A Monte Carlo investigation of the spatial resolution performance of a small-animal PET scanner designed for mouse brain imaging studies

Our laboratory has developed PET detectors with depth-encoding accuracy of ∼2 mm based on finely pixelated crystals with a tapered geometry, readout at both ends with position-sensitive avalanche photodiodes (PSAPDs). These detectors are currently being used in our laboratory to build a one-ring high resolution PET scanner for mouse brain imaging studies. Due to the inactive areas around the PSAPDs, large gaps exist between the detector modules which can degrade the image spatial resolution obtained using analytical reconstruction with filtered backprojection (FBP). In this work, the Geant4-based GATE Monte Carlo package was used to assist in determining whether gantry rotation was necessary and to assess the expected spatial resolution of the system. The following factors were investigated: rotating vs. static gantry modes with and without compensation of missing data using the discrete cosine transform (DCT) method, two levels of depth-encoding, and positron annihilation effects for ¹⁸F. Our results indicate that while the static scanner produces poor quality FBP images with streak and ring artifacts, the image quality was greatly improved after compensation of missing data. The simulation indicates that the expected FWHM system spatial resolution is 0.70 ± 0.05 mm, which approaches the predicted limit of 0.5 mm FWHM due to positron range, photon non-colinearity and physical detector element size effects. We conclude that excellent reconstructed resolution without gantry rotation is possible even using FBP if the gaps are appropriately handled and that this design can approach the resolution limits set by positron annihilation physics.     

Monte Carlo-based assessment of the trade-off between spatial resolution,field-of-view and scattered radiation in the variable resolution X-ray CT scanner

ObjectiveThe purpose of this work is to evaluate the impact of optimization of magnification on performance parameters of the variable resolution X-ray (VRX) CT scanner.MethodsA realistic model based on an actual VRX CT scanner was implemented in the GATE Monte Carlo simulation platform. To evaluate the influence of system magnification, spatial resolution, field-of-view (FOV) and scatter-to-primary ratio of the scanner were estimated for both fixed and optimum object magnification at each detector rotation angle. Comparison and inference between these performance parameters were performed angle by angle to determine appropriate object position at each opening half angle.ResultsOptimization of magnification resulted in a trade-off between spatial resolution and FOV of the scanner at opening half angles of 90°–12°, where the spatial resolution increased up to 50% and the scatter-to-primary ratio decreased from 4.8% to 3.8% at a detector angle of about 90° for the same FOV and X-ray energy spectrum. The disadvantage of magnification optimization at these angles is the significant reduction of the FOV (up to 50%). Moreover, magnification optimization was definitely beneficial for opening half angles below 12° improving the spatial resolution from 7.5 cy/mm to 20 cy/mm. Meanwhile, the FOV increased by more than 50% at these angles.ConclusionIt can be concluded that optimization of magnification is essential for opening half angles below 12°. For opening half angles between 90° and 12°, the VRX CT scanner magnification should be set according to the desired spatial resolution and FOV.     

Assessment of the variation in CT scanner performance (image quality and Hounsfield units) with scan parameters,for image optimisation in radiotherapy treatment planning

PurposeTo define a method and investigate how the adjustment of scan parameters affected the image quality and Hounsfield units (HUs) on a CT scanner used for radiotherapy treatment planning. A lack of similar investigations in the literature may be a contributing factor in the apparent reluctance to optimise radiotherapy CT protocols.MethodA Catphan phantom was used to assess how image quality on a Toshiba Aquilion LB scanner changed with scan parameters. Acquisition and reconstruction field-of-view (FOV), collimation, image slice thickness, effective mAs per rotation and reconstruction algorithm were varied. Changes were assessed for HUs of different materials, high contrast spatial resolution (HCSR), contrast-noise ratio (CNR), HU uniformity, scan direction low contrast and CT dose-index.ResultsCNR and HCSR varied most with reconstruction algorithm, reconstruction FOV and effective mAs. Collimation, but not image slice width, had a significant effect on CT dose-index with narrower collimation giving higher doses. Dose increased with effective mAs. Highest HU differences were seen when changing reconstruction algorithm: 56 HU for densities close to water and 117 HU for bone-like materials. Acquisition FOV affected the HUs but reconstruction FOV and effective mAs did not.ConclusionsAll the scan parameters investigated affected the image quality metrics. Reconstruction algorithm, reconstruction FOV, collimation and effective mAs were most important. Reconstruction algorithm and acquisition FOV had significant effect on HU. The methodology is applicable to radiotherapy CT scanners when investigating image quality optimisation, prior to assessing the impact of scan protocol changes on clinical CT images and treatment plans.     

Full field spatially-variant image-based resolution modelling reconstruction for the HRRT

Accurate characterisation of the scanner's point spread function across the entire field of view (FOV) is crucial in order to account for spatially dependent factors that degrade the resolution of the reconstructed images. The HRRT users' community resolution modelling reconstruction software includes a shift-invariant resolution kernel, which leads to transaxially non-uniform resolution in the reconstructed images. Unlike previous work to date in this field, this work is the first to model the spatially variant resolution across the entire FOV of the HRRT, which is the highest resolution human brain PET scanner in the world. In this paper we developed a spatially variant image-based resolution modelling reconstruction dedicated to the HRRT, using an experimentally measured shift-variant resolution kernel. Previously, the system response was measured and characterised in detail across the entire FOV of the HRRT, using a printed point source array. The newly developed resolution modelling reconstruction was applied on measured phantom, as well as clinical data and was compared against the HRRT users' community resolution modelling reconstruction, which is currently in use. Results demonstrated improvements both in contrast and resolution recovery, particularly for regions close to the edges of the FOV, with almost uniform resolution recovery across the entire transverse FOV. In addition, because the newly measured resolution kernel is slightly broader with wider tails, compared to the deliberately conservative kernel employed in the HRRT users' community software, the reconstructed images appear to have not only improved contrast recovery (up to 20% for small regions), but also better noise characteristics.     

Assessment of temporal resolution of multi-detector row computed tomography in helical acquisition mode using the impulse method

The purpose of this study was to propose a method for assessing the temporal resolution (TR) of multi-detector row computed tomography (CT) (MDCT) in the helical acquisition mode using temporal impulse signals generated by a metal ball passing through the acquisition plane. An 11-mm diameter metal ball was shot along the central axis at approximately 5 m/s during a helical acquisition, and the temporal sensitivity profile (TSP) was measured from the streak image intensities in the reconstructed helical CT images. To assess the validity, we compared the measured and theoretical TSPs for the 4-channel modes of two MDCT systems. A 64-channel MDCT system was used to compare TSPs and image quality of a motion phantom for the pitch factors P of 0.6, 0.8, 1.0 and 1.2 with a rotation time R of 0.5 s, and for two R/P combinations of 0.5/1.2 and 0.33/0.8. Moreover, the temporal transfer functions (TFs) were calculated from the obtained TSPs. The measured and theoretical TSPs showed perfect agreement. The TSP narrowed with an increase in the pitch factor. The image sharpness of the 0.33/0.8 combination was inferior to that of the 0.5/1.2 combination, despite their almost identical full width at tenth maximum values. The temporal TFs quantitatively confirmed these differences. The TSP results demonstrated that the TR in the helical acquisition mode significantly depended on the pitch factor as well as the rotation time, and the pitch factor and reconstruction algorithm affected the TSP shape.     

In phantom evaluation of targeting accuracy in MRI-based brain radiosurgery

PurposeTo determine the targeting accuracy of brain radiosurgery when planning procedures employing different MRI and MRI + CT combinations are adopted.Materials and methodA new phantom, the BrainTool, has been designed and realized to test image co-registration and targeting accuracy in a realistic anatomical situation. The phantom was created with a 3D printer and materials that mimic realistic brain MRI and CT contrast using a model extracted from a synthetic MRI study of a human brain. Eight markers distributed within the BrainTool provide for assessment of the accuracy of image registrations while two cavities that host an ionization chamber are used to perform targeting accuracy measurements with an iterative cross-scan method. Two procedures employing 1.5 T MRI-only or a combination of MRI (taken with 1.5 T or 3 T scanners) and CT to carry out Gamma Knife treatments were investigated. As distortions can impact targeting accuracy, MR images were preliminary evaluated to assess image deformation extent using GammaTool phantom.ResultsMR images taken with both scanners showed average and maximum distortion of 0.3 mm and 1 mm respectively. The marker distances in co-registered images resulted below 0.5 mm for both MRI scans. The targeting mismatches obtained were 0.8 mm, 1.0 mm and 1.2 mm for MRI-only and MRI + CT (1,5T and 3 T), respectively.ConclusionsProcedures using a combination of MR and CT images provide targeting accuracies comparable to those of MRI-only procedures. The BrainTool proved to be a suitable tool for carrying out co-registration and targeting accuracy of Gamma Knife brain radiosurgery treatments.     

Dynamic in vivo 3D atlantoaxial spine kinematics during upright rotation

Diagnosing dysfunctional atlantoaxial motion is challenging given limitations of current diagnostic imaging techniques. Three-dimensional imaging during upright functional motion may be useful in identifying dynamic instability not apparent on static imaging. Abnormal atlantoaxial motion has been linked to numerous pathologies including whiplash, cervicogenic headaches, C2 fractures, and rheumatoid arthritis. However, normal C1/C2 rotational kinematics under dynamic physiologic loading have not been previously reported owing to imaging difficulties. The objective of this study was to determine dynamic three-dimensional in vivo C1/C2 kinematics during upright axial rotation. Twenty young healthy adults performed full head rotation while seated within a biplane X-ray system while radiographs were collected at 30 images per second. Six degree-of-freedom kinematics were determined for C1 and C2 via a validated volumetric model-based tracking process. The maximum global head rotation (to one side) was 73.6 ± 8.3°, whereas maximum C1 rotation relative to C2 was 36.8 ± 6.7°. The relationship between C1/C2 rotation and head rotation was linear through midrange motion (±20° head rotation from neutral) in a nearly 1:1 ratio. Coupled rotation between C1 and C2 included 4.5 ± 3.1° of flexion and 6.4 ± 8.2° of extension, and 9.8 ± 3.8° of contralateral bending. Translational motion of C1 relative to C2 was 7.8 ± 1.5 mm ipsilaterally, 2.2 ± 1.2 mm inferiorly, and 3.3 ± 1.0 mm posteriorly. We believe this is the first study describing 3D dynamic atlantoaxial kinematics under true physiologic conditions in healthy subjects. C1/C2 rotation accounts for approximately half of total head axial rotation. Additionally, C1 undergoes coupled flexion/extension and contralateral bending, in addition to inferior, lateral and posterior translation.     

Magnetic particle imaging for artifact-free imaging of intracranial flow diverter stents: A phantom study

PurposeMagnetic Particle Imaging (MPI) is a new, background- and radiation-free tomographic imaging method that enables near real-time imaging of superparamagnetic iron-oxide nanoparticles (SPIONs) with high temporal and spatial resolution. This phantom study aims to investigate the potential of MPI for visualization of the stent lumen in intracranial flow diverters (FD).MethodsNitinol FD of different dimensions (outer diameter: 3.5 mm, 4.0 mm, 5.5 mm; total length: 22–40 mm) were scanned in vascular phantoms in a custom-built MPI scanner (in-plane resolution: ~ 2 mm, field of view: 65 mm length, 29 mm diameter). Phantoms were filled with diluted (1:50) SPION tracer agent Ferucarbotran (10 µmol (Fe)/ml; NaCL). Each phantom was measured in 32 different projections (overall acquisition time per image: 3200 ms, 5 averages). After image reconstruction from raw data, two radiologists assessed image quality using a 5-point Likert scale. The signal intensity profile was measured using a semi-automatic evaluation tool.ResultsMPI visualized the lumen of all FD without relevant differences between the stented vessel phantom and the reference phantom. At 3.5 mm image quality was slightly inferior to the larger diameters. The FD themselves neither generated an MPI signal nor did they lead to relevant imaging artifacts. Ratings of both radiologists showed no significant difference, interrater reliability was good (ICC 0.84). A quantitative evaluation of the signal intensity profile did not reveal any significant differences (p > 0.05) either.ConclusionMPI visualizes the lumen of nitinol FD stents in vessel phantoms without relevant stent-induced artifacts.     

Analysis of image sharpness reproducibility on a novel engineered micro-CT scanner with variable geometry and embedded recalibration software

This study investigates the reproducibility of the reconstructed image sharpness, after modifications of the geometry setup, for a variable magnification micro-CT (μCT) scanner. All the measurements were performed on a novel engineered μCT scanner for in vivo imaging of small animals (Xalt), which has been recently built at the Institute of Clinical Physiology of the National Research Council (IFC-CNR, Pisa, Italy), in partnership with the University of Pisa. The Xalt scanner is equipped with an integrated software for on-line geometric recalibration, which will be used throughout the experiments. In order to evaluate the losses of image quality due to modifications of the geometry setup, we have made 22 consecutive acquisitions by changing alternatively the system geometry between two different setups (Large FoV – LF, and High Resolution – HR). For each acquisition, the tomographic images have been reconstructed before and after the on-line geometric recalibration. For each reconstruction, the image sharpness was evaluated using two different figures of merit: (i) the percentage contrast on a small bar pattern of fixed frequency (f = 5.5 lp/mm for the LF setup and f = 10 lp/mm for the HR setup) and (ii) the image entropy. We have found that, due to the small-scale mechanical uncertainty (in the order of the voxel size), a recalibration is necessary for each geometric setup after repositioning of the system’s components; the resolution losses due to the lack of recalibration are worse for the HR setup (voxel size = 18.4 μm). The integrated on-line recalibration algorithm of the Xalt scanner allowed to perform the recalibration quickly, by restoring the spatial resolution of the system to the reference resolution obtained after the initial (off-line) calibration.     

Fully Automated Whole-Head Segmentation with Improved Smoothness and Continuity,with Theory Reviewed

Individualized current-flow models are needed for precise targeting of brain structures using transcranial electrical or magnetic stimulation (TES/TMS). The same is true for current-source reconstruction in electroencephalography and magnetoencephalography (EEG/MEG). The first step in generating such models is to obtain an accurate segmentation of individual head anatomy, including not only brain but also cerebrospinal fluid (CSF), skull and soft tissues, with a field of view (FOV) that covers the whole head. Currently available automated segmentation tools only provide results for brain tissues, have a limited FOV, and do not guarantee continuity and smoothness of tissues, which is crucially important for accurate current-flow estimates. Here we present a tool that addresses these needs. It is based on a rigorous Bayesian inference framework that combines image intensity model, anatomical prior (atlas) and morphological constraints using Markov random fields (MRF). The method is evaluated on 20 simulated and 8 real head volumes acquired with magnetic resonance imaging (MRI) at 1 mm3 resolution. We find improved surface smoothness and continuity as compared to the segmentation algorithms currently implemented in Statistical Parametric Mapping (SPM). With this tool, accurate and morphologically correct modeling of the whole-head anatomy for individual subjects may now be feasible on a routine basis. Code and data are fully integrated into SPM software tool and are made publicly available. In addition, a review on the MRI segmentation using atlas and the MRF over the last 20 years is also provided, with the general mathematical framework clearly derived.     

Faisabilité et intérêt d’un nouvel algorithme de reconstruction des images TEP-18FDG (AW OSEM DR) pour la délimitation des volumes cibles en radiothérapie. À propos d’un cas de néoplasie ORL

ObjectiveWe compared two reconstruction methods for 18fluorodeoxyglucose (¹⁸F-FDG) positron emission tomography (PET) images with “attenuation weighted ordered subset expectation maximization” using either the manufacturer-provided (AW-OSEM) or a “Detector response” (AW-OSEM DR) tomographic operator. We looked at the feasibility of using the latter reconstruction for radiotherapy target volumes definition in cancers of the superior aero-digestive tract (VADS). In this preliminary study, we first assessed the spatial resolution of images obtained with AW-OSEM and AW-OSEM DR on a Biograph™ 6, and secondly target volumes of radiotherapy “Gross Tumor Volume” (GTV), “Clinical Target Volume” (CTV) and “Planning Target Volume” (PTV) obtained with each of these reconstruction methods.Material and methodsThe spatial resolution was measured on a test object containing 4 radioactive point sources. Furthermore, radiotherapy target volumes have been defined with the software Eclipse™ on injected scanner (CT IV) and PET/CT (PET AW-OSEM and PET AW-OSEM DR) images.ResultsSpatial resolution was improved with AW-OSEM DR algorithm reconstruction compared to images obtained with AW-OSEM reconstruction (from 7.5 mm down to 5.4 mm for the highest reduction). GTV from AW-OSEM DR reconstruction with 42 and 50% of the “Standard uptake value maximum” (SUVmax) semi-automatic threshold (1.2 and 0.7 cm³ respectively) were lower than those obtained with AW-OSEM (3.6 and 2.2 cm³ respectively). They were also lower than GTV defined with CT IV (5.5 cm³). It was the same for CTV and PTV.ConclusionThis study showed that AW-OSEM DR reconstruction method allows less impaired spatial resolution than AW-OSEM. In the case of radiotherapy target volumes delineation, AW-OSEM DR may decrease the GTV, CTV and PTV and therefore the risk of side effects associated with organs at risk.     

Conceptual Design and Simulation Study of an ROI-Focused Panel-PET Scanner

Positron emission tomography (PET) is an important imaging modality for clincial use. Conventionally, the PET scanner is generally built to provide a roomy enough transverse field-of-view (FOV) for imaging most adults’ torsos. However, in many cases, the region-of-interest (ROI) for imaging is usually a small area inside the human body. Therefore, to fulfill a PET system which provides an FOV comparable in size to the target ROI seems appealing and more cost effective. Meanwhile, such a PET system has the potential for portable or bedside application with the reduced system size. In this work, we have investigated the feasibility of using dual-headed panel-detectors to build an ROI-focused PET scanner. A novel windowed list-mode ordered subset expectation maximization method was developed to perform the ROI image reconstruction. With this method, the ROI of the object can be reconstructed from the coincidences whose position determined by time-of-flight (TOF) measurements was inside the ROI. Monte Carlo simulation demonstrates the feasibility of detecting lesions not less than 1 cm in diameter, with a 300 ps full width at half maximum timing resolution. As a critical system performance, the impact of TOF information on image quality has been studied and the required TOF capability was assessed. With enhanced timing resolution, the distortions and artifacts were reduced effectively. The further improved TOF capability also shows a noticeable improvement of detection performance for low uptake lesions, as well as the recovery speed of lesion contrast, which is of practical significance in the lesion detection task.     

Digital breast tomosynthesis: Dose and image quality assessment

The aim of this work was to evaluate how different acquisition geometries and reconstruction parameters affect the performance of four digital breast tomosynthesis (DBT) systems (Senographe Essential – GE, Mammomat Inspiration – Siemens, Selenia Dimensions – Hologic and Amulet Innovality – Fujifilm) on the basis of a physical characterization.Average Glandular Dose (AGD) and image quality parameters such as in-plane/in-depth resolution, signal difference to noise ratio (SDNR) and artefact spread function (ASF) were examined.Measured AGD values resulted below EUREF limits for 2D imaging. A large variability was recorded among the investigated systems: the mean dose ratio DBT/2D ranged between 1.1 and 1.9.In-plane resolution was in the range: 2.2 mm⁻¹–3.8 mm⁻¹ in chest wall-nipple direction. A worse resolution was found for all devices in tube travel direction.In-depth resolution improved with increasing scan angle but was also affected by the choice of reconstruction and post-processing algorithms. The highest z-resolution was provided by Siemens (50°, FWHM = 2.3 mm) followed by GE (25°, FWHM = 2.8 mm), while the Fujifilm HR showed the lowest one, despite its wide scan angle (40°, FWHM = 4.1 mm).The ASF was dependent on scan angle: smaller range systems showed wider ASF curves; however a clear relationship was not found between scan angle and ASF, due to the different post processing and reconstruction algorithms.SDNR analysis, performed on Fujifilm system, demonstrated that pixel binning improves detectability for a fixed dose/projection.In conclusion, we provide a performance comparison among four DBT systems under a clinical acquisition mode.     

Measuring humeral head translation using fluoroscopy: A validation study

Numerous techniques have been employed to monitor humeral head translation due to its involvement with several shoulder pathologies. However, most of the techniques were not validated. The objective of this study is to compare the accuracy of manual digitization and contour registration in measuring superior translation of the humeral head. Eight pairs of cadaver scapulae and humerii bones were harvested for this study. Each scapula and humerus was secured in a customized jig that allowed for control of humeral head translations and a vise that permitted rotations of the scapula about three axes. Fluoroscopy was used to take images of the shoulder bones. Scapular orientation was manipulated in different positions while the humerus was at 90° of humeral elevation in the scapular plane. Humeral head translation was measured using the two methods and was compared to the known translation. Additionally, accuracy of the contour registration method to measure 2-D scapular rotations was assessed. The range for the root mean square (RMS) error for manual digitization method was 0.27 mm - 0.43 mm and the contour registration method had a RMS error ranging from 0.18 mm - 0.40 mm. In addition, the RMS error for the scapular angle rotation using the contour registration method was 2.4°. Both methods showed acceptable errors. However, on average, the contour registration method showed lesser measurement error compared to the manual digitization method. In addition, the contour registration method was able to show good accuracy in measuring rotation that is useful in 2-D image analysis.     

Glenohumeral joint kinematics measured by intracortical pins,reflective markers,and computed tomography: A novel technique to assess acromiohumeral distance

Combination of biplane fluoroscopy and CT-scan provides accurate 3D measurement of the acromiohumeral distance (AHD) during dynamic tasks. However, participants performed only two and six trials in previous experiments to respect the recommended radiation exposure per year. Our objective was to propose a technique to assess the AHD in 3D during dynamic tasks without this limitation. The AHD was computed from glenohumeral kinematics obtained using markers fitted to pins drilled into the scapula and the humerus combined with 3D bone geometry obtained using CT-scan. Four participants performed range-of-motion, daily-living, and sports activities. Sixty-six out of 158 trials performed by each participant were analyzed. Two participants were not considered due to experimental issues. AHD decreased with arm elevation. Overall, the smallest AHD occurred in abduction (1.1 mm (P1) and 1.2 mm (P2)). The smallest AHD were 2.4 mm (P1) and 3.1 mm (P2) during ADL. It was 2.8 mm (P1) and 1.1 mm (P2) during sports activities. The humeral head greater and lesser tuberosities came the nearest to the acromion. The proposed technique increases the number of trials acquired during one experiment compared to previous. The identification of movements maximizing AHD is possible, which may provide benefits for shoulder rehabilitation.     

Posterior kV-CBCT scanning of the head and neck region minimizes doses to critical organs with sustained image quality

We evaluated the absorbed dose to critical organs, as well as the image quality, at different partial angles in kV-CBCT (Cone Beam Computed Tomography) scanning of the head and neck region. CBCT images of phantom from a 200° rotation were performed by using three different scanning paths, anterior, posterior, and right lateral with Catphan504 and RANDO phantoms. Critical organ dose was measured using TLD 100H in the RANDO phantom. The image quality of those phantoms was evaluated, using HU uniformity, HU linearity, contrast-to-noise ratio, low contrast visibility and spatial resolution with the Catphan504 dataset; and 5-point grading scales for the RANDO phantom dataset by five radiation oncologists. The image qualities from Catphan504 and RANDO phantom of every scanning path were comparable, with no statistically significant difference (p ≥ 0.05). However, there was a significant difference in the critical organ dose in all paths (p < 0.05), depending on the critical organ location and the scanning direction. Scanning directions show no effects on the image quality. Differences in absorbed dose to critical organs should were evaluated. The posterior scanning path for the CBCT was deemed preferable due because of considerably lower doses to several critical organs of the head and neck region.     

An inhomogeneous most likely path formalism for proton computed tomography

PurposeMultiple Coulomb scattering (MCS) poses a challenge in proton CT (pCT) image reconstruction. The assumption of straight paths is replaced with Bayesian models of the most likely path (MLP). Current MLP-based pCT reconstruction approaches assume a water scattering environment. We propose an MLP formalism based on accurate determination of scattering moments in inhomogeneous media.MethodsScattering power relative to water (RScP) was calculated for a range of human tissues and investigated against relative stopping power (RStP). Monte Carlo simulation was used to compare the new inhomogeneous MLP formalism to the water approach in a slab geometry and a human head phantom. An MLP-Spline-Hybrid method was investigated for improved computational efficiency.ResultsA piecewise-linear correlation between RStP and RScP was shown, which may assist in iterative pCT reconstruction. The inhomogeneous formalism predicted Monte Carlo proton paths through a water cube with thick bone inserts to within 1.0 mm for beams ranging from 210 to 230 MeV incident energy. Improvement in accuracy over the conventional MLP ranged from 5% for a 230 MeV beam to 17% for 210 MeV. There was no noticeable gain in accuracy when predicting 200 MeV proton paths through a clinically relevant human head phantom. The MLP-Spline-Hybrid method reduced computation time by half while suffering negligible loss of accuracy.ConclusionsWe have presented an MLP formalism that accounts for material composition. In most clinical cases a water scattering environment can be assumed, however in certain cases of significant heterogeneity the proposed algorithm may improve proton path estimation.     

Practical considerations for using K3 cameras in CDS mode for high-resolution and high-throughput single particle cryo-EM

Detector technology plays a pivotal role in high-resolution and high-throughput cryo-EM structure determination. Compared with the first-generation, single-electron counting direct detection camera (Gatan K2), the latest K3 camera is faster, larger, and now offers a correlated-double sampling mode (CDS). Importantly this results in a higher DQE and improved throughput compared to its predecessor. In this study, we focused on optimizing camera data collection parameters for daily use within a cryo-EM facility and explored the balance between throughput and resolution. In total, eight data sets of murine heavy-chain apoferritin were collected at different dose rates and magnifications, using 9-hole image shift data collection strategies. The performance of the camera was characterized by the quality of the resultant 3D reconstructions. Our results demonstrated that the Gatan K3 operating in CDS mode outperformed standard (nonCDS) mode in terms of reconstruction resolution in all tested conditions with 8 electrons per pixel per second being the optimal dose rate. At low magnification (64kx) we were able to achieve reconstruction resolutions of 149% of the physical Nyquist limit (1.8 Å with a 1.346 Å physical pixel size). Low magnification allows more particles to be collected per image, aiding analysis of heterogeneous samples requiring large data sets. At moderate magnification (105kx, 0.834 Å physical pixel size) we achieved a resolution of 1.65 Å within 8-h of data collection, a condition optimal for achieving high-resolution on well behaved samples. Our results also show that for an optimal sample like apoferritin, one can achieve better than 2.5 Å resolution with 5 min of data collection. Together, our studies validate the most efficient ways of imaging protein complexes using the K3 direct detector and will greatly benefit the cryo-EM community.     

A convolution neural network for higher resolution dose prediction in prostate volumetric modulated arc therapy

PurposeThis study aims to investigate the feasibility of using convolutional neural networks to predict an accurate and high resolution dose distribution from an approximated and low resolution input dose.MethodsSixty-six patients were treated for prostate cancer with VMAT. We created the treatment plans using the Acuros XB algorithm with 2 mm grid size, followed by the dose calculated using the anisotropic analytical algorithm with 5 mm grid with the same plan parameters. U-net model was used to predict 2 mm grid dose from 5 mm grid dose. We investigated the two models differing for the training data used as input, one used just the low resolution dose (D model) and the other combined the low resolution dose with CT data (DC model). Dice similarity coefficient (DSC) was calculated to ascertain how well the shape of the dose-volume is matched. We conducted gamma analysis for the following: DVH from the two models and the reference DVH for all prostate structures.ResultsThe DSC values in the DC model were significantly higher than those in the D model (p < 0.01). For the CTV, PTV, and bladder, the gamma passing rates in the DC model were significantly higher than those in the D model (p < 0.002–0.02). The mean doses in the CTV and PTV for the DC model were significantly better matched to those in the reference dose (p < 0.0001).ConclusionsThe proposed U-net model with dose and CT image used as input predicted more accurate dose.     

Technical Note: Quality assessment of virtual monochromatic spectral images on a dual energy CT scanner

PurposeTo assess the quality of images obtained on a dual energy computed tomography (CT) scanner.MethodsImage quality was assessed on a 64 detector-row fast kVp-switching dual energy CT scanner (Revolution GSI, GE Medical Systems). The Catphan phantom and a low contrast resolution phantom were employed. Acquisitions were performed at eight different radiation dose levels that ranged from 9 mGy to 32 mGy. Virtual monochromatic spectral images (VMI) were reconstructed in the 40–140 keV range using all available kernels and iterative reconstruction (IR) at four different blending levels. Modulation Transfer Function (MTF) curves, image noise, image contrast, noise power spectrum and contrast to noise ratio were assessed.ResultsIn-plane spatial resolution at the 10% of the MTF curve was 0.60 mm⁻¹. In-plane spatial resolution was not modified with VMI energy and IR blending level. Image noise was reduced from 16.6 at 9 mGy to 6.7 at 32 mGy, while peak frequency remained within 0.14 ± 0.01 mm⁻¹. Image noise was reduced from 14.3 at IR 10% to 11.5 at IR 50% at a constant peak frequency. The lowest image noise and maximum peak frequency were recorded at 70 keV.ConclusionsOur results have shown how objective image quality is varied when different levels of radiation dose and different settings in IR are applied. These results provide CT operators an in depth understanding of the imaging performance characteristics in dual energy CT.