Method to Determine the Standard Deviation of SUV Parameters

31 Some of the parameters used for the quantification of Positron Emission Tomography (PET) images 32 are the Standardized Uptake Value (SUV)Max, SUVMean and SUVPeak. In order to assess the 33 significance of an increasing or decreasing of these parameters for diagnostic purposes it is relevant 34 to determine their standard deviation. In this study we present a method to determine the range of 35 statistical variation of the SUV in PET images. 36 Our method is based on dividing an original dataset into subsets of shorter time-frames. The variation 37 between the SUV parameters of the subsets is used to estimate the standard deviation of the of the 38 original acquisition. This method was tested on images of a NEMA quality phantom with acquisition 39 time of 150 s per bed position and foreground to background activity ratio of F 18 -2-fluoro-2-deoxy-D- 40 glucose (FDG) of 10:1. This original dataset has been reconstructed with different reconstruction 41 lengths, generating new data subsets. The SUVMax, Mean and Peak were calculated for each image 42 in the subsets. Their standard deviation has been calculated per subset for the different spheres 43 included in the phantom. The variation of each subset has then been used to estimate the expected 44 variation between images at 150 s reconstruction length. 45 We report the largest standard deviation of the SUV parameters for the smallest sphere, and the 46 smallest variation for the largest sphere. The expected variation at 150 s reconstruction length does 47 not exceed 6% for the smallest sphere and 2% for the largest sphere, but we report an higher 48 coefficient of variation (up to 30%) for shorter reconstruction lengths. We also report significant 49 differences in the variation of SUV parameters for the larger spheres. With the presented method we 50 are able to determine the standard deviation of SUV parameters only due to and the statistical 51 variation. The method enables us to evaluate the effect of parameter selection and lesion size on the 52 standard deviation, and therefore to evaluate its relevance on the total variation of the SUV value 53 between studies. quantifications parameters, when the images are acquired with the same scanner and software reconstruction, is crucial to provide a reliable interpretation of the data. In this study we present a method to estimate the standard deviation of the SUV parameters on different SUV parameters and lesion sizes of images acquired with the


56
Positron Emission Tomography (PET) has become an indispensable diagnostic tool over the last 57 decades. Computed Tomography (CT) is added to the PET modality for the purpose of attenuation 58 correction and furthermore PET-CT imaging provides a combined view of functional and morphological 59 information.
The radio-ligand FDG has ensured the success of PET-imaging. The glucose component of the molecule provides a higher uptake of FDG in malignant than in healthy cells [1], the fluor-18 62 component provides the detectability in the PET-CT system.

63
PET-CT images can be reported visually by the nuclear medicine physician, however an important 64 advantage of PET imaging is that the uptake can be quantified in absolute measures.

73
SUVMax has a low inter and intra observer variability but a high technical statistical variation. The

74
SUVMean on the other hand has a lower technical variation but a higher inter and intra observer 75 variability, since the borders of the volume are a determining factor of the result.

76
The SUVPeak is introduced as a "best of both worlds" parameter, it calculates the voxels in a limited 77 volume around the voxel with the maximum value. same scanner and reconstructed with the same method. The basics of the method is that one PET acquisition is divided into a number of time-frames and that the variation between the SUV 92 quantification of the separate time-frames is used to estimate the standard deviation of the total 93 acquisition. The proposed method determines the standard deviation of SUV parameters such that the  calibration, synchronization, image and scan reconstruction parameters is not present. We quantify the 101 variability of the SUV parameters between subsets of images derived from the same acquisition and 102 reconstructed in the same way. In comparison with test-retest method, our framework compares 103 images with less variable conditions and we focus on the impact of the reconstruction on images from 104 the same subset. With this method we want to quantify a degree of technical variability that is scanner-105 and reconstruction-specific and that might have a relative larger impact in images with small lesions, 106 low photon counts and/or high noise.
A NEMA NU2-2007 image quality phantom was imaged on a Philips Gemini TF PET/CT system 120 (Philips Healthcare, Andover, MA). PET reconstructions were made using scanner's default ordered 121 subset expectation maximization (OSEM) reconstruction algorithm with 33 subsets, 3 iterations, matrix 122 size of 144 × 144, and voxels of 4 × 4 × 4 mm. No Gaussian filter was applied. The reconstruction 123 corrected for geometrical response and detector efficiency (normalization), random coincidences, 124 scatter, and attenuation. Data were stored in list-mode, to be able to reconstruct different acquisition 125 times. All list-mode reconstructions were decay-corrected to the start time of the acquisition.

126
The phantom acquisitions were made according to the requirements for the EANM/EARL FDG-

127
PET/CT accreditation [7]. The phantom was composed by a fillable torso compartment acting as     (1)

182
With SD1 and SD2 being the standard deviations and RL1 and RL2 being the reconstruction lengths.

183
[5]. By using the measured variation of the SUV values in a subset as SD1 and the length of the 184 reconstruction of the specific subset as RL1, it is possible to estimate the variation SD2 between 185 repetitions of scans at the total acquisition time RL2=150 s.

187
Since we divided our acquisition into 14 different subsets, we could calculate 14 different estimations 188 of the SD2, using the described method above. We validated our method by testing whether the value 189 of the estimated SD2 was independent of the acquisition length of the images in the subset.

190
The results were tested to verify if is a significant difference between the spheres was present. The

197
The method that we describe in this paper for estimating the variation of SUV parameters between

198
PET images includes three main steps:

199
• acquisition of a dataset of a specific length

200
• generation and reconstruction of subsets

201
• estimation of the variation in the original dataset by using the subsets, according to Formula 1.

202
As an example for this method we used a 150 s acquisition for the original dataset, then we divided it

256
The data were statistically analysed to verify if the difference between the estimated SD of the SUV 257 parameters was significant between spheres (same parameter, different sphere diameter, so 258 difference between columns in Tab.1) and between SUV parameters (different SUV parameter, same 259 sphere diameter, so difference between rows in Tab.1).

260
Concerning the differences between spheres, we report that the difference between the estimated SD 261 of the SUV parameters of the sphere with d=10 mm and d=13 mm and with d=28 mm and d=37 mm is the sphere with d=13 mm and d=17 mm is also significant. The difference between estimated SD of 264 the SUVMean 2D and 3D is significant between each sphere diameter.

265
Concerning the differences between SUV parameters, we report no significant difference between the 266 estimated SD of the SUV parameters of the two smaller spheres (d=10, 13 mm

292
(images with high standard deviation between voxels) or for small lesions. In our case we observe that the variation is higher for shorter reconstruction lengths, suggesting that the contribution of the 294 technical variation might be higher, for example, for images acquired with a shorter acquisition time or 295 with a low counts emitter. The higher variation at shorter reconstruction lengths reaches values up to 296 30%, in our case, for the sphere with 10 mm diameter. This suggests that, when performing 297 quantification of PET images on small lesion, the effect of the technical variability evaluated with this 298 method might not be negligible when compared with the variation used for diagnostical purposes. Our 299 method might also be used to define the minimal required acquisition length: when the technical 300 statistical variation of the SUV has become negligible to the test-retest variation, a longer acquisition 301 time might not add value.

303
In Table 1 we report the result of the estimated standard deviation of the SUV quantifications. We

349
For this study we worked with a foreground to background activity ratio of 10:1. In order to further 350 verify the method with other uptakes it would be possible to repeat the study with other ratio's, as for hours in order to analyse the variation with other levels of noise. As previously discussed, a higher 353 coefficient of variation is to be expected for noisier images.

356
In this study we present a method to estimate the standard deviation of different SUV parameters. We   Coe cient of variation of the SUVMax 2D as a function of the reconstruction length.

Figure 3
Coe cient of variation of the SUVMax 3D as a function of the reconstruction length. Coe cient of variation of the SUVMean 2D as a function of the reconstruction length.

Figure 5
Coe cient of variation of the SUVMean 3D as a function of the reconstruction length.

Figure 6
Coe cient of variation of the SUVPeak as a function of the reconstruction length.