Oncol Biol Phys 2004;60:139-40.
Kumar V, Nath K, Berman CG, Kim J, Tanvetyanon T, Chiappori AA, et al
. Variance of SUVs for FDG-PET/CT is greater in clinical practice than under ideal study settings. Clin Nucl Med 2013;38:175-82.
Thorwarth D, Beyer T, Boellaard R, de Ruysscher D, Grgic A, Lee JA, et al
. Integration of FDG-PET/CT into external beam radiation therapy planning: Technical aspects and recommendations on methodological approaches. Nuklearmedizin 2012;51:140-53.
Jacob V, Astner ST, Bundschuh RA, Busch R, Souvatzoglou M, Wendl C, et al
. Evaluation of the SUV values calculation and 4D PET integration in the radiotherapy treatmentplanning system. Radiother Oncol 2011;98:323-9.
Scripes PG, Yaparpalvi R. Technical aspects of positron emission tomography/computed tomography in radiotherapy treatment planning. Semin Nucl Med 2012;42:283-8.
Uto F, Shiba E, Onoue S, Yoshimura H, Takada M, Tsuji Y, et al.
Phantom study on radiotherapy planning using PET/CT--delineation of GTV by evaluating SUV. J Radiat Res 2010;51:157-64.
Perrin R, Evans PM, Webb S, Partridge M. The use of PET images for radiotherapy treatment planning: An error analysis using radiobiological endpoints. Med Phys 2010;37:516-31.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2]