Point Spread Function Reconstruction Research Articles

Adaptive optics telemetry standard

Context. The amount of adaptive optics (AO) telemetry generated by visible/near-infrared ground-based observatories is ever greater, leading to a growing need for a standardised data exchange format to support performance analysis, AO research, and development activities that involve large-scale telemetry mining, processing, and curation. Aims. This paper introduces the Adaptive Optics Telemetry (AOT) data exchange format as a standard for sharing AO telemetry from visible/infrared ground-based observatories. AOT is based on the flexible image transport system (FITS) and aims to provide unambiguous and consistent data access across various systems and configurations, including natural and single- or multiple-laser guide-star AO systems. Methods. We designed AOT with a focus on two key use cases: atmospheric turbulence parameter estimation and point-spread function reconstruction. We prototyped and tested the design using existing AO telemetry datasets from multiple systems: single conjugate with natural and laser guide stars, tomographic systems with multi-channel wavefront sensors, and single- and multi-wavefront correctors in systems featuring either a Shack-Hartmann or Pyramid as the main wavefront sensor. Results. The AOT file structure has been thoroughly defined, with specified data fields, descriptions, data types, units, and expected dimensions. To support this format, we have developed a Python package that enables the data conversion, reading, writing, and exploration of AOT files; it has been made publicly available and is compatible with a general-purpose Python package manager. We have demonstrated the flexibility of the AOT format by packaging data from five different instruments, installed on different telescopes.

Model-based iterative reconstruction for direct imaging with point spread function encoded echo planar MRI

BackgroundEcho planar imaging (EPI) is a fast measurement technique commonly used in magnetic resonance imaging (MRI), but is highly sensitive to measurement non-idealities in reconstruction. Point spread function (PSF)-encoded EPI is a multi-shot strategy which alleviates distortion, but acquisition of encodings suitable for direct distortion-free imaging prolongs scan time. In this work, a model-based iterative reconstruction (MBIR) framework is introduced for direct imaging with PSF-EPI to improve image quality and acceleration potential. MethodsAn MBIR platform was developed for accelerated PSF-EPI. The reconstruction utilizes a subspace representation, is regularized to promote local low-rankedness (LLR), and uses variable splitting for efficient iteration. Comparisons were made against standard reconstructions from prospectively accelerated PSF-EPI data and with retrospective subsampling. Exploring aggressive partial Fourier acceleration of the PSF-encoding dimension, additional comparisons were made against an extension of Homodyne to direct PSF-EPI in numerical experiments. A neuroradiologists' assessment was completed comparing images reconstructed with MBIR from retrospectively truncated data directly against images obtained with standard reconstructions from non-truncated datasets. ResultsImage quality results were consistently superior for MBIR relative to standard and Homodyne reconstructions. As the MBIR signal model and reconstruction allow for arbitrary sampling of the PSF space, random sampling of the PSF-encoding dimension was also demonstrated, with quantitative assessments indicating best performance achieved through nonuniform PSF sampling combined with partial Fourier. With retrospective subsampling, MBIR reconstructs high-quality images from sub-minute scan datasets. MBIR was shown to be superior in a neuroradiologists' assessment with respect to three of five performance criteria, with equivalence for the remaining two. ConclusionsA novel image reconstruction framework is introduced for direct imaging with PSF-EPI, enabling arbitrary PSF space sampling and reconstruction of diagnostic-quality images from highly accelerated PSF-encoded EPI data.

Visualization of small brain nuclei with a high-spatial resolution, clinically available whole-body PET scanner

ObjectiveTo verify the visibility of physiological 18F-fluorodeoxyglucose (18F-FDG) uptake in nuclei in and around the brainstem by a whole-body (WB) silicon photomultiplier positron emission tomography (SiPM-PET) scanner with point-spread function (PSF) reconstruction using various iteration numbers.MethodsTen healthy subjects (5 men, 5 women; mean age, 56.0 ± 5.0 years) who underwent 18F-FDG PET/CT using a WB SiPM-PET scanner and magnetic resonance imaging (MRI) of the brain including a spin-echo three-dimensional sampling perfection with application-optimized contrasts using different flip angle evolutions fluid-attenuated inversion recovery (3D-FLAIR) and a 3D-T1 magnetization-prepared rapid gradient-echo (T1-MPRAGE) images were enrolled. Each acquired PET image was reconstructed using ordered-subset expectation maximization (OSEM) with iteration numbers of 4, 16, 64, and 256 (subset 5 fixed) + time-of-flight (TOF) + PSF. The reconstructed PET images and 3D-FLAIR images for each subject were registered to individual T1-MPRAGE volumes using normalized mutual information criteria. For each MR-coregistered individual PET image, the pattern of FDG uptake in the inferior olivary nuclei (ION), dentate nuclei (DN), midbrain raphe nuclei (MRN), inferior colliculi (IC), mammillary bodies (MB), red nuclei (RN), subthalamic nuclei (STN), lateral geniculate nuclei (LGN), medial geniculate nuclei (MGN), and superior colliculi (SC) was visually classified into the following three categories: good, clearly distinguishable FDG accumulation; fair, obscure contour of FDG accumulation; poor, FDG accumulation indistinguishable from surrounding uptake.ResultsAmong individual 18F-FDG PET images with OSEM iterations of 4, 16, 64, and 256 + TOF + PSF, the iteration numbers that showed the best visibility in each structure were as follows: ION, MRN, LGN, MGN, and SC, iteration 64; DN, iteration 16; IC, iterations 16, 64, and 256; MB, iterations 64 and 256; and RN and STN, iterations 16 and 64, respectively. Of the four iterations, the 18F-FDG PET image of iteration 64 visualized FDG accumulation in small structures in and around the brainstem most clearly (good, 98 structures; fair, 2 structures).ConclusionsA clinically available WB SiPM-PET scanner is useful for visualizing physiological FDG uptake in small brain nuclei, using a sufficiently high number of iterations for OSEM with TOF and PSF reconstructions.

PET imaging and quantification of small animals using a clinical SiPM-based camera

BackgroundSmall-animal PET imaging is an important tool in preclinical oncology. This study evaluated the ability of a clinical SiPM-PET camera to image several rats simultaneously and to perform quantification data analysis.MethodsIntrinsic spatial resolution was measured using 18F line sources, and image quality was assessed using a NEMA NU 4-2018 phantom. Quantification was evaluated using a fillable micro-hollow sphere phantom containing 4 spheres of different sizes (ranging from 3.95 to 7.86 mm). Recovery coefficients were computed for the maximum (Amax) and the mean (A50) pixel values measured on a 50% isocontour drawn on each sphere. Measurements were performed first with the phantom placed in the centre of the field of view and then in the off-centre position with the presence of three scattering sources to simulate the acquisition of four animals simultaneously. Quantification accuracy was finally validated using four 3D-printed phantoms mimicking rats with four subcutaneous tumours each. All experiments were performed for both 18F and 68Ga radionuclides.ResultsRadial spatial resolutions measured using the PSF reconstruction algorithm were 1.80 mm and 1.78 mm for centred and off-centred acquisitions, respectively. Spill-overs in air and water and uniformity computed with the NEMA phantom centred in the FOV were 0.05, 0.1 and 5.55% for 18F and 0.08, 0.12 and 2.81% for 68Ga, respectively. Recovery coefficients calculated with the 18F-filled micro-hollow sphere phantom for each sphere varied from 0.51 to 1.43 for Amax and from 0.40 to 1.01 for A50. These values decreased from 0.28 to 0.92 for Amax and from 0.22 to 0.66 for A50 for 68 Ga acquisition. The results were not significantly different when imaging phantoms in the off-centre position with 3 scattering sources. Measurements performed with the four 3D-printed phantoms showed a good correlation between theoretical and measured activity in simulated tumours, with r2 values of 0.99 and 0.97 obtained for 18F and 68Ga, respectively.ConclusionWe found that the clinical SiPM-based PET system was close to that obtained with a dedicated small-animal PET device. This study showed the ability of such a system to image four rats simultaneously and to perform quantification analysis for radionuclides commonly used in oncology.

3D FRONSAC with PSF reconstruction

PurposeThis study extends the Fast ROtary Nonlinear Spatial ACquisition (FRONSAC) method to include 3D acquisitions and reconstructions. It uses a transform domain reconstruction which is needed to make 3D reconstructions practical and provides new insights into how parallel imaging performance is enhanced by FRONSAC encoding. MethodsThis work developed the first examples of FRONSAC incorporated into a 3D acquisition. 3D FRONSAC was tested on human subjects with both simple gradient echo and MPRAGE Cartesian acquisitions. The quality of the 3D FRONSAC images was evaluated using structural similarity index measure (SSIM), and normalized root mean square error (NRMSE). ResultsFRONSAC encoding did not significantly modify the contrast obtained in either sequence, but it substantially improves the image quality of undersampled reconstruction. FRONSAC images have reduced undersampling ghosts and consistently improved SSIM and NRMSE. ConclusionsAcquisition and reconstruction of 3D FRONSAC images are feasible, and the additional FRONSAC encoding improves image quality in highly undersampled images.

Deep learning study on the mechanism of edge artifacts in point spread function reconstruction for numerical brain images.

Non-blinded image deblurring with deep learning was performed on blurred numerical brain images without point spread function (PSF) reconstruction to obtain edge artifacts (EA)-free images. This study uses numerical simulation to investigate the mechanism of EA in PSF reconstruction based on the spatial frequency characteristics of EA-free images. In 256 × 256 matrix brain images, the signal values of gray matter (GM), white matter, and cerebrospinal fluid were set to 1, 0.25, and 0.05, respectively. We assumed ideal projection data of a two-dimensional (2D) parallel beam with no degradation factors other than detector response blur to precisely grasp EA using the PSF reconstruction algorithm from blurred projection data. The detector response was assumed to be a shift-invariant and one-dimensional (1D) Gaussian function with 2-5mm full width at half maximum (FWHM). Images without PSF reconstruction (non-PSF), PSF reconstruction without regularization (PSF) and with regularization of relative difference function (PSF-RD) were generated by ordered subset expectation maximization (OSEM). For non-PSF, the image deblurring with a deep image prior (DIP) was applied using a 2D Gaussian function with 2-5mm FWHM. The 1D object-specific modulation transfer function (1D-OMTF), which is the ratio of 1D amplitude spectrum of the original and reconstructed images, was used as the index of spatial frequency characteristics. When the detector response was greater than 3mm FWHM, EA in PSF was observed in GM borders and narrow GM. No remarkable EA was observed in the DIP, and the FWHM estimated from the recovery coefficient for the deblurred image of non-PSF at 5mm FWHM was reduced to 3mm or less. PSF of 5mm FWHM showed higher spatial frequency characteristics than that of DIP up to around 2.2cycles/cm but was lower than the latter after 3cycles/cm. PSF-RD showed almost the same spatial frequency characteristics as that of DIP above 3 cycles/cm but was inferior below 3cycles/cm. PSF-RD has a lower spatial resolution than DIP. Unlike DIP, PSF lacks high-frequency components around the Nyquist frequency, generating EA. PSF-RD mitigates EA while simultaneously suppressing the signal, diminishing spatial resolution.

Off-axis point spread function reconstruction forsingle conjugate adaptive optics.

Modern giant segmented mirror telescopes (GSMTs) such as the Extremely Large Telescope, which is currently under construction, depend heavily on adaptive optics (AO) systems to correct for atmospheric distortions. However, a residual blur always remains in the astronomical images corrected by single conjugate AO (SCAO) systems due to fitting and bandwidth errors, which can mathematically be described by a convolution of the true image with a point spread function (PSF). Due to the nature of the turbulent atmosphere and its correction, the PSF is spatially varying, which is known as an anisoplanatic effect. The PSF serves, e.g.,as a quality measure for science images and therefore needs to be known as accurately as possible. In this paper, we present an algorithm for PSF reconstruction from pupil-plane data in directions apart from the guide star direction in an SCAO system. Our algorithm is adapted to the needs of GSMTs focused on estimating the contribution of the anisoplanatic and generalized fitting error to the PSF. Results obtained in an end-to-end simulation tool show a qualitatively good reconstruction of the PSF compared to the PSF calculated directly from the simulated incoming wavefront as well as stable performance with respect to imprecise knowledge of atmospheric parameters.

AIROPA IV: Validating point spread function reconstruction on various science cases

We present an analysis of six independent on-sky datasets taken with the Keck-II/NIRC2 instrument. Using the off-axis point spread function (PSF) reconstruction software AIROPA, we extract stellar astrometry, photometry, and other fitting metrics to characterize the performance of this package. We test the effectiveness of AIROPA to reconstruct the PSF across the field of view in varying atmospheric conditions, number and location of PSF reference stars, stellar crowding, and telescope position angle (PA). We compare the astrometric precision and fitting residuals between a static PSF model and a spatially varying PSF model that incorporates instrumental aberrations and atmospheric turbulence during exposures. Most of the fitting residuals we measure show little to no improvement in the variable-PSF mode over the single-PSF mode. For one of the data sets, we find photometric performance is significantly improved (by ∼10 × ) by measuring the trend seen in photometry as a function of off-axis location. For nearly all other metrics we find comparable astrometric and photometric precision across both PSF modes, with a ∼13 % smaller astrometric uncertainty in variable-PSF mode in the best case. We largely confirm that the spatially variable PSF does not significantly improve the astrometric and other PSF fitting residuals over the static PSF for on-sky observations. We attribute this to unaccounted instrumental aberrations that are not characterized through afternoon adaptive optics (AO) bench calibrations.

Bayesian penalized likelihood PET reconstruction impact on quantitative metrics in diffuse large B-cell lymphoma.

Evaluate the quantitative, subjective (Deauville score [DS]) and reader agreement differences between standard ordered subset expectation maximization (OSEM) and Bayesian penalized likelihood (BPL) positron emission tomography (PET) reconstruction methods. A retrospective review of 104 F-18 fluorodeoxyglucose PET/computed tomography (CT) exams among 52 patients with diffuse large B-cell lymphoma. An unblinded radiologist moderator reviewed both BPL and OSEM PET/CT exams. Four blinded radiologists then reviewed the annotated cases to provide a visual DS for each annotated lesion. Significant (P < .001) differences in BPL and OSEM PET methods were identified with greater standard uptake value (SUV) maximum and SUV mean for BPL. The DS was altered in 25% of cases when BPL and OSEM were reviewed by the same radiologist. Interobserver DS agreement was higher for OSEM (>1 cm lesion = 0.89 and ≤1 cm lesion = 0.84) compared to BPL (>1 cm lesion = 0.85 and ≤1 cm lesion = 0.81). Among the 4 readers, average intraobserver visual DS agreement between OSEM and BPL was 0.67 for lesions >1cm and 0.4 for lesions ≤1 cm. F-18 Fluorodeoxyglucose PET/CT of diffuse large B-cell lymphoma reconstructed with BPL has higher SUV values, altered DSs and reader agreement when compared to OSEM. This report finds volumetric PET measurements such as metabolic tumor volume to be similar between BPL and OSEM PET reconstructions. Efforts such as adoption of European Association Research Ltd accreditation should be made to harmonize PET data with an aim at balancing the need for harmonization and sensitivity for lesion detection.

Image-to-image translation for wavefront and point spread function estimation

We develop and evaluate a new approach to phase estimation for observational astronomy that can be used for accurate point spread function reconstruction. Phase estimation is required where a terrestrial observatory uses an adaptive optics (AO) system to assist astronomers in acquiring sharp, high-contrast images of faint and distant objects. Our approach is to train a conditional adversarial artificial neural network architecture to predict phase using the wavefront sensor data from a closed-loop AO system. We present a detailed simulation study under different turbulent conditions, using the retrieved residual phase to obtain the point spread function of the simulated instrument. Compared to the state-of-the-art model-based approach in astronomy, our approach is not explicitly limited by modeling assumptions, e.g., independence between terms, such as bandwidth and anisoplanatism—and is conceptually simple and flexible. We use the open-source COMPASS tool for end-to-end simulations. On key quality metrics, specifically the Strehl ratio and Halo distribution in our application domain, our approach achieves results better than the model-based baseline.

TDCOSMO

The importance of alternative methods for measuring the Hubble constant, such as time-delay cosmography, is highlighted by the recent Hubble tension. It is paramount to thoroughly investigate and rule out systematic biases in all measurement methods before we can accept new physics as the source of this tension. In this study, we perform a check for systematic biases in the lens modelling procedure of time-delay cosmography by comparing independent and blind time-delay predictions of the system WGD 2038−4008 from two teams using two different software programs: GLEE and LENSTRONOMY. The predicted time delays from the two teams incorporate the stellar kinematics of the deflector and the external convergence from line-of-sight structures. The un-blinded time-delay predictions from the two teams agree within 1.2σ, implying that once the time delay is measured the inferred Hubble constant will also be mutually consistent. However, there is a ∼4σ discrepancy between the power-law model slope and external shear, which is a significant discrepancy at the level of lens models before the stellar kinematics and the external convergence are incorporated. We identify the difference in the reconstructed point spread function (PSF) to be the source of this discrepancy. When the same reconstructed PSF was used by both teams, we achieved excellent agreement, within ∼0.6σ, indicating that potential systematics stemming from source reconstruction algorithms and investigator choices are well under control. We recommend that future studies supersample the PSF as needed and marginalize over multiple algorithms or realizations for the PSF reconstruction to mitigate the systematics associated with the PSF. A future study will measure the time delays of the system WGD 2038−4008 and infer the Hubble constant based on our mass models.

Point spread function reconstruction for SOUL + LUCI LBT data

Here, we present the status of an ongoing project aimed at developing a point spread function (PSF) reconstruction software for adaptive optics (AO) observations. In particular, we test for the first time the implementation of pyramid wave-front sensor data on our algorithms. As a first step in assessing its reliability, we applied the software to bright, on-axis, point-like sources using two independent sets of observations, acquired with the single-conjugated AO upgrade for the Large Binocular Telescope. Using only telemetry data, we reconstructed the PSF by carefully calibrating the instrument response. The accuracy of the results has been first evaluated using the classical metric: specifically, the reconstructed PSFs differ from the observed ones by <2 % in Strehl ratio and 4.5% in full-width at half maximum. Moreover, the recovered encircled energy associated with the PSF core is accurate at 4% level in the worst case. The accuracy of the reconstructed PSFs has then been evaluated by considering an idealized scientific test-case consisting in the measurements of the morphological parameters of a compact galaxy. In the future, our project will include the analysis of anisoplanatism, low signal-to-noise ratio regimes, and the application to multi-conjugated AO observations.

P15.11.A 18F-Fluciclovine PET/CT to distinguish radiation necrosis from tumour progression in brain metastases treated with stereotactic radiosurgery: results of a prospective pilot study

Abstract Background Amino acid PET radiopharmaceutical, 18F-fluciclovine, shows increased uptake in brain tumors relative to normal tissue and may be a useful tool for detecting recurrent brain metastases. Here, we report results from a prospective pilot study evaluating the use of 18F-fluciclovine PET/CT to distinguish radiation necrosis from tumour progression among patients with brain metastases treated with stereotactic radiosurgery (SRS). Material and Methods The primary objective was to estimate the accuracy of 18F-fluciclovine PET/CT in distinguishing radiation necrosis from tumour progression. The trial included adults with brain metastases who underwent SRS and presented with a follow up MRI brain (with DSC MR perfusion) which was equivocal for radiation necrosis versus tumour progression. Within 30 days of equivocal MRI brain, patients underwent an 18F-fluciclovine PET/CT (Siemens mCT) acquired 5-15 min post-injection with images generated by PSF reconstruction. Quantitative metrics for each lesion were documented and lesion to normal brain SUVmean ratios were calculated. The reference standard for diagnosis of radiation necrosis vs tumour progression was clinical follow up with MRI brain every 2-4 months until multidisciplinary consensus or tissue confirmation. Results Of 16 patients enrolled between 7/2019-11/2020, 1 patient died prior to diagnosis, allowing 15 evaluable subjects with 20 lesions. Primary histology was NSCLC in 9 (45%) lesions, breast in 7 (35%), melanoma in 3 (15%), and endometrial in 1 (5%). The final diagnosis was radiation necrosis in 16 (80%) lesions and tumour progression in 4 (20%). SUVmax was a statistically significant predictor of tumour progression (P = 0.011), with higher SUVmax values indicative of tumour progression. The area under the ROC curve was 0.833 (95% CI: 0.590, 1.0). A cutoff of 4.3 provided a sensitivity to identify tumour progression of 1.0 (4/4) and specificity to rule out tumour progression of 0.63 (10/16). SUVmean (P = 0.018), SUVpeak (P = 0.007), and SUVpeak/normal (P = 0.002) also reached statistical significance as predictors of tumour progression, with higher SUVmax values indicative of tumour progression. SUVmax/normal (P = 0.1) and SUVmean/normal (P = 0.5) were not statistically significant. The AUC for SUVmax was not significantly higher than the AUCs for the other quantitative variables (P-values &amp;gt; 0.2). Conclusion In this prospective pilot study, 18F Fluciclovine PET/CT demonstrated promising accuracy to distinguish radiation necrosis from tumour progression among patients with brain metastases previously treated with SRS. Using SUVmax, a cutpoint of 4.3 provided a sensitivity of 1.0 and specificity of 0.63. Confirmatory phase II and III studies are ongoing.

Positron range in combination with point-spread-function correction: an evaluation of different implementations for [124I]-PET imaging

AimTo evaluate the effect of combining positron range correction (PRC) with point-spread-function (PSF) correction and to compare different methods of implementation into iterative image reconstruction for 124I-PET imaging.Materials and methodsUniform PR blurring kernels of 124I were generated using the GATE (GEANT4) framework in various material environments (lung, water, and bone) and matched to a 3D matrix. The kernels size was set to 11 × 11 × 11 based on the maximum PR in water and the voxel size of the PET system. PET image reconstruction was performed using the standard OSEM algorithm, OSEM with PRC implemented before the forward projection (OSEM+PRC simplified) and OSEM with PRC implemented in both forward- and back-projection steps (full implementation) (OSEM+PRC). Reconstructions were repeated with resolution recovery, point-spread function (PSF) included. The methods and kernel variation were validated using different phantoms filled with 124I acquired on a Siemens mCT PET/CT system. The data was evaluated for contrast recovery and image noise.ResultsContrast recovery improved by 2–10% and 4–37% with OSEM+PRC simplified and OSEM+PRC, respectively, depending on the sphere size of the NEMA IQ phantom. Including PSF in the reconstructions further improved contrast by 4–19% and 3–16% with the PSF+PRC simplified and PSF+PRC, respectively. The benefit of PRC was more pronounced within low-density material. OSEM-PRC and OSEM-PSF as well as OSEM-PSF+PRC in its full- and simplified implementation showed comparable noise and convergence. OSEM-PRC simplified showed comparably faster convergence but at the cost of increased image noise.ConclusionsThe combination of the PSF and PRC leads to increased contrast recovery with reduced image noise compared to stand-alone PSF or PRC reconstruction. For OSEM-PRC reconstructions, a full implementation in the reconstruction is necessary to handle image noise. For the combination of PRC with PSF, a simplified PRC implementation can be used to reduce reconstruction times.

AIROPA III: testing simulated and on-sky data

Adaptive optics (AO) images from the W. M. Keck Observatory have delivered numerous influential scientific results, including detection of multi-system asteroids, the supermassive black hole at the center of the Milky Way, and directly imaged exoplanets. Specifically, the precise and accurate astrometry these images yield was used to measure the mass of the supermassive black hole using orbits of the surrounding star cluster. Despite these successes, one of the major obstacles to improved astrometric measurements is the spatial and temporal variability of the point-spread function delivered by the instruments. Anisoplanatic and Instrumental Reconstruction of Off-axis PSFs for AO (AIROPA) is a software package for the astrometric and photometric analysis of AO images using point-spread function fitting together with the technique of point-spread function reconstruction. In AO point-spread function reconstruction, the knowledge of the instrument performance and of the atmospheric turbulence is used to predict the long-exposure point-spread function of an observation. We present the results of our tests using AIROPA on both simulated and on-sky images of the Galactic Center. We find that our method is very reliable in accounting for the static aberrations internal to the instrument, but it does not improve significantly the accuracy on sky, possibly due to uncalibrated telescope aberrations.

Analyzing long-term performance of the Keck-II adaptive optics system

We present an analysis of the long-term performance of the W. M. Keck observatory laser guide star adaptive optics (LGS-AO) system and explore factors that influence the overall AO performance most strongly. Astronomical surveys can take years or decades to finish, so it is worthwhile to characterize the AO performance on such timescales in order to better understand future results. The Keck telescopes have two of the longest-running LGS-AO systems in use today, and as such they represent an excellent test-bed for processing large amounts of AO data. We use a Keck-II near infrared camera 2 (NIRC2) LGSAO surve of the Galactic Center (GC) from 2005 to 2019 for our analysis, combining image metrics with AO telemetry files, multiaperture scintillation sense/differential imaging motion monitor turbulence profiles, seeing information, weather data, and temperature readings in a compiled dataset to highlight areas of potential performance improvement. We find that image quality trends downward over time, despite multiple improvements made to Keck-II and its AO system, resulting in a 9 mas increase in the average full width at half maximum (FWHM) and a 3% decrease in the average Strehl ratio over the course of the survey. Image quality also trends upward with ambient temperature, possibly indicating the presence of uncorrected turbulence in the beam path. Using nine basic features from our dataset, we train a simple machine learning (ML) algorithm to predict the delivered image quality of NIRC2 given current atmospheric conditions, which could eventually be used for real-time observation planning and exposure time adjustments. A random forest algorithm trained on this data can predict the Strehl ratio of an image to within 18% and the FWHM to within 7%, which is a solid baseline for future applications involving more advanced ML techniques. The assembled dataset and coding tools are released to the public as a resource for testing new predictive control and point spread function-reconstruction algorithms.

Brain partial volume correction with point spreading function reconstruction in high-resolution digital PET: comparison with an MR-based method in FDG imaging

ObjectiveIn quantitative positron emission tomography (PET) of the brain, partial volume effect due mainly to the finite spatial resolution of the PET scanner (> 3 mm full width at half maximum [FWHM]) is a primary source of error in the measurement of tracer uptake, especially in small structures such as the cerebral cortex (typically < 3 mm thickness). The aim of this study was to evaluate the partial volume correction (PVC) performance of point spread function-incorporated reconstruction (PSF reconstruction) in combination with the latest digital PET scanner. This evaluation was performed through direct comparisons with magnetic resonance imaging (MR)-based PVC (used as a reference method) in a human brain study.MethodsTen healthy subjects underwent brain 18F-FDG PET (30-min acquisition) on a digital PET/CT system (Siemens Biograph Vision, 3.5-mm FWHM scanner resolution at the center of the field of view) and anatomical T1-weighted MR imaging for MR-based PVC. PSF reconstruction was applied with a wide range of iterations (4 to 256; 5 subsets). FDG uptake in the cerebral cortex was evaluated using the standardized uptake value ratio (SUVR) and compared between PSF reconstruction and MR-based PVC.ResultsCortical structures were visualized by PSF reconstruction with several tens of iterations and were anatomically well matched with the MR-derived cortical segments. Higher numbers of iterations resulted in higher cortical SUVRs, which approached those of MR-based PVC (1.76), although even with the maximum number of iterations they were still smaller by 16% (1.47), corresponding to approximately 1.5-mm FWHM of the effective spatial resolution.ConclusionWith the latest digital PET scanner, PSF reconstruction can be used as a PVC technique in brain PET, albeit with suboptimal resolution recovery. A relative advantage of PSF reconstruction is that it can be applied not only to cerebral cortical regions, but also to various small structures such as small brain nuclei that are hardly visualized on anatomical T1-weighted imaging, and thus hardly recovered by MR-based PVC.

A Cube-Based Dual-GPU List-Mode Reconstruction Algorithm for PET Imaging

Graphics processing units (GPUs) have been a popular choice for the acceleration of iterative tomographic reconstruction algorithms. In this work, two dual-GPU approaches to iterative list-mode reconstruction of PET data are studied. The developed techniques split the image into equal-size cubes and distribute them between the GPUs. Benefiting from caching a plane of the image into the fast shared memory, line-driven forward and back projection kernels are implemented incorporating a shift-invariant point spread function (PSF). In the first approach, lines were individually convolved with the PSF kernel whereas the second approach convolves the reconstructed image with the PSF before the forward projection and after the back projection. The algorithms were implemented using two CUDA-compatible devices with no CPU/GPU data transfer during the reconstruction process. List-mode data obtained from Monte Carlo simulations were used to evaluate the methods. One complete iteration of 107 events on an image of 220 × 220 × 110 voxels is done in less than 510 ms for a non-PSF reconstruction and in 520 to 5220 ms for a PSF reconstruction depending on the PSF and the reconstruction parameters. Dual-GPU memory layout, implementation details, sample images, and performance metrics of the developed algorithms are discussed.

Advanced quantitative evaluation of PET systems using the ACR phantom and NiftyPET software.

PurposeA novel phantom‐imaging platform, a set of software tools, for automated and high‐precision imaging of the American College of Radiology (ACR) positron emission tomography (PET) phantom for PET/magnetic resonance (PET/MR) and PET/computed tomography (PET/CT) systems is proposed.MethodsThe key feature of this platform is the vector graphics design that facilitates the automated measurement of the knife‐edge response function and hence image resolution, using composite volume of interest templates in a 0.5 mm resolution grid applied to all inserts of the phantom. Furthermore, the proposed platform enables the generation of an accurate μ‐map for PET/MR systems with a robust alignment based on two‐stage image registration using specifically designed PET templates. The proposed platform is based on the open‐source NiftyPET software package used to generate multiple list‐mode data bootstrap realizations and image reconstructions to determine the precision of the two‐stage registration and any image‐derived statistics. For all the analyses, iterative image reconstruction was employed with and without modeled shift‐invariant point spread function and with varying iterations of the ordered subsets expectation maximization (OSEM) algorithm. The impact of the activity outside the field of view (FOV) was assessed using two acquisitions of 30 min each, with and without the activity outside the FOV.ResultsThe utility of the platform has been demonstrated by providing a standard and an advanced phantom analysis including the estimation of spatial resolution using all cylindrical inserts. In the imaging planes close to the edge of the axial FOV, we observed deterioration in the quantitative accuracy, reduced resolution (FWHM increased by 1–2 mm), reduced contrast, and background uniformity due to the activity outside the FOV. Although it slows convergence, the PSF reconstruction had a positive impact on resolution and contrast recovery, but the degree of improvement depended on the regions. The uncertainty analysis based on bootstrap resampling of raw PET data indicated high precision of the two‐stage registration.ConclusionsWe demonstrated that phantom imaging using the proposed methodology with the metric of spatial resolution and multiple bootstrap realizations may be helpful in more accurate evaluation of PET systems as well as in facilitating fine tuning for optimal imaging parameters in PET/MR and PET/CT clinical research studies.

Impact of point-spread function reconstruction on dynamic and static 18F-DOPA PET/CT quantitative parameters in glioma.

Quantification of dynamic and static parameters extracted from 3,4-dihydroxy-6-[18F]-fluoro-L-phenylalanine (18F-DOPA, FDOPA) positron emission tomography (PET)/computed tomography (CT) plays a critical role for glioma assessment. The objective of the present study was to investigate the impact of point-spread function (PSF) reconstruction on these quantitative parameters. Fourteen patients with untreated gliomas and investigated with FDOPA PET/CT were analyzed. The distribution of the 14 cases was as follows: 6 astrocytomas-isocitrate dehydrogenase-mutant; 2 oligodendrogliomas/1p19q-codeleted-isocitrate dehydrogenase-mutant; and 6 isocitrate dehydrogenase-wild-type glioblastomas. A 0-20-min dynamic images (8×15, 2×30, 2×60, and 3×300 s post-injection) and a 0-20-min static image were reconstructed with and without PSF. Tumoral volumes-of-interest were generated on all of the PET series and the background volumes-of-interest were generated on the 0-20-min static image with and without PSF. Static parameters (SUVmax and SUVmean) of the tumoral and the background volumes-of-interest and kinetic parameters (K1 and k2) of the tumoral volumes-of-interest extracted from using full kinetic analysis were provided. PSF and non-PSF quantitative parameters values were compared. Thirty-three tumor volumes-of-interest and 14 background volumes-of-interest were analyzed. PSF images provided higher tumor SUVmax than non-PSF images for 23/33 VOIs [median SUVmax =3.0 (range, 1.4-10.2) with PSF vs. 2.7 (range, 1.4-9.1) without PSF; P<0.001] and higher tumor SUVmean for 13/33 volumes-of-interest [median SUVmean =2.0 (range, 0.8-7.6) with PSF vs. 2.0 (range, 0.8-7.4) without PSF; P=0.002]. K1 and k2 were significantly lower with PSF than without PSF [respectively median K1 =0.077 mL/ccm/min (range, 0.043-0.445 mL/ccm/min) with PSF vs. 0.101 mL/ccm/min (range, 0.055-0.578 mL/ccm/min) without PSF; P<0.001 and median k2 =0.070 min-1 (range, 0.025-0.146 min-1) with PSF vs. 0.081 min-1 (range, 0.027-0.180 min-1) without PSF; P<0.001]. Background SUVmax and SUVmean were statistically unaffected [respectively median SUVmax =1.7 (range, 1.3-2.0) with PSF vs. 1.7 (range, 1.3-1.9) without PSF; P=0.346 and median SUVmean =1.5 (range, 1.0-1.8) with PSF vs. 1.5 (range, 1.0-1.7) without PSF; P=0.371]. The present study confirms that PSF significantly increases tumor activity concentrations measured on PET images. PSF algorithms for quantitative PET/CT analysis should be used with caution, especially for quantification of kinetic parameters.