A review of harmonization strategies for quantitative PET.

PET can reveal in vivo biological processes at the molecular level. PET-derived quantitative values have been used as a surrogate marker for clinical decision-making in numerous clinical studies and trials. However, quantitative values in PET are variable depending on technical, biological, and physical factors. The variability may have a significant impact on a study outcome. Appropriate scanner calibration and quality control, standardization of imaging protocols, and any necessary harmonization strategies are essential to make use of PET as a biomarker with low bias and variability. This review summarizes benefits, limitations, and remaining challenges for harmonization of quantitative PET, including whole-body PET in oncology, brain PET in neurology, PET/MR, and non-18F PET imaging. This review is expected to facilitate harmonization of quantitative PET and to promote the contribution of PET-derived biomarkers to research and development in medicine.

Effect of Tumor-Pixel Positioning on the Variability of SUV Measurements in PET Images.

Objectives: The aim of this study was to investigate the effect on standardized uptake value (SUV) measurement variability of the positional relationship between objects of different sizes and the pixel of a positron emission tomography (PET) image. Methods: We used a NEMA IEC body phantom comprising six spheres with diameters of 10, 13, 17, 22, 28, and 37 mm. The phantom was filled with 18F solution and contained target-to-background ratios (TBRs) of 2, 4, and 8. The PET data were acquired for 30 min using a SIGNA PET/MR scanner. The PET images were reconstructed with the ordered subsets expectation maximization (OSEM) algorithm with and without point-spread function (PSF) correction (OSEM + PSF + Filter and OSEM + Filter, respectively). A Gaussian filter of 4 mm full width at half maximum was applied in all reconstructions, except for one model (OSEM + PSF + no Filter). The matrix sizes were 128×128, 192×192, 256×256 and 384×384. Reconstruction was performed by shifting the reconstruction center position by 1 mm in the range 0 to 3 mm in the upward or rightward direction for each parameter. For all reconstructed images, the SUVmax of each hot sphere was measured. To investigate the resulting variation in the SUVmax, the coefficient of variation (CV) of each SUVmax was calculated. Results: The CV of the SUVmax increased as the matrix size and the diameter of the hot sphere decreased in all reconstruction settings. With PSF correction, the CV of SUVmax increased as the TBR increased except when the TBR was 2. The CV of the SUVmax measured in the OSEM + PSF + no Filter images were larger than those measured in the OSEM + PSF + Filter images. The amount of this increase was higher for smaller spheres and larger matrix sizes and was independent of TBR. Conclusions: Shifting the reconstruction center position of the PET image causes variability in SUVmax measurements. To reduce the variability of SUV measurements, it is necessary to use sufficient matrix sizes to satisfy sampling criterion and appropriate filters.

Improvement in PET/CT image quality in overweight patients with PSF and TOF.

OBJECTIVE: The aim of this study was to evaluate the effect of the point spread function (PSF) and time of flight (TOF) on PET/CT images of overweight patients in relation to the iteration number and the acquisition time. METHODS: This study consisted of a phantom study and a clinical study. The NEMA IEC body phantom and a 40 cm diameter large phantom (LG phantom) simulating an overweight patient were used in this study. Both phantoms were filled with (18)F solution with a sphere to background ratio of 4:1. The PET data were reconstructed with the baseline ordered-subsets expectation maximization (OSEM) algorithm, with the OSEM + PSF model, with the OSEM + TOF model and with the OSEM + PSF + TOF model. The clinical study was a retrospective analysis of 66 patients who underwent (18)F-FDG PET/CT. The image quality was evaluated using the background variability (coefficient of variance, CVphantom and CVliver) and the contrast (CONTHOT and SNR). RESULTS: In phantom study, the CVphantom of the LG phantom was higher than that of the NEMA phantom. The PSF decreased the CVphantom of the LG phantom to the NEMA phantom level. The TOF information accelerated the CVphantom plateau earlier. The best relationship between the CVphantom and the CONTHOT was observed for the OSEM + PSF + TOF. In clinical study, the combination of PSF and TOF decreased the CVliver for overweight patients to that for normal weight patients while it increased the SNR similarly between two patient groups. CONCLUSION: The combination of the PSF and TOF correction improved the image quality of the LG phantom and overweight patients.

Impact of Time-of-Flight PET/CT with a Large Axial Field of View for Reducing Whole-Body Acquisition Time.

UNLABELLED: The aim of this study was to evaluate the imaging performance of 39- and 52-ring time-of-flight (TOF) PET/CT scanners. We also assessed the potential of reducing the scanning time using a 52-ring TOF PET/CT scanner. METHODS: PET/CT scanners with 39- and 52-ring lutetium oxyorthosilicate detectors were evaluated. The axial fields of view were 16.2 and 21.6 cm, respectively. We used a National Electrical Manufacturers Association International Electrotechnical Commission body phantom filled with an (18)F solution containing background activity of 5.31 and 2.65 kBq/mL for the studies. The sphere-to-background ratio was 4:1. The PET data were acquired for 10 min in 3-dimensional list mode and then reconstructed with both ordered-subsets reconstruction maximization and ordered-subsets reconstruction maximization plus point-spread function plus time-of-flight algorithms. PET images with different acquisition times were reconstructed (from 1 to 10 min). The image quality was physically assessed using the sensitivity, noise-equivalent counting rate, coefficient of variation of background activity, and relative recovery coefficient. RESULTS: The total system sensitivities of the 39- and 52-ring scanners were 5.6 and 9.3 kcps/MBq, respectively. Compared with the 39-ring scanner, the noise-equivalent counting rate of the 52-ring scanner was 60% higher for both the high-activity and the low-activity models. The recovery coefficient was consistent, irrespective of the number of detector rings. The coefficient of variation of the 52-ring scanner using a 3-min acquisition time was equivalent to that of the 39-ring scanner using a 4-min acquisition time. CONCLUSION: The image quality of the 52-ring scanner is superior to that of the 39-ring scanner. The acquisition time per bed position of the 52-ring system can be reduced by about 25% without compromising image quality. In addition, the number of bed positions required is 25% lower for the 52-ring system. Finally, the examination time required for a whole-body PET scan is considered to be reduced by about 40% if the 52-ring scanner is used.

Influences of point-spread function and time-of-flight reconstructions on standardized uptake value of lymph node metastases in FDG-PET.

PURPOSE: The purpose of this study was to investigate the effects of point-spread function (PSF) and time-of-flight (TOF) on the standardized uptake value (SUV) of lymph node metastasis in FDG-PET/CT. MATERIALS AND METHODS: This study evaluated 41 lymph node metastases in 15 patients who had undergone (18)F-FDG PET/CT. The lesion diameters were 2.5 cm or less. The mean short-axis diameter of the lymph nodes was 10.5 ± 3.7 mm (range 4.6-22.8mm). The PET data were reconstructed with baseline OSEM algorithm, with OSEM+PSF, with OSEM+TOF and with OSEM+PSF+TOF. A semi-quantitative analysis was performed using the maximum and mean SUV of lymph node metastases (SUVmax and SUVmean) and mean SUV of normal lung tissue (SUVlung). We also evaluated image quality using the signal-to-noise ratio in the liver (SNRliver). RESULTS: Both PSF and TOF increased the SUV of lymph node metastases. The combination of PSF and TOF increased the SUVmax by 43.3% and the SUVmean by 31.6% compared with conventional OSEM. By contrast, the SUVlung was not influenced by PSF and TOF. TOF significantly improved the SNRliver. CONCLUSION: PSF and TOF both increased the SUV of lymph node metastases. Although PSF and TOF are considered to improve small-lesion detectability, it is important to be aware that PSF and TOF influence the accuracy of quantitative measurements.

Benefits of point-spread function and time of flight for PET/CT image quality in relation to the body mass index and injected dose.

UNLABELLED: The PET image quality of overweight patients and patients who receive low injected doses deteriorates because of increases in statistical noise. The purpose of this study was to investigate the benefits of the point-spread function (PSF) and time-of-flight (TOF) for PET/CT image quality in such patients. METHODS: The PET images were reconstructed using the baseline ordered-subsets expectation-maximization algorithm (OSEM), OSEM + PSF, OSEM + TOF, and OSEM + PSF + TOF. In the phantom study, we used a National Electrical Manufacturers Association body phantom with different radioactivity concentrations and analyzed image quality using the coefficient of variance in the background (CVphantom). In the clinical study, we retrospectively studied 39 patients who underwent clinical F-FDG PET/CT. The patients were classified into groups based on body mass index and injected dose. Image quality was evaluated using the CV in the liver (CVliver). RESULTS: In the phantom study, PSF and TOF improved the CVphantom, especially in low-activity models. Among all of the reconstructions, the best CVphantom was obtained with OSEM + PSF + TOF. In the clinical study, the CVliver of the low-dose group with OSEM + PSF + TOF was comparable to that of the high-dose group with conventional OSEM. CONCLUSIONS: Point-spread function and TOF improved PET/CT image quality for overweight patients who received a lower injected dose. Therefore, the use of PSF and TOF is suggested to maintain the image quality of such patients without extending scanning times. It is greatly beneficial to obtain sufficient image quality for larger patients, especially in delivery institutions where the injection dose cannot be easily increased.

Improvement in PET/CT image quality with a combination of point-spread function and time-of-flight in relation to reconstruction parameters.

UNLABELLED: The aim of this study was to investigate the effects of the point-spread function (PSF) and time-of-flight (TOF) on improving (18)F-FDG PET/CT images in relation to reconstruction parameters and noise-equivalent counts (NEC). METHODS: This study consisted of a phantom study and a retrospective analysis of 39 consecutive patients who underwent clinical (18)F-FDG PET/CT. The body phantom of the National Electrical Manufacturers Association and International Electrotechnical Commission with a 10-mm-diameter sphere was filled with an (18)F-FDG solution with a 4:1 radioactivity ratio compared with the background. The PET data were reconstructed with the baseline ordered-subsets expectation maximization (OSEM) algorithm, with the OSEM+PSF model, with the OSEM+TOF model, and with the OSEM+PSF+TOF model. We evaluated image quality by visual assessment, the signal-to-noise ratio of the 10-mm sphere (SNR(10 mm)), the contrast of the 10-mm sphere, and the coefficient of variance in the phantom study and then determined the optimal reconstruction parameters. We also examined the effects of PSF and TOF on the quality of clinical images using the signal-to-noise ratio in the liver (SNR(liver)) in relation to the NEC in the liver (NEC(liver)). RESULTS: In the phantom study, the SNR(10 mm) was the highest for the OSEM+PSF+TOF model, and the highest value was obtained at iteration 2 for algorithms with the TOF and at iteration 3 for those without the TOF. In terms of a postsmoothing filter full width at half maximum (FWHM), the high SNR(10 mm) was obtained with no filtering or was smaller than 2 mm for algorithms with PSF and was 4-6 mm for those without PSF. The balance between the contrast recovery and noise is different for algorithms with either PSF or TOF. A combination of PSF and TOF improved SNR(10 mm), contrast, and coefficient of variance, especially with a small-FWHM gaussian filter. In the clinical study, the SNR(liver) of the low-NEC(liver) group in the OSEM+PSF+TOF model was compared with that of the high-NEC(liver) group in conventional OSEM. The PSF+TOF improved the SNR(liver) by about 24.9% ± 9.81%. CONCLUSION: A combination of PSF and TOF clearly improves image quality, whereas optimization of the reconstruction parameters is necessary to obtain the best performance for PSF or TOF. Furthermore, this combination has the potential to provide good image quality with either lower activity or shorter acquisition time, thus improving patient comfort and reducing the radiation burden.

Determination of the optimal acquisition protocol of breath-hold PET/CT for the diagnosis of thoracic lesions.

OBJECTIVE: The aim of this study was to determine the optimal acquisition scan protocol for deep inspiration breath-hold (BH) fluoro-2-deoxy-D-glucose positron emission tomography (PET) for the examination of thoracic lesions. METHODS: We studied 32 thoracic lesions in 21 patients. Whole-body PET/computed tomography (CT) scanning with free breathing (FB) was performed for 3 min per bed position, followed by a BH-CT and five BH-PET for 20 s each. Summed BH images with total acquisition times of 40, 60, 80 and 100 s were generated (BH × 2, BH × 3, BH × 4 and BH × 5, respectively). The displacements between PET and CT images, the lesion volume of the PET image, the maximum standardized uptake value (SUVmax) and the quality of the PET image were assessed in relation to the clinical characteristics of each patient and the summation of the BH-PET images. RESULTS: BH-PET decreased the tumor volume significantly (FB: 7.23 ± 9.70 cm³, BH × 5: 4.71 ± 5.14 cm³, P<0.01) and increased the SUVmax (FB: 6.27 ± 5.41, BH × 5: 7.53 ± 6.28, P<0.01). The displacement between the PET and CT images was improved significantly in the BH scans (FB: 0.77 ± 0.53 cm, BH × 5: 0.36 ± 0.24 cm, P<0.01). In addition, aging and lung function of patients influenced the reproducibility of BH-PET. The summed BH-PET images, obtained by summation of three or more BH-PET images (total acquisition time of 60 s or more), achieved good image quality. CONCLUSION: BH-PET/CT improved the misregistration between PET and CT images and increased the SUVmax of thoracic lesions. The recommended number of BH-PET images for summation with 20 s of acquisition time is three or more.

A simple table lookup method for PET/CT partial volume correction using a point-spread function in diagnosing lymph node metastasis.

OBJECTIVE: We evaluated the partial volume effect in PET/CT images and developed a simple correction method to address this problem. METHODS: Six spheres and the background in the phantom were filled with F-18 and we thus obtained 4 different sphere-to-background (SB) ratios. Thirty-nine cervical lymph nodes in 7 patients with papillary thyroid carcinoma (15 malignant and 24 benign) were also examined as a preliminary clinical study. First, we developed recovery coefficient (RC) curves normalized to the maximum counts of the 37-mm sphere. Next, we developed a correction table to determine the true SB ratio using three parameters, including the maximum counts of both the sphere and background and the lesion diameter, by modifying the approximation formula of the RC curves including the point-spread function correction. The full width at half maximum in this formula is estimated with the function of the SB ratio. RESULTS: In the phantom study, a size-dependent underestimation of the radioactivity was observed. The degree of decline of RC was influenced by the SB ratio. In preliminary clinical examination, the difference in the SUV(max) between malignant and benign LNs thus became more prominent after the correction. The PV correction slightly improved the diagnostic accuracy from 95 to 100%. CONCLUSIONS: We developed a simple table lookup correction method for the partial volume effect of PET/CT. This new method is considered to be clinically useful for the diagnosis of cervical LN metastasis. Further examination with a greater number of subjects is required to corroborate its clinical usefulness.

Phantom study on three-dimensional target volume delineation by PET/CT-based auto-contouring.

OBJECTIVE: The aim of this study was to determine an appropriate threshold value for delineation of the target volume in PET/CT and to investigate whether we could delineate a target volume by phantom studies. METHODS: A phantom consisted of six spheres (phi 10-37 mm) filled with 18F solution. Data acquisition was performed PET/CT in non-motion and motion status with high 18F solution and in non-motion status with low 18F solution. In non-motion phantom experiments, we determined two types of threshold value, an absolute SUV (T(SUV)) and a percentage of the maximum SUV (T%). Delineation using threshold values was applied for all spheres and for selected large spheres (a diameter of 22 mm or larger). In motion phantom experiments, data acquisition was performed in a static mode (sPET) and a gated mode (gPET). CT scanning was performed with helical CT (HCT) and 4-dimensional CT (4DCT). RESULTS: The appropriate threshold values were aT% = 27% and aT(SUV) = 2.4 for all spheres, and sT% = 30% and sT(SUV) = 4.3 for selected spheres. For all spheres in sPET/HCT in motion, the delineated volumes were 84%-129% by the aT% and 34%-127% by the aT(SUV). In gPET/4DCT in motion, the delineated volumes were 94-103% by the aT% and 51-131% by the aT(SUV). For low radioactivity spheres, the delineated volumes were all underestimated. CONCLUSION: A threshold value of T% = 27% was proposed for auto-contouring of lung tumors. Our results also suggested that the respiratory gated data acquisition should be performed in both PET and CT for target volume delineation.