Ordered Subsets Expectation Maximization Algorithm Research Articles

Accuracy of Rhenium-188 SPECT/CT activity quantification for applications in radionuclide therapy using clinical reconstruction methods

The main applications of 188Re in radionuclide therapies include trans-arterial liver radioembolization and palliation of painful bone-metastases. In order to optimize 188Re therapies, the accurate determination of radiation dose delivered to tumors and organs at risk is required. Single photon emission computed tomography (SPECT) can be used to perform such dosimetry calculations. However, the accuracy of dosimetry estimates strongly depends on the accuracy of activity quantification in 188Re images.In this study, we performed a series of phantom experiments aiming to investigate the accuracy of activity quantification for 188Re SPECT using high-energy and medium-energy collimators. Objects of different shapes and sizes were scanned in Air, non-radioactive water (Cold-water) and water with activity (Hot-water). The ordered subset expectation maximization algorithm with clinically available corrections (CT-based attenuation, triple-energy window (TEW) scatter and resolution recovery was used). For high activities, the dead-time corrections were applied. The accuracy of activity quantification was evaluated using the ratio of the reconstructed activity in each object to this object’s true activity. Each object’s activity was determined with three segmentation methods: a 1% fixed threshold (for cold background), a 40% fixed threshold and a CT-based segmentation. Additionally, the activity recovered in the entire phantom, as well as the average activity concentration of the phantom background were compared to their true values. Finally, Monte-Carlo simulations of a commercial -camera were performed to investigate the accuracy of the TEW method.Good quantification accuracy (errors <10%) was achieved for the entire phantom, the hot-background activity concentration and for objects in cold background segmented with a 1% threshold. However, the accuracy of activity quantification for objects segmented with 40% threshold or CT-based methods decreased (errors >15%), mostly due to partial-volume effects. The Monte-Carlo simulations confirmed that TEW-scatter correction applied to 188Re, although practical, yields only approximate estimates of the true scatter.

Continuous Analog of Accelerated OS-EM Algorithm for Computed Tomography

The maximum-likelihood expectation-maximization (ML-EM) algorithm is used for an iterative image reconstruction (IIR) method and performs well with respect to the inverse problem as cross-entropy minimization in computed tomography. For accelerating the convergence rate of the ML-EM, the ordered-subsets expectation-maximization (OS-EM) with a power factor is effective. In this paper, we propose a continuous analog to the power-based accelerated OS-EM algorithm. The continuous-time image reconstruction (CIR) system is described by nonlinear differential equations with piecewise smooth vector fields by a cyclic switching process. A numerical discretization of the differential equation by using the geometric multiplicative first-order expansion of the nonlinear vector field leads to an exact equivalent iterative formula of the power-based OS-EM. The convergence of nonnegatively constrained solutions to a globally stable equilibrium is guaranteed by the Lyapunov theorem for consistent inverse problems. We illustrate through numerical experiments that the convergence characteristics of the continuous system have the highest quality compared with that of discretization methods. We clarify how important the discretization method approximates the solution of the CIR to design a better IIR method.

Respiratory motion correction in 4D-PET by simultaneous motion estimation and image reconstruction (SMEIR)

In conventional 4D positron emission tomography (4D-PET), images from different frames are reconstructed individually and aligned by registration methods. Two issues that arise with this approach are as follows: (1) the reconstruction algorithms do not make full use of projection statistics; and (2) the registration between noisy images can result in poor alignment. In this study, we investigated the use of simultaneous motion estimation and image reconstruction (SMEIR) methods for motion estimation/correction in 4D-PET. A modified ordered-subset expectation maximization algorithm coupled with total variation minimization (OSEM-TV) was used to obtain a primary motion-compensated PET (pmc-PET) from all projection data, using Demons derived deformation vector fields (DVFs) as initial motion vectors. A motion model update was performed to obtain an optimal set of DVFs in the pmc-PET and other phases, by matching the forward projection of the deformed pmc-PET with measured projections from other phases. The OSEM-TV image reconstruction was repeated using updated DVFs, and new DVFs were estimated based on updated images. A 4D-XCAT phantom with typical FDG biodistribution was generated to evaluate the performance of the SMEIR algorithm in lung and liver tumors with different contrasts and different diameters (10–40 mm). The image quality of the 4D-PET was greatly improved by the SMEIR algorithm. When all projections were used to reconstruct 3D-PET without motion compensation, motion blurring artifacts were present, leading up to 150% tumor size overestimation and significant quantitative errors, including 50% underestimation of tumor contrast and 59% underestimation of tumor uptake. Errors were reduced to less than 10% in most images by using the SMEIR algorithm, showing its potential in motion estimation/correction in 4D-PET.

Similarity-regulation of OS-EM for accelerated SPECT reconstruction

Ordered subsets expectation maximization (OS-EM) is widely used to accelerate image reconstruction in single photon emission computed tomography (SPECT). Speedup of OS-EM over maximum likelihood expectation maximization (ML-EM) is close to the number of subsets used. Although a high number of subsets can shorten reconstruction times significantly, it can also cause severe image artifacts such as improper erasure of reconstructed activity if projections contain few counts. We recently showed that such artifacts can be prevented by using a count-regulated OS-EM (CR-OS-EM) algorithm which automatically adapts the number of subsets for each voxel based on the estimated number of counts that the voxel contributed to the projections. While CR-OS-EM reached high speed-up over ML-EM in high-activity regions of images, speed in low-activity regions could still be very slow. In this work we propose similarity-regulated OS-EM (SR-OS-EM) as a much faster alternative to CR-OS-EM. SR-OS-EM also automatically and locally adapts the number of subsets, but it uses a different criterion for subset regulation: the number of subsets that is used for updating an individual voxel depends on how similar the reconstruction algorithm would update the estimated activity in that voxel with different subsets. Reconstructions of an image quality phantom and in vivo scans show that SR-OS-EM retains all of the favorable properties of CR-OS-EM, while reconstruction speed can be up to an order of magnitude higher in low-activity regions. Moreover our results suggest that SR-OS-EM can be operated with identical reconstruction parameters (including the number of iterations) for a wide range of count levels, which can be an additional advantage from a user perspective since users would only have to post-filter an image to present it at an appropriate noise level.

Quantitative accuracy of (177)Lu SPECT reconstruction using different compensation methods: phantom and patient studies.

BackgroundIn targeted radionuclide therapy (TRT), accurate quantification using SPECT/CT images is important for optimizing radiation dose delivered to both the tumour and healthy tissue. Quantitative SPECT images are regularly reconstructed using the ordered subset expectation maximization (OSEM) algorithm with various compensation methods such as attenuation (A), scatter (S) and detector and collimator response (R). In this study, different combinations of the compensation methods are applied during OSEM reconstruction and the effect on the 177Lu quantification accuracy is studied in an anthropomorphic torso phantom. In addition, the phantom results are reflected to (177)Lu-DOTA-Tyr3-octreotate (177Lu-DOTATATE)-treated patient data and kidney absorbed dose estimates.MethodsThe torso phantom was imaged with nine various sized (0.4–104.4 cm3) spherical inserts, filled with known 177Lu activity ranging from 0.5 to 105.5 MBq. Images were reconstructed using OSEM algorithm using A, AR and ARS compensation method combinations. The compensation method combinations were compared by calculating the concentration recovery coefficient (cRC) for each insert. In addition, ten 177Lu-DOTATATE-treated patient’s post-therapy dosimetry acquisitions were reconstructed, and the absorbed dose to kidneys was estimated.ResultscRC values depend on the insert size for all compensation methods. AR and ARS produced significantly higher cRC values than attenuation correction alone. There were no cRC value differences between the methods for the smallest 1-cm-diameter insert, cRC being 0.18. However, the collimator and detector response compensation method (R) made the 1.3-cm-diameter insert clearly visible and improved cRC estimate from 0.19 to 0.43. ARS produced slightly higher cRC values for small- and medium-sized inserts than AR. On the patient data, a similar trend could be seen. AR and ARS produced higher kidney activities than using attenuation correction alone; the total absorbed doses to the right and left kidneys were on average 15 and 20 % higher for AR and 19 and 25 % higher for ARS, respectively. The effective half-life decay estimated from time-activity curves however showed no notable difference between the compensation methods.ConclusionsThe highest cRC values were achieved by applying ARS compensation during reconstruction. The results were notably higher than those using attenuation correction alone. Similarly, higher activity estimates and thus higher absorbed dose estimates were found in patient data when all compensation methods were applied. ARS improved cRC especially in small-sized sources, and it thus might aid tumour dosimetry for 177Lu PRRT treatments.

Applications of the Medipix3-CT in combination with iterative reconstruction techniques

The pixelated semiconductor detectors of the Medipix family with their photon-counting abilities offer the possibility of high quality X-ray radiography as well as computed tomography. The generated signal from each photon is amplified and shaped before it is compared to an energy threshold. For a photon with an energy above the threshold the counter is incremented by one count. Photons below the operator-defined threshold do not increment the counter and therefore do not participate in the image formation. Furthermore, compared to other detectors like scintillators, an additional conversion step is dispensed due to the direct converting nature of photon-counting detectors, leading to a higher signal-to-noise-ratio. Additionally, the photon processing capabilities of photon-counting detectors allow photons to be weighted equally and not proportional to their energy as it is the case for charge integrating devices, where high energy photons are weighted stronger than low energy photons. Compared to integrating devices, this leads to an increase in contrast for images of both high and low contrast objects, hence improve object information. The use of photon-counting detectors in combination with iterative reconstruction techniques based on OSEM (ordered subset expectation maximization) algorithms is the basis of our computed tomography scans for material analysis. Due to its ability to operate with highly undersampled data sets, iterative reconstruction offers the possibility to decrease dose in CT scans. In order to identify the limits of the data set reduction, a first series of scans was performed to test, under real conditions, the CT-image quality when a strongly reduced amount of projections is used for reconstruction. In addition, the effect of a total variation minimization tool on these undersampled data sets was evaluated. Furthermore, this paper includes a number of recent CT-results with scans performed at two different setups within our facility.

Accelerated time-of-flight (TOF) PET image reconstruction using TOF bin subsetization and TOF weighting matrix pre-computation

Time-of-flight (TOF) positron emission tomography (PET) technology has recently regained popularity in clinical PET studies for improving image quality and lesion detectability. Using TOF information, the spatial location of annihilation events is confined to a number of image voxels along each line of response, thereby the cross-dependencies of image voxels are reduced, which in turns results in improved signal-to-noise ratio and convergence rate. In this work, we propose a novel approach to further improve the convergence of the expectation maximization (EM)-based TOF PET image reconstruction algorithm through subsetization of emission data over TOF bins as well as azimuthal bins. Given the prevalence of TOF PET, we elaborated the practical and efficient implementation of TOF PET image reconstruction through the pre-computation of TOF weighting coefficients while exploiting the same in-plane and axial symmetries used in pre-computation of geometric system matrix. In the proposed subsetization approach, TOF PET data were partitioned into a number of interleaved TOF subsets, with the aim of reducing the spatial coupling of TOF bins and therefore to improve the convergence of the standard maximum likelihood expectation maximization (MLEM) and ordered subsets EM (OSEM) algorithms. The comparison of on-the-fly and pre-computed TOF projections showed that the pre-computation of the TOF weighting coefficients can considerably reduce the computation time of TOF PET image reconstruction. The convergence rate and bias-variance performance of the proposed TOF subsetization scheme were evaluated using simulated, experimental phantom and clinical studies. Simulations demonstrated that as the number of TOF subsets is increased, the convergence rate of MLEM and OSEM algorithms is improved. It was also found that for the same computation time, the proposed subsetization gives rise to further convergence. The bias-variance analysis of the experimental NEMA phantom and a clinical FDG-PET study also revealed that for the same noise level, a higher contrast recovery can be obtained by increasing the number of TOF subsets. It can be concluded that the proposed TOF weighting matrix pre-computation and subsetization approaches enable to further accelerate and improve the convergence properties of OSEM and MLEM algorithms, thus opening new avenues for accelerated TOF PET image reconstruction.

Validation of semi-quantitative methods for DAT SPECT: influence of anatomical variability and partial volume effect

The aim of this work was to evaluate the influence of anatomical variability between subjects and of the partial volume effect (PVE) on the standardized Specific Uptake Ratio (SUR) in [123I]FP-bib SPECT studies. To this end, magnetic resonance (MR) images of 23 subjects with differences in the striatal volume of up to 44% were segmented and used to generate a database of 138 Monte Carlo simulated SPECT studies. Data included normal uptakes and pathological cases. Studies were reconstructed by filtered back projection (FBP) and the ordered-subset expectation-maximization algorithm. Quantification was carried out by applying a reference method based on regions of interest (ROIs) derived from the MR images and ROIs derived from the Automated Anatomical Labelling map. Our results showed that, regardless of anatomical variability, the relationship between calculated and true SUR values for caudate and putamen could be described by a multiple linear model which took into account the spill-over phenomenon caused by PVE ( for caudate and ≥0.980 for putamen) and also by a simple linear model (R2 ≥ 0.952 for caudate and ≥0.973 for putamen). Calculated values were standardized by inverting both linear systems. Differences between standardized and true values showed that, although the multiple linear model was the best approach in terms of variability ( ≥ 11.79 for caudate and ≤7.36 for putamen), standardization based on a simple linear model was also suitable ( ≥ 12.44 for caudate and ≤12.57 for putamen).

Performance evaluation of a 90°-rotating dual-head small animal PET system

Dual-head PET system suffers from poor resolution in the dimension perpendicular to the detector heads and vastly increased computations required by the accurate estimation of the system response matrix (SRM). The problem of poor x-axis resolution is caused by missing data owing to the intrinsic geometry flaws of the dual-head PET, which leads to high SRM similarity between the adjacent voxels along the x-axis. To compensate for this problem, our dual-head PET system prototype incorporates a -rotating mechanism. In addition, the cubic field of view (FOV) and cubic voxels are adopted to achieve a better geometrical symmetry, which avoids the re-estimation of SRM after rotation. The spatial resolution, sensitivity, noise equivalent count rate, scatter fraction and pileup fraction under different head separation distances were evaluated by a series of simulations, and the results were compared with those of the non-rotation system. The SRM was estimated using the Monte Carlo method and the three-dimensional images were reconstructed using the ordered subsets expectation maximization algorithm. Experimental results reveal a significantly improved x-axis resolution via the proposed rotation mechanism. Our system can achieve x-axis spatial resolution of 0.79 mm, 0.74 mm and 0.72 mm at the center of FOV under the head separation distance of 5 cm, 10 cm and 15 cm, respectively.

High-resolution brain SPECT imaging by combination of parallel and tilted detector heads.

To improve the spatial resolution of brain single-photon emission computed tomography (SPECT), we propose a new brain SPECT system in which the detector heads are tilted towards the rotation axis so that they are closer to the brain. In addition, parallel detector heads are used to obtain the complete projection data set. We evaluated this parallel and tilted detector head system (PT-SPECT) in simulations. In the simulation study, the tilt angle of the detector heads relative to the axis was 45°. The distance from the collimator surface of the parallel detector heads to the axis was 130 mm. The distance from the collimator surface of the tilted detector heads to the origin on the axis was 110 mm. A CdTe semiconductor panel with a 1.4 mm detector pitch and a parallel-hole collimator were employed in both types of detector head. A line source phantom, cold-rod brain-shaped phantom, and cerebral blood flow phantom were evaluated. The projection data were generated by forward-projection of the phantom images using physics models, and Poisson noise at clinical levels was applied to the projection data. The ordered-subsets expectation maximization algorithm with physics models was used. We also evaluated conventional SPECT using four parallel detector heads for the sake of comparison. The evaluation of the line source phantom showed that the transaxial FWHM in the central slice for conventional SPECT ranged from 6.1 to 8.5 mm, while that for PT-SPECT ranged from 5.3 to 6.9 mm. The cold-rod brain-shaped phantom image showed that conventional SPECT could visualize up to 8-mm-diameter rods. By contrast, PT-SPECT could visualize up to 6-mm-diameter rods in upper slices of a cerebrum. The cerebral blood flow phantom image showed that the PT-SPECT system provided higher resolution at the thalamus and caudate nucleus as well as at the longitudinal fissure of the cerebrum compared with conventional SPECT. PT-SPECT provides improved image resolution at not only upper but also at central slices of the cerebrum.

WE-AB-204-09: Respiratory Motion Correction in 4D-PET by Simultaneous Motion Estimation and Image Reconstruction (SMEIR)

Purpose: In conventional 4D-PET, images from different frames are reconstructed individually and aligned by registration methods. Two issues with these approaches are: 1) Reconstruction algorithms do not make full use of all projections statistics; and 2) Image registration between noisy images can Result in poor alignment. In this study we investigated the use of simultaneous motion estimation and image reconstruction (SMEIR) method for cone beam CT for motion estimation/correction in 4D-PET. Methods: Modified ordered-subset expectation maximization algorithm coupled with total variation minimization (OSEM- TV) is used to obtain a primary motion-compensated PET (pmc-PET) from all projection data using Demons derived deformation vector fields (DVFs) as initial. Motion model update is done to obtain an optimal set of DVFs between the pmc-PET and other phases by matching the forward projection of the deformed pmc-PET and measured projections of other phases. Using updated DVFs, OSEM- TV image reconstruction is repeated and new DVFs are estimated based on updated images. 4D XCAT phantom with typical FDG biodistribution and a 10mm diameter tumor was used to evaluate the performance of the SMEIR algorithm. Results: Image quality of 4D-PET is greatly improved by the SMEIR algorithm. When all projections are used to reconstruct a 3D-PET, motion blurring artifacts are present, leading to a more than 5 times overestimation of the tumor size and 54% tumor to lung contrast ratio underestimation. This error reduced to 37% and 20% for post reconstruction registration methods and SMEIR respectively. Conclusion: SMEIR method can be used for motion estimation/correction in 4D-PET. The statistics is greatly improved since all projection data are combined together to update the image. The performance of the SMEIR algorithm for 4D-PET is sensitive to smoothness control parameters in the DVF estimation step.

Quantitative SPECT/CT Imaging of (177)Lu with In Vivo Validation in Patients Undergoing Peptide Receptor Radionuclide Therapy.

The purpose of this study is to extend an established SPECT/CT quantitation protocol to (177)Lu and validate it in vivo using urine samples, thus providing a basis for 3D dosimetry of (177)Lu radiotherapy and improvement over current planar methods which improperly account for anatomical variations, attenuation, and overlapping organs. In our quantitation protocol, counts in images reconstructed using an ordered subset-expectation maximization algorithm are converted to kilobecquerels per milliliter using a calibration factor derived from a phantom experiment. While varying reconstruction parameters, we tracked the ratio of image to true activity concentration (recovery coefficient, RC) in hot spheres and a noise measure in a homogeneous region. The optimal parameter set was selected as the point where recovery in the largest three spheres (16, 8, and 4 ml) stagnated, while the noise continued to increase. Urine samples were collected following 12 SPECT/CT acquisitions of patients undergoing [(177)Lu]DOTATATE therapy, and activity concentrations were measured in a well counter. Data was reconstructed using parameters chosen in the phantom experiment, and estimated activity concentration from the images was compared to the urine values to derive RCs. In phantom data, our chosen parameter set yielded RCs in 16, 8, and 4 ml spheres of 80.0, 74.1, and 64.5 %, respectively. For patients, the mean bladder RC was 96.1 ± 13.2% (range, 80.6-122.4 %), with a 95 % confidence interval between 88.6 and 103.6 %. The mean error of SPECT/CT concentrations was 10.1 ± 8.3% (range, -19.4-22.4 %). Our results show that quantitative (177)Lu SPECT/CT in vivo is feasible but could benefit from improved reconstruction methods. Quantifying bladder activity is analogous to determining the amount of activity in the kidneys, an important task in dosimetry, and our results provide a useful benchmark for future efforts.

4D numerical observer for lesion detection in respiratory-gated PET.

Respiratory-gated positron emission tomography (PET)/computed tomography protocols reduce lesion smearing and improve lesion detection through a synchronized acquisition of emission data. However, an objective assessment of image quality of the improvement gained from respiratory-gated PET is mainly limited to a three-dimensional (3D) approach. This work proposes a 4D numerical observer that incorporates both spatial and temporal informations for detection tasks in pulmonary oncology. The authors propose a 4D numerical observer constructed with a 3D channelized Hotelling observer for the spatial domain followed by a Hotelling observer for the temporal domain. Realistic (18)F-fluorodeoxyglucose activity distributions were simulated using a 4D extended cardiac torso anthropomorphic phantom including 12 spherical lesions at different anatomical locations (lower, upper, anterior, and posterior) within the lungs. Simulated data based on Monte Carlo simulation were obtained using geant4 application for tomographic emission (GATE). Fifty noise realizations of six respiratory-gated PET frames were simulated by GATE using a model of the Siemens Biograph mMR scanner geometry. PET sinograms of the thorax background and pulmonary lesions that were simulated separately were merged to generate different conditions of the lesions to the background (e.g., lesion contrast and motion). A conventional ordered subset expectation maximization (OSEM) reconstruction (5 iterations and 6 subsets) was used to obtain: (1) gated, (2) nongated, and (3) motion-corrected image volumes (a total of 3200 subimage volumes: 2400 gated, 400 nongated, and 400 motion-corrected). Lesion-detection signal-to-noise ratios (SNRs) were measured in different lesion-to-background contrast levels (3.5, 8.0, 9.0, and 20.0), lesion diameters (10.0, 13.0, and 16.0 mm), and respiratory motion displacements (17.6-31.3 mm). The proposed 4D numerical observer applied on multiple-gated images was compared to the conventional 3D approach applied on the nongated and motion-corrected images. On average, the proposed 4D numerical observer improved the detection SNR by 48.6% (p < 0.005), whereas the 3D methods on motion-corrected images improved by 31.0% (p < 0.005) as compared to the nongated method. For all different conditions of the lesions, the relative SNR measurement (Gain = SNRObserved/SNRNongated) of the 4D method was significantly higher than one from the motion-corrected 3D method by 13.8% (p < 0.02), where Gain4D was 1.49 ± 0.21 and Gain3D was 1.31 ± 0.15. For the lesion with the highest amplitude of motion, the 4D numerical observer yielded the highest observer-performance improvement (176%). For the lesion undergoing the smallest motion amplitude, the 4D method provided superior lesion detectability compared with the 3D method, which provided a detection SNR close to the nongated method. The investigation on a structure of the 4D numerical observer showed that a Laguerre-Gaussian channel matrix with a volumetric 3D function yielded higher lesion-detection performance than one with a 2D-stack-channelized function, whereas a different kind of channels that have the ability to mimic the human visual system, i.e., difference-of-Gaussian, showed similar performance in detecting uniform and spherical lesions. The investigation of the detection performance when increasing noise levels yielded decreasing detection SNR by 27.6% and 41.5% for the nongated and gated methods, respectively. The investigation of lesion contrast and diameter showed that the proposed 4D observer preserved the linearity property of an optimal-linear observer while the motion was present. Furthermore, the investigation of the iteration and subset numbers of the OSEM algorithm demonstrated that these parameters had impact on the lesion detectability and the selection of the optimal parameters could provide the maximum lesion-detection performance. The proposed 4D numerical observer outperformed the other observers for the lesion-detection task in various lesion conditions and motions. The 4D numerical observer shows substantial improvement in lesion detectability over the 3D observer method. The proposed 4D approach could potentially provide a more reliable objective assessment of the impact of respiratory-gated PET improvement for lesion-detection tasks. On the other hand, the 4D approach may be used as an upper bound to investigate the performance of the motion correction method. In future work, the authors will validate the proposed 4D approach on clinical data for detection tasks in pulmonary oncology.

SU‐E‐J‐100: Reconstruction of Prompt Gamma Ray Three Dimensional SPECT Image From Boron Neutron Capture Therapy(BNCT)

Purpose:Purpose of paper is to confirm the feasibility of acquisition of three dimensional single photon emission computed tomography (SPECT) image from boron neutron capture therapy (BNCT) using Monte Carlo simulation.Methods:In case of simulation, the pixelated SPECT detector, collimator and phantom were simulated using Monte Carlo n particle extended (MCNPX) simulation tool. A thermal neutron source (&lt;1 eV) was used to react with the boron uptake region (BUR) in the phantom. Each geometry had a spherical pattern, and three different BURs (A, B and C region, density: 2.08 g/cm3) were located in the middle of the brain phantom. The data from 128 projections for each sorting process were used to achieve image reconstruction. The ordered subset expectation maximization (OSEM) reconstruction algorithm was used to obtain a tomographic image with eight subsets and five iterations. The receiver operating characteristic (ROC) curve analysis was used to evaluate the geometric accuracy of reconstructed image.Results:The OSEM image was compared with the original phantom pattern image. The area under the curve (AUC) was calculated as the gross area under each ROC curve. The three calculated AUC values were 0.738 (A region), 0.623 (B region), and 0.817 (C region). The differences between length of centers of two boron regions and distance of maximum count points were 0.3 cm, 1.6 cm and 1.4 cm.Conclusion:The possibility of extracting a 3D BNCT SPECT image was confirmed using the Monte Carlo simulation and OSEM algorithm. The prospects for obtaining an actual BNCT SPECT image were estimated from the quality of the simulated image and the simulation conditions. When multiple tumor region should be treated using the BNCT, a reasonable model to determine how many useful images can be obtained from the SPECT could be provided to the BNCT facilities.This research was supported by the Leading Foreign Research Institute Recruitment Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, Information and Communication Technologies (ICT) and Future Planning (MSIP)(Grant No.200900420) and the Radiation Technology Research and Development program (Grant No.2013043498), Republic of Korea.

Statistical iterative reconstruction for streak artefact reduction when using multidetector CT to image the dento-alveolar structures.

When metallic prosthetic appliances and dental fillings exist in the oral cavity, the appearance of metal-induced streak artefacts is not avoidable in CT images. The aim of this study was to develop a method for artefact reduction using the statistical reconstruction on multidetector row CT images. Adjacent CT images often depict similar anatomical structures. Therefore, reconstructed images with weak artefacts were attempted using projection data of an artefact-free image in a neighbouring thin slice. Images with moderate and strong artefacts were continuously processed in sequence by successive iterative restoration where the projection data was generated from the adjacent reconstructed slice. First, the basic maximum likelihood-expectation maximization algorithm was applied. Next, the ordered subset-expectation maximization algorithm was examined. Alternatively, a small region of interest setting was designated. Finally, the general purpose graphic processing unit machine was applied in both situations. The algorithms reduced the metal-induced streak artefacts on multidetector row CT images when the sequential processing method was applied. The ordered subset-expectation maximization and small region of interest reduced the processing duration without apparent detriments. A general-purpose graphic processing unit realized the high performance. A statistical reconstruction method was applied for the streak artefact reduction. The alternative algorithms applied were effective. Both software and hardware tools, such as ordered subset-expectation maximization, small region of interest and general-purpose graphic processing unit achieved fast artefact correction.

An Initial Study for PET Imaging Simulation

Simulation play an important role in the research of PET/CT imaging technology. In this paper, GATE (Geant4 Application Tomography Emission) software packages and NCAT(dynamic NURBS-based cardiac-torso phantom) were used to simulate GE ST PET/CT imaging. GATE provides abundant of functions to simulate both PET/CT imaging procedure and geometric phantom generation. NCAT can generate voxlised body torso phantom. Three different kinds of digital phantoms were designed and generated for PET/CT imaging simulation. The simulation output of GATE was converted to the data format that the STIR (Software for Tomographic Image Reconstruction) requires to reconstruct image by the OSEM (ordered subsets expectation-maximization) algorithm. The experiment results validate that GATE and NCAT are able to simulate PET/CT imaging.

Improved visualization of perfusion defects by respiratory-gated SPECT

Single-photon emission computed tomography ventilation/perfusion (SPECT V/Q) imaging is recommended both by the Society of Nuclear Medicine and by the European Association of Nuclear Medicine for the diagnosis of pulmonary embolism. However, respiratory motion produces image blurring and degradation of detail in the lungs. We have investigated respiratory gating of SPECT images, correcting for motion to reduce blur and improve image definition. Wedge-shaped defects of different sizes ranging from 15 to 4 mm were fixed in the lung cavities of an anthropomorphic lung phantom to simulate perfusion defects. Gated and nongated SPECT images were obtained using a double-headed SPECT system. Three-dimensional movement was introduced using a purpose-built moving platform with two motion frequencies of 10 and 20 cycles/min. Motion was tracked with a respiratory-gating system. Gated SPECT data were acquired in 16 discrete data bins in synchronization with the breathing cycle. The images were reconstructed using ordered-subset expectation maximization algorithms and corrected for rigid motion. Contrast and contrast-to-noise ratios (CNRs) were measured to quantify any improvement in the gated motion-corrected images. Visualization of defects in the reconstructed images was performed by seven observers and analyzed using alternative free-response receiver operating characteristic analysis. Assessment of gated and nongated SPECT phantom images demonstrated that motion adversely affected the detectability of defects. Quantification of data demonstrated that, in the controlled simulation, image quality, defect definition, observer confidence, contrast, and CNR were increased after applying motion correction. Improvement in CNRs was found to be significant using alternative free-response receiver operating characteristic analysis (P=0.0002). Respiratory-gated motion-corrected SPECT images enhanced the visualization of defects compared with matched moving/nongated images in a realistic moving phantom. This approach may be particularly valuable for SPECT V/Q imaging and may improve the diagnosis of pulmonary embolism.

Advantages of Improved Time Resolution for TOF PET at Very Low Statistics

In this paper, Monte Carlo generated data are used to simulate a time-of-flight (TOF) positron emission tomography (PET) scanner with improving time resolution, down to 200-ps time resolution. TOF and non-TOF ordered subsets expectation maximization (OSEM) reconstructions were compared in terms of contrast recovery and signal-to-noise ratio, with the purpose of characterizing the differences of TOF and non-TOF at very low counts. This study shows that there is a lower limit to the number of counts in a PET scan, in order to obtain a quantitative PET image. Below this limit, the reliability of the measurement is very poor. The use of TOF information has the beneficial effect of lowering this statistical limit, allowing for shorter reliable PET scans. Another interesting finding derives from the comparison with TOF filtered back projection (FBP) images: if time resolution can be improved, TOF FBP images become competitive with TOF MLEM and OSEM algorithms, not only in terms of contrast recovery but also in terms of reduced noise level.

Comparison of a Noise-Weighted Filtered Backprojection Algorithm with the Standard MLEM Algorithm for Poisson Noise

Iterative maximum-likelihood expectation maximization and ordered-subset expectation maximization algorithms are excellent for image reconstruction and usually provide better images than filtered backprojection (FBP). Recently, an FBP algorithm able to incorporate noise weighting during reconstruction was developed. This paper compares the performance of the noise-weighted FBP algorithm and the iterative maximum-likelihood expectation maximization algorithm with Poisson noise-corrupted emission data generated by computer simulations and a SPECT experimental study. The results show comparable performance for these 2 algorithms.

High quality ordered subset expectation maximization reconstruction algorithm based on multi-resolution for PET images

In Positron Emission Tomography(PET) imaging,Maximum Likelihood Expectation Maximization(MLEM) algorithm cannot be directly applied to clinical diagnosis due to suppressing noise ineffectively and converging slowly.Although Ordered Subset Expectation Maximization(OSEM) algorithm converges fast,it will lead to a significant decline in the quality of the reconstructed image.To address this problem,multi-resolution technology was introduced into the subset of the OSEM reconstruction algorithm to suppress noise and stabilize solving process.The experimental results indicate that the new algorithm overcomes the drawback of the traditional algorithm on degrading the reconstructed image and has the advantage of fast convergence.The proposed reconstruction algorithm can obtain a higher Signal-to-Noise Ratio(SNR) and a superior visual effect.