ABSTRACT
Objective:To compare the effect of different β values on the semi-quantitative accuracy and image quality of 68Ga-prostate specific membrane antigen (PSMA) PET/CT imaging after partial volume effect correction (PVC). Methods:In the model experiment, image reconstruction was carried out based on block sequential regularized expectation maximization algorithm (BSREMA) with the range of β values from 100 to 1 000. Recovery coefficient (RC), contrast recovery (CR) and background variability (BV) were measured to evaluate semi-quantitative accuracy and image quality. In the clinical study, image data of 21 prostate cancer patients (age 45-78 years) who underwent 68Ga-PSMA PET/CT examination in the First Hospital of Shanxi Medical University from March 2019 to February 2020 were retrospectively collected. A total of 29 abdominal imaging positive lymph nodes were divided into the small lymph node group (diameter <10 mm; n=12) and the large lymph node group (10 mm≤diameter≤30 mm; n=17). SUV parameters including SUV max, SUV mean and peak of SUV (SUV peak) and the influence of different β values on the SUV parameters were evaluated. The signal to noise ratio (SNR) and subjective scores were used to evaluate image quality. Independent-sample t test, Kappa test and Pearson correlation analysis were used to analyze data. Results:The model experiment showed that CR, RC and BV decreased with the increase of β values. The image quality, image clarity, lesion significance, and total image scores given by nuclear medicine physicians showed strong consistency ( Kappa values: 0.65-0.87, P values: 0.026-0.043). The small lymph node group had the highest score (13 and 14) with β value of 600, while the large lymph node group had the highest score (13 and 14) with β value of 700. SNR of the two groups increased steadily within β values from 100 to 600 ( t values: 2.49-8.99, P values: 0.023-0.038). When the β value was higher than 600, SNR of the small lymph node group reached a plateau ( t values: 1.28-2.00, P values: 0.072-0.098), while the SNR of the large lymph node group continued to increase ( t values: 2.98-4.63, P values: 0.012-0.029). Before PVC, there were significant negative correlations between SUV parameters and β values ( r values: from -0.94 to -0.64, P values: 0.039-0.046). After PVC, it was found that SUV mean and SUV max still had significant negative correlation with β values ( r values: from -0.78 to -0.68, P values: 0.035-0.042), while the SUV peak showed no significant correlation with β values ( r values: -0.22, -0.28, P values: 0.069, 0.126). Conclusions:Based on subjective scores and semi-quantitative indicators, 68Ga-PSMA PET/CT is superior to select β values of 600 and 700 for image reconstruction based on BSREMA. The SUV peak of small lesions is stable after PVC and the clinical value should be explored in further.
ABSTRACT
Objective:To improve the quality of 18F-fluorodeoxyglucose (FDG) PET images at different acquisition times through deep learning (DL) PET image reconstruction methods. Methods:A total of 45 patients (20 males, 25 females; age (52.0±13.6) years) with malignant tumors and PET/CT scans from September 2020 to October 2020 in the Department of Nuclear Medicine of the First Hospital of Shanxi Medical University were included in this retrospective study. The short acquisition time 30 s/bed PET images from the raw list mode were selected as the input of DL model. DL image reconstruction model, based on the Unet algorithm, was trained to output imitated PET images with full dose standard acquisition time (3 min). The image quality evaluation and quantitative analysis were carried out for four groups of images: DL images, 30 s, 90 s, and 120 s images, respectively. The quality of PET images in four groups was evaluated using the five-point method. Liver background activities, lesions quantification parameters (maximum standardized uptake value (SUV max), mean standardized uptake value (SUV mean), standard deviation (SD), signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR)), and first-order texture features (skewness, kurtosis, uniformity, entropy) were measured. Kappa test, χ2 test and one-way analysis of variance (least significant difference t test) were used for data analysis. Results:The image quality scores between four groups were highly consistent ( Kappa=0.799, P<0.001). The number of patients with scores≥3 in DL, 30 s, 90 s and 120 s groups were 6, 4, 7 and 8, respectively ( χ2=125.47, P<0.001). The liver SD of DL group was significantly lower than that of 30 s group (0.26±0.07 vs 0.43±0.11; F=3.58, t=-7.91, P<0.05). The liver SNR of DL group was higher than that of 30 s group (11.04±4.36 vs 5.41±1.41; F=10.22, t=5.40, P<0.05). The liver SD and SNR of DL group were similar to those of 90 s group (0.39±0.16, 8.46±3.34; t values: -0.87 and 2.17, both P>0.05). In 18 tumor lesions with high uptake, SNR and CNR of DL group were significantly higher than those of 30 s group (60.21±29.26 vs 38.38±16.54, 22.26±15.85 vs 15.41±9.51; F values: 13.09 and 7.05; t values: 5.20 and 4.04, both P<0.001). There were statistically significant differences among four groups in the first-order texture features ( F values: 4.30-9.65, all P<0.05), but there was no significant difference between DL group and 120 s group ( t values: from -1.25 to 0.15, all P>0.05). Conclusion:DL reconstruction model can improve the quality of short-frame PET images, which meets the needs of clinical diagnosis, efficacy evaluation and radiomics research.
ABSTRACT
Objective:To compare four reconstruction algorithms of 18F-fluorodeoxyglucose (FDG) PET/CT on standardized uptake value (SUV) of pulmonary nodules. Methods:A total of 46 patients (27 males, 19 females; median age: 66 (range: 44-82) years) with solid pulmonary nodules from February 2018 to July 2019 in the First Hospital of Shanxi Medical University who performed 18F-FDG PET/CT imaging were enrolled. All PET/CT images were retrospectively reconstructed by using four algorithms reconstructions including ordered subset expectation maximization (OSEM), OSEM+ time of flight (TOF), OSEM+ TOF+ point spread function (PSF) and block sequential regularized expectation maximization (BSREM) (G1-G4). Nodule and background parameters were analyzed semi-quantitatively and visually. The maximum of SUV(SUV max), mean of SUV(SUV mean) and peak of SUV (SUV peak) were collected by the region of interest (ROI). Nodules were divided into small nodule group (diameter ≤10 mm) and large nodule group (10 mm < diameter ≤30 mm). Kruskal-Wallis rank sum test and Bonferroni method were performed to compare the differences of SUVs between G1-G4, and Spearman correlation analysis was used to analyze the correlation between the change rate of SUV (%ΔSUV) and the diameter of nodules. The receiver operating characteristic (ROC) curve analysis was used to analyze the diagnostic efficacy of SUV for the differential diagnosis of pulmonary nodules and to get the optimal threshold. Results:There were 114 pulmonary nodules (large nodules, n=55; small nodules, n=59). In visual analysis, the visual detection rates of small nodules in G4 were 55.93%(33/59), 44.07%(26/59), 20.34%(12/59) higher than those in G1-G3. Of 114 pulmonary nodules in 46 patients, there were differences in SUV max and SUV mean between G1-G4 (median SUV max : 2.65-5.29, median SUV mean: 2.05-2.99; H values: 20.628 and 17.749, respectively, both P<0.001), G4 had significant increases compared to G1 in SUV max (median 5.29 and 2.65, P<0.001) and SUV mean (median 2.99 and 2.05, P<0.001). The %ΔSUV max (median: 4.45%-52.96%) and %ΔSUV mean (median: 1.69%-47.56%) were negatively correlated with the diameter of nodules (9.75(6.20, 16.58) mm; r s values: -0.371 to -0.354, -0.371 to -0.320, all P<0.001). In 59 small nodules, G1 significantly increased the SUV max of G4 (median 4.05 and 2.14, H=18.327, P<0.001), while G4 significantly increased the SUV mean of G1 and G3 (median 2.31, 1.26 and 1.53, H=16.808, P<0.05). There was no significant difference in SUVs between G1-G4 in 55 large nodules ( H values: 0.812-7.290, all P>0.05). The optimal threshold values of SUV max in G1-G4 were 4.335, 5.185, 5.410, 5.745 and the area of under curves (AUCs) were 0.747, 0.699, 0.756, 0.778 respectively. The AUC of SUV mean and SUV peak also showed a similar trend. Conclusion:Among the four reconstruction algorithms, BRERM can not only enhance the image quality, but also significantly improve the SUV max and SUV mean of lung nodules diameter below 10 mm, and thus its diagnostic threshold of SUV should be appropriately increased.