Point Spread Function Reconstruction Research Articles

PRIME: PSF Reconstruction and Identification for Multiple-source characterization Enhancement – application to Keck NIRC2 imager

ABSTRACT In order to enhance the scientific exploitation of adaptive optics (AO)-assisted observations, we investigate a novel hybrid concept to improve the parametric estimation of point spread function (PSF) called PSF Reconstruction and Identification for Multiple-source characterization Enhancement (PRIME). PRIME uses both focal and pupil-plane measurements to estimate jointly the model parameters related to the atmosphere [$C_n^2(h)$, seeing] and the AO system (e.g. optical gains and residual low-order errors). Photometry and astrometry are provided as by-products. The parametric model in use is flexible enough to be scaled with field location and wavelength, making it a proper choice for optimized on-axis and off-axis data-reduction across the spectrum. Here, we present the methodology and validate PRIME on engineering and binary Keck II telescope NIRC2 images. We also present applications of PSF model parameters retrieval using PRIME: (i) calibrate the PSF model for observations void of stars on the acquired images, i.e. optimize the PSF reconstruction process, (ii) update the AO error breakdown mutually constrained by the telemetry and the images in order to speculate on the origin of the missing error terms and evaluate their magnitude, and (iii) measure photometry and astrometry with an application to the triple system Gl569 images.

Impact of Point-Spread Function Reconstruction on 68Ga-DOTATATE PET/CT Quantitative Imaging Parameters.

OBJECTIVE. The objective of our study was to investigate the impact of point-spread function (PSF) reconstruction and lesion size on 68Ga-tetraazacyclododecanetetraacetic acid-DPhe1-Tyr3-octreotate (DOTATATE) PET/CT quantitative parameters. MATERIALS AND METHODS. A total of 38 patients with 42 68Ga-DOTATATE PET/CT studies and 125 lesions were included. Scans were reconstructed with and without PSF modulation. For each lesion, the maximum standardized uptake value (SUVmax) and peak standardized uptake value (SUVpeak), metabolic tumor volume (MTV), total lesion somatostatin avidity, and tumor somatostatin receptor expression heterogeneity using the AUC method were measured. Intraclass correlation coefficient (ICC) and Bland-Altman analyses were used to compare PSF and non-PSF values. A subgroup analysis was performed to determine the impact of lesion size. RESULTS. Of the 125 lesions, 51 were in the liver, 31 in lymph nodes, 17 in bone, eight in pancreas, four in lung, and 14 in other sites. The ICCs between PSF and non-PSF values were excellent for SUVmax, SUVpeak, MTV, and total lesion somatostatin avidity (ICC = 0.97-0.99), and the ICC was good for tumor somatostatin receptor expression heterogeneity (ICC = 0.81). Comparison of PSF with non-PSF values showed a bias (mean percentage change ± SD) of 27.5% ± 14.7% for SUVmax, 15.5% ± 9.5% for SUVpeak, -18.6% ± 37.6% for MTV, 0.8% ± 28.1% for total lesion somatostatin avidity, and -7.1% ± 11.0% for tumor somatostatin receptor expression heterogeneity. Comparison of PSF with non-PSF values for lesions less than 2 cm (n = 75) showed corresponding biases greater than those for lesions 2 cm or larger (n = 50). CONCLUSION. PSF reconstruction effected higher values for SUVmax and SUVpeak, produced decreased values for tumor somatostatin receptor expression heterogeneity, and had a variable effect on MTV and total lesion somatostatin avidity depending on lesion size.

Testing Shear Recovery with Field Distortion

Abstract The tilt, rotation, or offset of each CCD with respect to the focal plane, as well as the distortion of the focal plane itself, causes shape distortions in the observed objects, an effect typically known as field distortion (FD). We point out that FD provides a unique way of quantifying the accuracy of cosmic shear measurement. The idea is to stack the shear estimators from galaxies that share similar FD-induced shape distortions. Given that the latter can be calculated with parameters from astrometric calibrations, the accuracy of the shear estimator can be directly tested on real images. It provides a way to calibrate the multiplicative and additive shear recovery biases within the scientific data itself, without requiring simulations or any external data sets. We use the CFHTLenS images to test the Fourier_Quad shear recovery method. We highlight some details in our image-processing pipeline, including background removal, source identification and deblending, astrometric calibration, star selection for point-spread function reconstruction, noise reduction, etc. We show that in the shear ranges of −0.005 ≲ g 1 ≲ 0.005 and −0.008 ≲ g 2 ≲ 0.008, the multiplicative biases are at the level of ≲0.04. Slight additive biases on the order of ∼5 × 10−4 (6σ) are identified for sources provided by the official CFHTLenS catalog (not using its shear catalog), but are minor (4σ) for the source catalog generated by our Fourier_Quad pipeline.

Point Spread Function Reconstruction for Integrated 18F-FET PET/MRI in Patients With Glioma: Does It Affect SUVs and Respective Tumor-to-Background Ratios?

Semiquantitative F-FET PET assessment using the tumor's SUV or tumor-to-background ratios (TBRs) can separate gliomas from peritumoral tissue or progression from pseudoprogression. This study investigated if point spread function (PSF) reconstruction of F-FET PET data affects SUV-based dignity assessment. This study is a retrospective analysis of 87 glioma patients (female, 36; male, 51; age, 48 [13-81] years) undergoing F-FET PET/MRI for staging (n = 17) or restaging (n = 70). PET was reconstructed using ordered-subset expectation maximization with and without PSF. Lesions were delineated with semiautomated background-adapted thresholding relative to SUVmax; background was delineated contralaterally. Comparative measurements with a National Electrical Manufacturers Association International Electrotechnical Commission PET body phantom (sphere-to-background ratios, 8:1 and 4:1) were performed. PSF showed significantly higher tumor SUVmax (median difference, +0.1; interquartile range, 0.04-0.18), SUVmean (+0.05; 0.03-0.08), TBRmax|mean (+0.1; 0.04-0.2), and TBRmean|mean (+0.06; 0.03-0.09) than non-PSF (P < 0.001). Background SUVmean was unaffected. In patients and phantom, differences between PSF and non-PSF increased with TBR and decreased with lesion's PET volume. Differences only exceeded 0.2 SUV for SUVmax or 0.1 SUV for SUVmean if TBR was greater than 3 and lesion's PET volume was less than 10 mL (d = 27 mm). Dignity assessment by PSF and non-PSF was concordant in all patients examined for staging (cutoff, TBRmean|mean > 1.6; positive, 14; negative, 3) and restaging (cutoff, TBRmax|mean > 2.0; positive, 67; negative, 3). PSF increased tumor SUVmax and SUVmean compared with non-PSF F-FET PET/MRI data, especially in small lesions with high TBR (>3). However, dignity assessment using established TBR cutoffs was not affected.

基于PSF重构和改进的最大后验估计的自适应光学图像复原算法

基于PSF重构和改进的最大后验估计的自适应光学图像复原算法

331. Performance of point spread function reconstruction algorithm on a new PET-CT TOF scanner

Purpose Aim of this study was to evaluate the performance of PSF reconstruction algorithm of a PET-TC TOF scanner. Comparison in term of image quality was performed in order to investigate potentialities and differences between PSF algorithm and the standard OSEM algorithm routinely applied in the clinical use. Methods The phantom NEMA 2012 Body Image Quality (18F, 213 kBq/ml, □ 10 mm 13 mm 17 mm 22 mm hot spheres (background ratio 4 ÷ 1 ), □ 28 mm 37 mm cold spheres) was acquired with Philips IngenuityTF PET-TC scanner (TOF, AC-SC). Raw data were reconstructed both with standard OSEM 3D and with PSF, using 20 different paired iteration values (1–500) and regularization levels (1–20). Seven regions of interest (ROIs) were contoured: 4 centred on “hot” spheres, 2 on “cold” spheres and 1 on background. Mean value (MV) and standard deviation (SD) of every ROI were calculated: signal to noise ratio (SNR) and contrast to noise ratio (CNR) were compared. Spatial resolution was evaluated as steepness of the 10 mm-sphere profile. Results PSF images showed resolution higher (e.g. 3 mm for 30 iterations) than the standard OSEM (6 mm). As shown in figure 1a the more iteration number the higher artifacts presence, especially in the background region. MV and SD values relative to the “hot spheres” strongly depend on iteration number and regularization level; variations up to +1000% were found (500 iterations and 20 regularization) both on MV and SD values. No variations were noticed on background MV, but SD values increase up to +600%. As consequence SNR and CNR are affected by reconstruction parameters (Fig. 1b). Conclusions PSF images showed resolution higher than standard OSEM, but also huge artifacts and unacceptable variations in measured activity (and SUV). The use of PSF reconstruction for quantitative purposes is not recommended and its use should be designed for specific uses.

Optimization of a dedicated protocol using a small-voxel PSF reconstruction for head-and-neck 18FDG PET/CT imaging in differentiated thyroid cancer

Background18FDG PET/CT is crucial before neck surgery for nodal recurrence localization in iodine-refractory differentiated or poorly differentiated thyroid cancer (DTC/PDTC). A dedicated head-and-neck (HN) acquisition performed with a thin matrix and point-spread-function (PSF) modelling in addition to the whole-body PET study has been shown to improve the detection of small cancer deposits. Different protocols have been reported with various acquisition times of HN PET/CT. We aimed to compare two reconstruction algorithms for disease detection and to determine the optimal acquisition time per bed position using the Siemens Biograph6 with extended field-of-view.MethodsTwenty-one consecutive and unselected patients with DTC/PDTC underwent HN PET/CT acquisition using list-mode. PET data were reconstructed, mimicking five different acquisition times per bed position from 2 to 10 min. Each PET data set was reconstructed using 3D-ordered subset expectation maximisation (3D-OSEM) or iterative reconstruction with PSF modelling with no post filtering (PSFallpass). These reconstructions resulted in 210 anonymized datasets that were randomly reviewed to assess 18FDG uptake in cervical lymph nodes or in the thyroid bed using a 5-point scale. Noise level, maximal standard uptake values (SUVmax), tumour/background ratios (TBRs) and dimensions of the corresponding lesion on the CT scan were recorded. In surgical patients, the largest tumoral size of each lymph node metastasis was measured by a pathologist.ResultsThe 120 HN PET studies of the 12 patients with at least 1 18FDG focus scored malignant formed the study group. Noise level significantly decreased between 2 and 4 min for both 3D-OSEM and PSFallpass reconstructions (p < 0.01). TBRs were similar for all the acquisition times for both 3D-OSEM and PSFallpass reconstructions (p = 0.25 and 0.44, respectively). The detection rate of malignant foci significantly improved from 2 to 10 min for PSFallpass reconstruction (20/26 to 26/26; p = 0.01) but not for 3D-OSEM (15/26 to 19/26; p = 0.26). For each of the five acquisition times, PSFallpass detected more malignant foci than 3D-OSEM (p < 0.01). In the seven surgical patients, PSFallpass evidenced smaller malignant lymph nodes than 3D-OSEM at 8 and 10 min. At 10 min, the mean size of the lymph node metastases neither detected with PSFallpass nor 3D-OSEM was 3 ± 0.6 mm vs 5.8 ± 1.1 mm for those detected with PSFallpass only and 10.9 ± 3.3 for those detected with both reconstructions (p < 0.001).ConclusionsPSFallpass HN PET improves lesion detectability as compared with 3D-OSEM HN PET. PSFallpass with an acquisition time between 8 and 10 min provides the best performance for tumour detection.

How to harmonize SUVs obtained by hybrid PET/CT scanners with and without point spread function correction

State of the art point-spread function (PSF) corrections implemented in positron emission tomography/computed tomography (PET/CT) reconstruction improved image quality and diagnostic performance but caused an increase in the standardized uptake value (SUV) compared to a conventional OSEM reconstruction system. The EANM suggested one produce two reconstructions, one optimised for maximum lesion detection and one for semi-quantitative analysis. In this work we investigated an alternative methodology, using a single reconstruction data set together with a post-reconstruction algorithm for SUV harmonization. Data acquisition was performed on a Siemens Biograph mCT system equipped with lutetium oxyorthosilicat crystals, PSF and time-of-flight algorithms and on a General Electric Discovery STE system equipped with BGO crystals. Both a EANM double reconstruction method and a dedicated post-reconstruction algorithm (marketed as EQ-filter) were tested to harmonize the quantitative values of the two PET/CT scanners. For phantom measurements we used a NEMA IQ phantom and a Jaszczak cylindrical phantom equipped with small spheres (lesion to background ratios of 8:1 and 4:1). Several different reconstruction settings were tested in order to provide a general methodology. Data obtained by phantom measurements were validated on seven oncologic patients who performed a one-bed extra acquisition on a different scanner. The evaluation regarded 39 small lesions (diameters: 0.3–2.6 cm) and was performed by two experienced nuclear medicine physicians. The SUV recoveries measured with the PSF reconstruction exceeded those obtained by the OSEM reconstruction with deviations ranging from 16% to 150%. These discrepancies resulted below 7% applying the optimized value of the EQ.filter or the double-reconstruction methods. For each reconstruction setting the optimal value of the EQ.filter was identified in order to minimize these discrepancies. Patient data, analyzed by Wilcoxon statistical test, confirmed and validated phantom measurements. EQ.filter can harmonize SUV values between different PET/CT scanners using a single reconstruction optimized to maximum lesion detectability. In this way, the second reconstruction proposed by EANM/EARL is avoided.

Point spread function reconstruction for single-conjugate adaptive optics on extremely large telescopes

Modern ground-based telescopes like the planned extremely large telescope (ELT) depend heavily on adaptive optics (AO) systems to correct for atmospheric turbulence. Even though AO correction is used, the quality of astronomical images still is degraded due to the time delay stemming from the wavefront sensor integration time and temporal response of the deformable mirror(s) (DM). This results in a blur, which can be mathematically described by a convolution of the true image with the point spread function (PSF). We present an algorithm for single-conjugate adaptive optics PSF reconstruction adapted to the needs of ELTs in a storage efficient way. In particular, the classical PSF reconstruction algorithm is changed in several points to give a more accurate estimate for the post-AO PSF. Bilinear splines are used as basis functions to minimize the computational effort. Results obtained in an end-to-end simulation tool show qualitatively good reconstruction of the PSF compared with the PSF calculated directly from the simulated incoming wavefront. Furthermore, the used algorithm has a reasonable runtime and memory consumption.

Why harmonization is needed when using FDG PET/CT as a prognosticator: demonstration with EARL-compliant SUV as an independent prognostic factor in lung cancer.

To determine EARL-compliant prognostic SUV thresholds in a mature cohort of patients with locally advanced NSCLC, and to demonstrate how detrimental it is to use a threshold determined on an older-generation PET system with a newer PET/CT machine, and vice versa, or to use such a threshold with non-harmonized multicentre pooled data. This was a single-centre retrospective study including 139 consecutive stage IIIA-IIIB patients. PET data were acquired as per the EANM guidelines and reconstructed with unfiltered point spread function (PSF) reconstruction. Subsequently, a 6.3mm Gaussian filter was applied using the EQ.PET (Siemens Healthineers) methodology to meet the EANM/EARL harmonizing standards (PSFEARL). A multicentre study including non-EARL-compliant systems was simulated by randomly creating four groups of patients whose images were reconstructed with unfiltered PSF and PSF with Gaussian post-filtering of 3, 5, and 10mm. Identification of optimal SUV thresholds was based on a two-fold cross-validation process that partitioned the overall sample into learning and validation subsamples. Proportional Cox hazards models were used to estimate age-adjusted and multivariable-adjusted hazard ratios (HRs) and their 95% confidence intervals. Kaplan-Meier curves were compared using the log rank test. Median follow-up was 28months (1-104months). For the whole population, the estimated overall survival rate at 36months was 0.39 [0.31-0.47]. The optimal SUVmax cutoff value was 25.43 (95% CI: 23.41-26.31) and 8.47 (95% CI: 7.23-9.31) for the PSF and for the EARL-compliant dataset respectively. These SUVmax cutoff values were both significantly and independently associated with lung cancer mortality; HRs were 1.73 (1.05-2.84) and 1.92 (1.16-3.19) for the PSF and the EARL-compliant dataset respectively. When (i) applying the optimal PSF SUVmax cutoff on an EARL-compliant dataset and the optimal EARL SUVmax cutoff on a PSF dataset or (ii) applying the optimal EARL compliant SUVmax cutoff to a simulated multicentre dataset, the tumour SUVmax was no longer significantly associated with lung cancer mortality. The present study provides the PET community with an EARL-compliant SUVmax as an independent prognosticator for advanced NSCLC that should be confirmed in a larger cohort, ideally at other EARL accredited centres, and highlights the need to harmonize PET quantitative metrics when using them for risk stratification of patients.

Determination of the optical turbulence parameters from the adaptive optics telemetry: critical analysis and on-sky validation.

It has been demonstrated by several authors that the optical turbulence parameters associated with a given adaptive optics (AO) run-the seeing angle and outer scale-can be determined from a statistical analysis of the commands of the system's deformable mirror (DM). The higher the accuracy on these parameters, the more we can make use of them, allowing for instance a better estimation of the seeing statistics at the telescope location or a more accurate assessment of the performance of the AO system. In the context of a point spread function reconstruction project (PSF-R) for the W. M. Keck observatory AO system, we decided to identify, in the most exhaustive way, all the sources of systematic and random errors affecting the determination of the seeing angle and outer scale from the DM telemetry, and find ways to compensate/mitigate these errors to keep them under 10%. The seeing estimated using our improved DM-seeing method was compared with more than 70 nearly simultaneous seeing measurements from open-loop PSFs on the same optical axis, and with independent seeing-monitor measurements acquired at the same time but far from the telescope (DIMM/MASS): the correlation with the open-loop PSF is very good (the error is about 10%), validating the DM-seeing method for accurate seeing determination, while it is weak and sometimes completely uncorrelated with the DIMM/MASS seeing monitor data. We concluded that DM-based seeing can be very accurate if all the error terms are considered in the DM data processing, but that seeing taken from non-collocated seeing monitors is of no use even when moderate accuracy is required.

Evaluation of MR-based PET Images with a Homogeneous μmap by Using Integrated PET/MR

This study intends to obtain magnetic resonance (MR)-based positron emission tomography (PET) images with a relatively uniform μmap by using a National Electrical Manufacturers Association (NEMA) International Electrotechnical Commission (IEC) body phantom and to evaluate the PET phantom images by applying various reconstruction parameters in accordance with the protocol of the NEMA. In order to obtain a uniform MR-based μmap, we set the initial voltage of the RF transmitter adjustment algorithm to 75 V, and we obtained the attenuation-corrected PET image by applying the two-point vibe Dixon MR sequence. The images reconstructed using the 3D ordered subset expectation maximization (OSEM) algorithm with the point spread function (PSF) showed the highest percentage of contrast recovery and the lowest percentage background variability compared to the other 3D reconstruction algorithms like the Filtered backprojection (FBP), 3D OSEM, and 3D OSEM with PSF algorithms. The 3D OSEM with PSF reconstruction algorithm is confirmed to be effective for maintaining the optimum image quality.

Protocols for Harmonized Quantification and Noise Reduction in Low-Dose Oncologic 18F-FDG PET/CT Imaging.

Oncologic 18F-FDG PET/CT acquisition and reconstruction protocols need to be optimized for both quantitative and detection tasks. To date, most studies have focused on either quantification or noise, leading to quantitative harmonization guidelines or appropriate noise levels. We developed and evaluated protocols that provide harmonized quantitation with optimal amounts of noise as a function of acquisition parameters and body mass. Methods: Multiple image acquisitions (n = 17) of the International Electrotechnical Commission/National Electrical Manufacturers Association PET image-quality phantom were performed with variable counting statistics. Phantom images were reconstructed with 3-dimensional ordered-subset expectation maximization (OSEM3D) and point-spread function (PSF) for harmonized quantification of the contrast recovery coefficient of the maximum pixel value (CRC max ). The lowest counting statistics that resulted in compliance with European Association of Nuclear Medicine recommendations for CRC max and CRC max variability were used as optimization metrics. Image noise in the liver of 48 typical oncologic 18F-FDG PET/CT studies was analyzed with OSEM3D and PSF harmonized reconstructions. We also evaluated 164 additional 18F-FDG PET/CT reconstructed list-mode images to derive analytic expressions that predict image quality and noise variability. Phantom-to-subject translational analysis was used to derive optimized acquisition and reconstruction protocols. Results: For harmonized quantitation levels, PSF reconstructions yielded decreased noise and lower CRC max variability than regular OSEM3D reconstructions, suggesting they could enable a decreased activity regimen for matched performance. Conclusion: PSF reconstruction with a 7-mm postprocessing filter can provide harmonized quantification performance and acceptable image noise levels with injected activity, duration, and mass settings using a 260 MBq⋅s/kg acquisition parameter at scan time. Similarly, OSEM3D with a 5-mm postprocessing filter can provide similar performance with 401 MBq⋅s/kg.

Numerical estimation of wavefront error breakdown in adaptive optics

Aims. Adaptive optics (AO) system performance is improved using post-processing techniques, such as point spread function (PSF) deconvolution. The PSF estimation involves characterization of the different wavefront (WF) error sources in the AO system. We propose a numerical error breakdown estimation tool that allows studying AO error source behavior such as their correlations. We also propose a new analytical model for anisoplanatism and bandwidth errors that were validated with the error breakdown estimation tool. This model is the first step for a complete AO residual error model that is expressed in deformable mirror space, leading to practical usage such as PSF reconstruction or turbulent parameters identification. Methods. We have developed in the computing platform for adaptive optics systems (COMPASS) code, which is an end-to-end simulation code using graphics processing units (GPU) acceleration, an estimation tool that provides a comprehensive error breakdown by the outputs of a single simulation run. We derive the various contributors from the end-to-end simulator at each iteration step: this method provides temporal buffers of each contributor. Then, we use this tool to validate a new model of anisoplanatism and bandwidth errors including their correlation. This model is based on a statistical approach that computes the error covariance matrices using structure functions. Results. On a SPHERE-like system, the comparison between a PSF computed from the error breakdown with a PSF obtained from classical end-to-end simulation shows that the statistics convergence limits converge very well, with a sub-percent difference in terms of Strehl ratio and ensquared energy at 5λ/D separation. A correlation analysis shows significant correlations between some contributors, especially WF measurement deviation error and bandwidth error due to centroid gain, and the well-known correlation between bandwidth and anisoplanatism errors is also retrieved. The model we propose for the two latter errors shows an SR and EE difference of about one percent compared to the end-to-end simulation, even if some approximations exist.

P094] Performance evaluation of point spread function reconstruction algorithm on a new PET-TC TOF scanner

Purpose Aim of this study was to evaluate the performance of PSF reconstruction algorithm of a new PET-TC TOF scanner. Comparison in term of image quality was performed in order to investigate potentialities and differences between PSF algorithm and the standard OSEM algorithm routinely applied in the clinical use. Methods The phantom NEMA 2012 Body Image Quality (18 F, 213 kBq/ml, diameter 10 mm, 13 mm, 17 mm and 22 mm hot spheres (background ratio 4 ÷ 1), diameter 28 mm and 37 mm cold spheres) was acquired with Philips Ingenuity TF PET-TC scanner (TOF, AC-SC). Raw data were reconstructed both with standard OSEM 3D and with PSF, using 20 different paired iteration values (1–500) and regularization levels (1–20). Seven regions of interest (ROIs) were contoured: 4 centered on ”hot” spheres, 2 on ”cold” spheres and 1 on background. Mean value (MV) and standard deviation (SD) of every ROI were calculated: signal to noise ratio (SNR) and contrast to noise ratio (CNR) were compared. Spatial resolution was evaluated as steep of the 10 mm-sphere profile. Results PSF images showed resolution higher (e.g. 3 mm for 30 iterations) than the standard OSEM (6 mm). The more iteration number the higher artifacts presence, especially in the background region. MV and SD values relative to the ”hot spheres” strongly depend on iteration number and regularization level; variations up to +1000% were found (500 iterations and 20 regularization) both on MV and SD values. No variations were noticed on background MV, but SD values increase up to +600%. As consequence SNR and CNR are strongly affected by reconstruction parameters. Conclusions PSF images showed resolution higher than standard OSEM, but also huge artifacts and unacceptable variations in measured activity (and SUV). The use of PSF reconstruction for quantitative purposes is not recommended and its use should be designed for specific uses.

Low-dose High-resolution 18F-FDG-PET/CT Using Time-of-flight and Point-spread Function Reconstructions: A Role in the Detection of Breast Carcinoma Axillary Lymph Node Metastases.

to evaluate the performance of N-staging assessment in clinically-proven T1 breast carcinoma by high-resolution 18F-fluorodeoxyglucose-positron-emission tomography/computed tomography (18F-FDG-PET/CT) using time-of-flight with point-spread function reconstruction. In 30 women with clinically proven T1 breast carcinoma, imaging before surgery was performed using 18F-FDG-PET/CT. The results of PET/CT in detection of lymph node metastases were compared with those obtained after pathological investigation of axillary biopsy. A four-ring PET subsystem with image reconstruction using time-of-flight and point-spread function was used with the radiopharmaceutical dose reduction to 2.5 MBq/kg. Axillary lymph node metastasis was confirmed by histology in 13 patients, but metastasis was suspected based on PET/CT in 12 of those patients, the absence of metastasis was surgically confirmed in 17 women, 15 of which were suspected based on PET/CT. The sensitivity for detection of axillary lymph node metastasis was 93.3%, with a specificity of 88.2% in the whole patient cohort. Additionally, distant metastatic spread was found in 13.3% of patients. The reconstruction of PET images with time-of-flight and point-spread function enabled the improvement of diagnostic performance in N-staging of breast carcinoma, even when the dose of radiopharmaceutical was reduced to 2.5 MBq/kg.

An Accurate Centroiding Algorithm for PSF Reconstruction

Abstract In this work, we present a novel centroiding method based on Fourier space Phase Fitting (FPF) for Point Spread Function (PSF) reconstruction. We generate two sets of simulations to test our method. The first set is generated by GalSim with an elliptical Moffat profile and strong anisotropy that shifts the center of the PSF. The second set of simulations is drawn from CFHT i band stellar imaging data. We find non-negligible anisotropy from CFHT stellar images, which leads to ∼0.08 scatter in units of pixels using a polynomial fitting method (Vakili &amp; Hogg). When we apply the FPF method to estimate the centroid in real space, the scatter reduces to ∼0.04 in S/N = 200 CFHT-like sample. In low signal-to-noise ratio (S/N; 50 and 100) CFHT-like samples, the background noise dominates the shifting of the centroid; therefore, the scatter estimated from different methods is similar. We compare polynomial fitting and FPF using GalSim simulation with optical anisotropy. We find that in all S/N (50, 100, and 200) samples, FPF performs better than polynomial fitting by a factor of ∼3. In general, we suggest that in real observations there exists anisotropy that shifts the centroid, and thus, the FPF method provides a better way to accurately locate it.

Diagnostic performance of 18F-FDG PET/CT using point spread function reconstruction on initial staging of rectal cancer: a comparison study with conventional PET/CT and pelvic MRI

BackgroundAccurate staging is crucial for treatment selection and prognosis prediction in patients with rectal cancer. Point spread function (PSF) reconstruction can improve spatial resolution and signal-to-noise ratio of PET imaging. The aim of this study was to evaluate the effectiveness of 18F-FDG PET/CT with PSF reconstruction for initial staging in rectal cancer compared with conventional PET/CT and pelvic MRI.MethodsA total of 59 patients with rectal cancer underwent preoperative 18F-FDG PET/CT and pelvic MRI. The maximum standardized uptake value (SUVmax) and lesion to background (L/B) ratio of possible metastatic lymph nodes, and metabolic tumor volumes (MTVs) of primary tumors were calculated. For N and T (T1-2 vs T3-4) staging, sensitivities, specificities, positive predictive values, negative predictive values, and accuracies were compared between conventional PET/CT [reconstructed with ordered subset expectation maximization (OSEM)], PSF-PET/CT (reconstructed with OSEM+PSF), and pelvic MRI. Histopathologic analysis was the reference standard.ResultsFor N staging, PSF-PET/CT provided higher sensitivity (78.6%) than conventional PET/CT (64.3%), and pelvic MRI (57.1%), and all techniques showed high specificity (PSF-PET: 95.4%, conventional PET: 96.7%, pelvic MRI: 93.5%). SUVmax and L/B ratio were significantly higher in PSF-PET/CT than conventional-PET/CT (p < 0.001). The accuracy for T staging in PSF-PET/CT (69.4%) was not significantly different to conventional PET/CT (73.5%) and pelvic MRI (73.5%). MTVs of PSF and conventional PET showed a significant difference among T stages (p < 0.001), with higher values in advanced stages. In M staging, both PSF and conventional PET/CT diagnosed all distant metastases correctly.ConclusionsPSF-PET/CT produced images with higher lesion-to-background contrast than conventional PET/CT, which allowed improved detection of lymph node metastasis without compromising specificity, and showed comparable diagnostic value to MRI in local staging. PSF-PET/CT is likely to have a great value for initial staging in rectal cancer.

PSF reconstruction for Compton-based prompt gamma imaging

Compton-based prompt gamma (PG) imaging has been proposed for in vivo range verification in proton therapy. However, several factors degrade the image quality of PG images, some of which are due to inherent properties of a Compton camera such as spatial resolution and energy resolution. Moreover, Compton-based PG imaging has a spatially variant resolution loss. In this study, we investigate the performance of the list-mode ordered subset expectation maximization algorithm with a shift-variant point spread function (LM-OSEM-SV-PSF) model. We also evaluate how well the PG images reconstructed using an SV-PSF model reproduce the distal falloff of the proton beam. The SV-PSF parameters were estimated from simulation data of point sources at various positions. Simulated PGs were produced in a water phantom irradiated with a proton beam. Compared to the LM-OSEM algorithm, the LM-OSEM-SV-PSF algorithm improved the quality of the reconstructed PG images and the estimation of PG falloff positions. In addition, the 4.44 and 5.25 MeV PG emissions can be accurately reconstructed using the LM-OSEM-SV-PSF algorithm. However, for the 2.31 and 6.13 MeV PG emissions, the LM-OSEM-SV-PSF reconstruction provides limited improvement. We also found that the LM-OSEM algorithm followed by a shift-variant Richardson–Lucy deconvolution could reconstruct images with quality visually similar to the LM-OSEM-SV-PSF-reconstructed images, while requiring shorter computation time.

Bayesian penalised likelihood reconstruction (Q.Clear) of 18F-fluciclovine PET for imaging of recurrent prostate cancer: semi-quantitative and clinical evaluation.

Objective:18F-Fluciclovine (FACBC) is an amino acid PET radiotracer approved for recurrent prostate cancer imaging. We investigate the use of Bayesian penalised likelihood (BPL) reconstruction for 18F-fluciclovine PET.Methods:15 18F-fluciclovine scans were reconstructed using ordered subset expectation maximisation (OSEM), OSEM + point spread function (PSF) modelling and BPL using β-values 100–600. Lesion maximum standardised uptake value (SUVmax), organ SUVmean and standard deviation were measured.Deidentified reconstructions (OSEM, PSF, BPL using β200–600) from 10 cases were visually analysed by two readers who indicated their most and least preferred reconstructions, and scored overall image quality, noise level, background marrow image quality and lesion conspicuity.Results:Comparing BPL to OSEM, there were significant increments in lesion SUVmax and signal-to-background up to β400, with highest gain in β100 reconstructions (mean ΔSUVmax 3.9, p < 0.0001). Organ noise levels increased on PSF, β100 and β200 reconstructions. Across BPL reconstructions, there was incremental reduction in organ noise with increasing β, statistically significant beyond β300–500 (organ-dependent). Comparing with OSEM and PSF, lesion signal-to-noise was significantly increased in BPL reconstructions where β ≥ 300 and ≥ 200 respectively.On visual analysis, β 300 had the first and second highest scores for image quality, β500 and β600 equal highest scores for marrow image quality and least noise, PSF and β 200 had first and second highest scores for lesion conspicuity. For overall preference, one reader preferred β 300 in 9/10 cases and the other preferred β 200 in all cases.Conclusion:BPL reconstruction of 18F-fluciclovine PET images improves signal-to-noise ratio, affirmed by overall reader preferences. On balance, β300 is suggested for 18F-fluciclovine whole body PET image reconstruction using BPL.Advances in knowledge:The optimum β is different to that previously published for 18F-fluorodeoxyglucose, and has practical implications for a relatively new tracer in an environment with modern reconstruction technologies.