In the development of positron emission tomography (PET) detectors, understanding and optimizing scintillator light collection is critical for achieving high performance, particularly when the design incorporates depth-of-interaction (DOI) encoding or time-of-flight information. Monte-Carlo simulations play an important role in guiding research in detector designs and popular software such as GATE now include models of light transport in scintillators. Although current simulation toolkits are able to provide accurate models of perfectly polished surfaces, they do not successfully predict light output for other surface finishes, for example those often used in DOI-encoding detectors. The lack of accuracy of those models mainly originates from a simplified description of rough surfaces as an ensemble of micro-facets determined by the distribution of their normal, typically a Gaussian distribution. The user can specify the standard deviation of this distribution, but this parameter does not provide a full description of the surface reflectance properties. We propose a different approach based on 3D measurements of the surface using atomic force microscopy (AFM). Polished and rough (unpolished) crystals were scanned to compute the surface reflectance properties. The angular distributions of reflectance and reflected rays were computed and stored in look-up tables (LUTs). The LUTs account for the effect of incidence angle and were integrated in a light transport model. Crystals of different sizes were simulated with and without reflector. The simulated maximum light output and the light output as a function of DOI showed very good agreement with experimental characterization of the crystals, indicating that our approach provides an accurate model of polished and rough surfaces and could be used to predict light collection in scintillators. This model is based on a true 3D representation of the surface, makes no assumption about the surface and provides insight on the optical behaviour of rough crystals that can play a critical role in optimizing the design of PET detectors. This approach is also compatible with existing simulation toolkits and next steps include the implementation in GATE.
Positron emission tomography (PET); PET detectors; light transport; rough surface; light collection; reflectance; simulation; AFM; unpolished crystals
Cardiac motion and Partial Volume Effects (PVE) are two of the main causes of image degradation in cardiac PET. Motion generates artifacts and blurring while PVE lead to erroneous myocardial activity measurements. Newly available simultaneous PET-MR scanners offer new possibilities in cardiac imaging as MRI can assess wall contractility while collecting PET perfusion data. In this perspective, we develop a list-mode iterative reconstruction framework incorporating both tagged-MR derived non-rigid myocardial wall motion and position dependent detector Point Spread Function (PSF) directly into the PET system matrix. In this manner, our algorithm performs both motion “deblurring” and PSF deconvolution while reconstructing images with all available PET counts. The proposed methods are evaluated in a beating non-rigid cardiac phantom whose hot myocardial compartment contains small transmural and non-transmural cold defects. In order to accelerate imaging time, we investigate collecting full and half k-space tagged MR data to obtain tagged volumes that are registered using non-rigid B-spline registration to yield wall motion information. Our experimental results show that tagged-MR based motion correction yielded an improvement in defect/myocardium contrast recovery of 34-206% as compared to motion uncorrected studies. Likewise, lesion detectability improved by respectively 115-136% and 62-235% with MR-based motion compensation as compared to gating and no motion correction and made it possible to distinguish non-transmural from transmural defects, which has clinical significance given inherent limitations of current single modality imaging in identifying the amount of residual ischemia. The incorporation of PSF modeling within the framework of MR-based motion compensation significantly improved defect/myocardium contrast recovery (5.1-8.5%, p<0.01) and defect detectability (39-56%, p<0.01). No statistical difference was found in PET contrast and lesion detectability based on motion fields obtained with half and full k-space tagged data.
This study investigated the dosimetric impact of uncompensated motion and motion compensation with dynamic multileaf collimator (DMLC) tracking for prostate intensity modulated arc therapy. Two treatment approaches were investigated; a conventional approach with a uniform radiation dose to the target volume and an intraprostatic lesion (IPL) boosted approach with an increased dose to a subvolume of the prostate. The impact on plan quality of optimizations with a leaf position constraint, which limited the distance between neighbouring adjacent MLC leaves, was also investigated. Deliveries were done with and without DMLC tracking on a linear acceleration with a high-resolution MLC. A cylindrical phantom containing two orthogonal diode arrays was used for dosimetry. A motion platform reproduced six patient-derived prostate motion traces, with the average displacement ranging from 1.0 to 8.9 mm during the first 75 seconds. A research DMLC tracking system was used for real-time motion compensation with optical monitoring for position input. The gamma index was used for evaluation, with measurements with a static phantom or the planned dose as reference, using 2% and 2 mm gamma criteria. The average pass rate with DMLC tracking was 99.9% (range 98.7–100%, measurement as reference), whereas the pass rate for untracked deliveries decreased distinctly as the average displacement increased, with an average pass rate of 61.3% (range 32.7–99.3%). Dose-volume histograms showed that DMLC tracking maintained the planned dose distributions in the presence of motion whereas traces with > 3 mm average displacement caused clear plan degradation for untracked deliveries. The dose to the rectum and bladder had an evident dependence on the motion direction and amplitude for untracked deliveries, and the dose to the rectum was slightly increased for IPL boosted plans compared to conventional plans for anterior motion with large amplitude. In conclusion, optimization using a leaf position constraint had minimal dosimetric effect, DMLC tracking improved the target and normal tissue dose distributions compared to no tracking for target motion >3 mm, with the DMLC tracking distributions showing generally good agreement between the planned and delivered doses.
motion management; intrafraction motion; integrated boost
The goal of this study was to evaluate the nature of the stem effect light produced within an optical fiber, to quantify its composition, and to evaluate the efficiency of the chromatic technique to remove the stem effect. Spectrometry studies were performed during irradiations of a bare PMMA optical fiber with kilovoltage x-rays from a superficial therapy unit, an Ir-192 high-dose-rate brachytherapy source, a Co-60 external-therapy unit, and megavoltage electrons and x-rays from a linear accelerator. Stem effect spectra can be accurately modeled by a linear combination of the Cerenkov light and fluorescence emitted spectra. Fluorescence light contributes more for lower-energy modalities. Cerenkov light contributes more as the energy increases above the threshold for its production. The chromatic stem effect removal technique is accurate in most of the situations. However, noticeable differences were obtained between very specific high-energy irradiation conditions. It would be advantageous to implement an additional channel in the chromatic stem effect removal chain or implement a spectral approach to independently remove the Cerenkov and the fluorescence components from the signal of interest. This would increase the accuracy and versatility of the actual chromatic stem effect removal technique.
fluorescence; Cerenkov; scintillation; fiber; stem effect
Contrast-enhanced ultrasound (US) imaging is potentially applicable to the clinical investigation of a wide variety of perfusion disorders. Quantitative analysis of perfusion is not widely performed, and is limited by the fact that data are acquired from a single tissue plane, a situation that is unlikely to accurately reflect global perfusion. Real-time perfusion information from a tissue volume in an experimental rabbit model of testicular torsion was obtained with a two-dimensional matrix phased array US transducer. Contrast-enhanced imaging was performed in 20 rabbits during intravenous infusion of the microbubble contrast agent Definity® before and after unilateral testicular torsion and contralateral orchiopexy. The degree of torsion was 0° in 4 (sham surgery), 180° in 4, 360° in 4, 540° in 4, and 720° in 4. An automated technique was developed to analyze the time history of US image intensity in experimental and control testes. Comparison of mean US intensity rate of change and of ratios between mean US intensity rate of change in experimental and control testes demonstrated good correlation with testicular perfusion and mean perfusion ratios obtained with radiolabeled microspheres, an accepted “gold standard”. This method is of potential utility in the clinical evaluation of testicular and other organ perfusion.
We present a method based on spectral theory for the shape analysis of carpal bones of the human wrist. We represent the cortical surface of the carpal bone in a coordinate system based on the eigensystem of the two-dimensional Helmholtz equation. We employ a metric—global point signature (GPS)—that exploits the scale and isometric invariance of eigenfunctions to quantify overall bone shape. We use a fast finite-element-method to compute the GPS metric. We capitalize upon the properties of GPS representation—such as stability, a standard Euclidean (ℓ2) metric definition, and invariance to scaling, translation and rotation—to perform shape analysis of the carpal bones of ten women and ten men from a publicly-available database. We demonstrate the utility of the proposed GPS representation to provide a means for comparing shapes of the carpal bones across populations.
carpal bone; shape comparison; global point signature; morphometry; eigenfunctions; wrist
Full quantitative analysis of brain PET data requires knowledge of the arterial input function into the brain. Such data are normally acquired by arterial sampling with corrections for delay and dispersion to account for the distant sampling site. Several attempts have been made to extract an image-derived input function (IDIF) directly from the internal carotid arteries that supply the brain and are often visible in brain PET images. We have devised a method of delineating the internal carotids in co-registered MR images using the level-set method and applying the segmentations to PET images using a novel centerline approach. Centerlines of the segmented carotids were modeled as cubic splines and re-registered in PET images summed over the early portion of the scan. Using information from the anatomical center of the vessel should minimize partial volume and spillover effects. Centerline time-activity curves were taken as the mean of the values for points along the centerline interpolated from neighboring voxels. A scale factor correction was derived from calculation of cerebral blood flow (CBF) using gold standard arterial blood measurements. We have applied the method to human subject data from multiple injections of [15O]water on the HRRT. The method was assessed by calculating the area under the curve (AUC) of the IDIF and the CBF, and comparing these to values computed using the gold standard arterial input curve. The average ratio of IDIF to arterial AUC (apparent recovery coefficient: aRC) across 9 subjects with multiple (n = 69) injections was 0.49 ± 0.09 at 0–30 s post tracer arrival, 0.45 ± 0.09 at 30–60 s, and 0.46 ± 0.09 at 60–90 s. Grey and white matter CBF values were 61.4 ± 11.0 and 15.6 ± 3.0 mL/min/100g tissue using sampled blood data. Using IDIF centerlines scaled by the average aRC over each subjects’ injections, gray and white matter CBF values were 61.3 ± 13.5 and 15.5 ± 3.4 mL/min/100g tissue. Using global average aRC values, the means were unchanged, and intersubject variability was noticeably reduced. This MR-based centerline method with local re-registration to [15O]water PET yields a consistent IDIF over multiple injections in the same subject, thus permitting the absolute quantification of CBF without arterial input function measurements.
Accurate quantification of pharmacokinetic parameters in dynamic contrast-enhanced (DCE) MRI may be affected by the passive diffusion of contrast agent (CA) within the tissue. By introducing an additional term into the standard Tofts-Kety (STK) model, we correct for the effects of CA diffusion. We first develop the theory describing a CA diffusion corrected STK model (DTK). The model is then tested in simulation with simple models of diffusion. The DTK model is also fit to 18 in vivo DCE-MRI acquisitions from murine models of cancer and results are compared to those from the STK model. The DTK model returned estimates with significantly lower error than the STK model (p≪0.001). In poorly-perfused (i.e., Ktrans≤0.05 min−1) regions the STK model returned unphysical ve values, while the DTK model estimated ve with less than 7% error in noise-free simulations. Results in vivo data revealed similar trends. For voxels with low Ktrans values and late peak concentration times the STK model returned ve estimates >1.0 in 40% of the voxels as compared to only 16% for the DTK model. The DTK model presented here shows promise in estimating accurate kinetic parameters in the presence of passive contrast agent diffusion.
DCE-MRI; contrast agent diffusion; cancer
An important element of radiation treatment planning for cancer therapy is the selection of beam angles (out of all possible coplanar and non-coplanar angles in relation to the patient) in order to maximize the delivery of radiation to the tumor site and minimize radiation damage to nearby organs-at-risk. This category of combinatorial optimization problem is particularly difficult because direct evaluation of the quality of treatment corresponding to any proposed selection of beams requires the solution of a large-scale dose optimization problem involving many thousands of variables that represent doses delivered to volume elements (voxels) in the patient. However, if the quality of angle sets can be accurately estimated without expensive computation, a large number of angle sets can be considered, increasing the likelihood of identifying a very high quality set. Using a computationally efficient surrogate beam set evaluation procedure based on single-beam data extracted from plans employing equally-spaced beams (eplans), we have developed a global search metaheuristic process based on the Nested Partitions framework for this combinatorial optimization problem. The surrogate scoring mechanism allows us to assess thousands of beam set samples within a clinically acceptable time frame. Tests on difficult clinical cases demonstrate that the beam sets obtained via our method are superior quality.
Previous research has demonstrated that a parameter extracted from a power function fit to the anatomical noise power spectrum, β, may be predictive of breast mass lesion detectability in x-ray based medical images of the breast. In this investigation, the value of β was compared with a number of other more widely used parameters, in order to determine the relationship between β and these other parameters.
This study made use of breast CT data sets, acquired on two breast CT systems developed in our laboratory. A total of 185 breast data sets in 183 women were used, and only the unaffected breast was used (where no lesion was suspected). The anatomical noise power spectrum was computed from two-dimensional regions of interest (ROI), was fit to a power function (NPS(f) = αf−β), and the exponent parameter (β) was determined using log/log linear regression. Breast density for each of the volume data sets was characterized in previous work. The breast CT data sets analyzed in this study were part of a previous study which evaluated the receiver operating characteristic (ROC) curve performance using simulated spherical lesions and a pre-whitened matched filter computer observer. This ROC information was used to compute the detectability index as well as the sensitivity at 95% specificity. The fractal dimension was computed from the same ROI's which were used for the assessment of β. The value of β was compared to breast density, detectability index, sensitivity, and fractal dimension, and the slope of these relationships was investigated to assess statistical significance from zero slope. A statistically significant non-zero slope was considered to be a positive association in this investigation.
All comparisons between β and breast density, detectability index, sensitivity at 95% specificity, and fractal dimension demonstrated statistically significant association with p < 0.001 in all cases. The value of β was also found to be associated with patient age and breast diameter, parameters both related to breast density. In all associations between other parameters, lower values of β were associated with increased breast cancer detection performance. Specifically, lower values of β were associated with lower breast density, higher detectability index, higher sensitivity, and lower fractal dimension values.
While causality was not and probably cannot be demonstrated, the strong, statistically significant association between the β metric and the other more widely used parameters suggest that β may be considered as a surrogate measure for breast cancer detection performance. These findings are specific to breast parenchymal patterns and mass lesions only.
Two new fluoroscopic fiducial tracking methods that exploit the spatial relationship among the multiple implanted fiducial to achieve fast, accurate and robust tracking are proposed in this paper. The spatial relationship between multiple implanted markers are modeled as Gaussian distributions of their pairwise distances over time. The means and standard deviations of these distances are learned from training sequences, and pairwise distances that deviate from these learned distributions are assigned a low spatial matching score. The spatial constraints are incorporated in two different algorithms: a stochastic tracking method and a detection based method. In the stochastic method, hypotheses of the “true” fiducial position are sampled from a pre-trained respiration motion model. Each hypothesis is assigned an importance value based on image matching score and spatial matching score. Learning the parameters of the motion model is needed in addition to the learning the distribution parameters of the pairwise distances in the proposed stochastic tracking approach. In the detection based method, a set of possible marker locations are identified by using a template matching based fiducial detector. The best location is obtained by optimizing the image matching score and spatial matching score through non-serial dynamic programming. In this detection based approach, there is no need to learn the respiration motion model. The two proposed algorithms are compared with a recent work using multiple hypothesis tracking algorithm which is denoted by MHT. Phantom experiments were performed using fluoroscopic videos captured with known motion relative to an anthropomorphic phantom. The patient experiments were performed using a retrospective study of 16 fluoroscopic videos of liver cancer patients with implanted fiducials. For the motion phantom data sets, the detection based approach has the smallest tracking error (μerr: 0.78 – 1.74 mm, σerr: 0.39 – 1.16 mm) for the images taken at low exposure (50 mAs). At higher exposure (500 mAs), the stochastic method gave the best performance (μerr:~ 0.39 mm, σerr:~ 0.27 mm). In contrast, the tracker (MHT) that does not model the spatial constraints only performs well when there is no occluded fiducial. With the RANDO phantom data, both of our proposed methods performed well and have the mean tracking errors around ~ 1.8 mm with the standard deviations ~ 0.93 mm at 100 mAs and ~ 0.91 mm with 0.88 mm standard deviation at 500 mAs. The MHT tracker has the largest tracking errors with mean ~ 4.8 mm) and standard deviation ~ 2.4 mm in both sessions with the Rondo phantom data. On the patient data sets, the detection based method gave the smallest error (μerr: 0.39 mm, σerr: ~ 0.19 mm). The stochastic method performed well (μerr: ~ 0.58 mm, σerr: ~ 0.39 mm) when the patient breathed consistently, the accuracy dropped to (μerr: ~ 1.55 mm) when the patient breathed differently across sessions.
Tumor blood volume and vascular permeability are well established indicators of tumor angiogenesis and important predictors in cancer diagnosis, planning, and treatment. In this work, we establish a novel preclinical imaging protocol which allows quantitative measurement of both metrics simultaneously. First, gold nanoparticles are injected and allowed to extravasate into the tumor, and then liposomal iodine nanoparticles are injected. Combining a previously optimized dual energy micro-CT scan using high-flux polychromatic x-ray sources (energies: 40 kVp, 80 kVp) with a novel post-reconstruction spectral filtration scheme, we are able to decompose the results into 3D iodine and gold maps, allowing simultaneous measurement of extravasated gold and intravascular iodine concentrations. Using a digital resolution phantom, the mean limits of detectability (mean CNR = 5) for each element are determined to be 2.3 mg/mL (18 mM) for iodine and 1.0 mg/mL (5.1 mM) for gold, well within the observed in vivo concentrations of each element (I: 0-24 mg/mL, Au: 0-9 mg/mL) and a factor of 10 improvement over the limits without post-reconstruction spectral filtration. Using a calibration phantom, these limits are validated and an optimal sensitivity matrix for performing decomposition using our micro-CT system is derived. Finally, using a primary mouse model of soft-tissue sarcoma, we demonstrate the in vivo application of the protocol to measure fractional blood volume and vascular permeability over the course of five days of active tumor growth.
Micro-CT; dual energy; small animal imaging; gold nanoparticles
Fluorescence tomography (FT) is a promising molecular imaging technique that can spatially resolve both fluorophore concentration and lifetime parameters. However, recovered fluorophore parameters highly depend on the size and depth of the object due to the ill-posedness of the FT inverse problem. Structural a priori information from another high spatial resolution imaging modality has been demonstrated to significantly improve FT reconstruction accuracy. In this study, we have constructed a combined magnetic resonance imaging (MRI) and FT system for small animal imaging. A photo-multiplier tube (PMT) is used as the detector to acquire frequency domain FT measurements. This is the first MR-compatible time-resolved FT system that can reconstruct both fluorescence concentration and lifetime maps simultaneously. The performance of the hybrid system is evaluated with phantom studies. Two different fluorophores, Indocyanine Green (ICG) and 3-3′ Diethylthiatricarbocyanine Iodide (DTTCI), which have similar excitation and emission spectra but different lifetimes, are utilized. The fluorescence concentration and lifetime maps are both reconstructed with and without the structural a priori information obtained from MRI for comparison. We show that the hybrid system can accurately recover both fluorescence intensity and lifetime within 10% error for two 4.2 mm-diameter cylindrical objects embedded in a 38 mm-diameter cylindrical phantom when MRI structural a priori information is utilized.
It has been widely believed that the non-negativity and the piecewise-constant constraints guarantee a unique solution to an interior tomographic problem. This letter points out that if the number of views is finite, the non-negativity and the piecewise-constant constraints do not guarantee a unique solution to an interior tomographic problem and the reconstruction could be biased by a constant polygon.
Image reconstruction; Computed tomography; Reconstruction algorithms; Inverse problems; Piecewise constant constraint; Compressed sensing; Optimization
In tomographic medical imaging, signal activity is typically estimated by summing voxels from a reconstructed image. We introduce an alternative estimation scheme that operates on the raw projection data and offers a substantial improvement, as measured by the ensemble mean-square error (EMSE), when compared to using voxel values from a maximum-likelihood expectation-maximization (MLEM) reconstruction. The scanning-linear (SL) estimator operates on the raw projection data and is derived as a special case of maximum-likelihood (ML) estimation with a series of approximations to make the calculation tractable. The approximated likelihood accounts for background randomness, measurement noise, and variability in the parameters to be estimated. When signal size and location are known, the SL estimate of signal activity is an unbiased estimator, i.e., the average estimate equals the true value. By contrast, standard algorithms that operate on reconstructed data are subject to unpredictable bias arising from the null functions of the imaging system. The SL method is demonstrated for two different tasks: 1) simultaneously estimating a signal's size, location, and activity; 2) for a fixed signal size and location, estimating activity. Noisy projection data are realistically simulated using measured calibration data from the multi-module multi-resolution (M3R) small-animal SPECT imaging system. For both tasks the same set of images is reconstructed using the MLEM algorithm (80 iterations), and the average and the maximum value within the ROI are calculated for comparison. This comparison shows dramatic improvements in EMSE for the SL estimates. To show that the bias in ROI estimates affects not only absolute values but also relative differences, such as those used to monitor response to therapy, the activity estimation task is repeated for three different signal sizes.
Time-of-flight (TOF) and point spread function (PSF) modeling have been shown to improve PET reconstructions, but the impact on physicians in the clinical setting has not been thoroughly investigated. A lesion detection and localization study was performed using simulated lesions in real patient images. Four reconstruction schemes were considered: ordinary Poisson OSEM (OP) alone and combined with TOF, PSF, and TOF+PSF. The images were presented to physicians experienced in reading PET images, and the performance of each was quantified using localization receiver operating characteristic (LROC). Numerical observers (non-prewhitening and Hotelling) were used to identify optimal reconstruction parameters, and observer SNR was compared to the performance of the physicians. The numerical models showed good agreement with human performance, and best performance was achieved by both when using TOF+PSF. These findings suggest a large potential benefit of TOF+PSF for oncology PET studies, especially in the detection of small, low-intensity, focal disease in larger patients.
Diffuse optical imaging using non-ionizing radiation is a non-invasive method that shows promise towards breast cancer diagnosis. Hand-held optical imagers show potential for clinical translation of the technology, yet they have not been used towards 3D tomography. Herein, 3D tomography of human breast tissue in vivo is demonstrated for the first time using a hand-held optical imager with automated coregistration facilities. Simulation studies are performed on breast geometries to demonstrate the feasibility of 3D tomographic imaging using a hand-held imager under perfect (1:0) and imperfect (100:1, 50:1) fluorescence absorption contrast ratios. Experimental studies are performed in vivo using a 1 μM ICG filled phantom target placed non-invasively underneath the flap of the breast tissue. Results show the ability to perform automated tracking and coregistered imaging of human breast tissue (with tracking accuracy on the order of ~1 cm). Three-dimensional tomography results demonstrated the ability to recover a single target placed at a depth of 2.5 cm, from both the simulated (at 1:0, 100:1 and 50:1 contrasts) and experimental cases on actual breast tissues. Ongoing efforts to improve target depth recovery are carried out via implementation of transmittance imaging in the hand-held imager.
A comprehensive set of photon fluence-to-dose response functions (DRFs) are presented for two radiosensitive skeletal tissues – active and total shallow marrow – within 15 and 32 bones sites, respectively, of the ICRP reference adult male. The functions were developed using fractional skeletal masses and associated electron absorbed fractions as reported for the UF hybrid adult male phantom, which in turn is based upon microCT images of trabecular spongiosa taken from a 40-year male cadaver. The new DRFs expand upon both the original set of seven functions produced in 1985, as well as a 2007 update calculated under the assumption of secondary electron escape from spongiosa. In the present study, it is assumed that photon irradiation of the skeleton will yield charged particle equilibrium across all spongiosa regions at energies exceeding 200 keV. Kerma factors for active marrow, inactive marrow, trabecular bone, and spongiosa at higher energies are calculated using the DRF algorithm setting the electron absorbed fraction for self-irradiation to unity. By comparing kerma factors and DRF functions, dose enhancement factors and mass energy-absorption coefficient (MEAC) ratios for active marrow to spongiosa were derived. These MEAC ratios compared well with those provided by the NIST Physical Reference Data Library (mean difference of 0.8%), and the dose enhancement factors for active marrow compared favorably with values calculated in the well-known study published by King and Spiers (1985) (mean absolute difference of 1.9 percentage points). Additionally, dose enhancement factors for active marrow were shown to correlate well with the shallow marrow volume fraction (R2 = 0.91). Dose enhancement factors for the total shallow marrow were also calculated for 32 bone sites
bone dosimetry; dose-response function; active marrow; total shallow marrow
Target tissues include the active bone marrow, associated with radiogenic leukemia, and total shallow marrow, associated with radiogenic bone cancer. Monoenergetic electron emissions are considered over the energy range 1 keV to 10 MeV for the following sources: bone marrow (active and inactive), trabecular bone (surfaces and volumes), and cortical bone (surfaces and volumes). Specific absorbed fractions are computed according to the MIRD schema, and are given as skeletal-averaged values in the paper with site-specific values reported in both tabular and graphical format in an electronic annex. The distribution of cortical bone and spongiosa at the macroscopic dimensions of the phantom, as well as the distribution of trabecular bone and marrow tissues at the microscopic dimensions of the phantom, are imposed through detailed analyses of whole-body ex-vivo CT images (1 mm resolution) and spongiosa-specific ex-vivo microCT images (30 μm resolution), respectively, taken from a 40-year male cadaver. The method utilized in this work includes: (1) explicit accounting for changes in marrow self-dose with variations in marrow cellularity, (2) explicit accounting for electron escape from spongiosa, (3) explicit consideration of spongiosa cross-fire from cortical bone, and (4) explicit consideration of the ICRP’s change in the surrogate tissue region defining the location of the osteoprogenitor cells (from a 10-μm endosteal layer covering the trabecular and cortical surfaces, to a 50-μm shallow marrow layer covering trabecular and medullary cavity surfaces). Skeletal-averaged values of absorbed fraction in the present model are noted to be very compatible with those weighted by the skeletal tissue distributions found in the ICRP Publication 110 adult male and female voxel phantoms, but are in many cases incompatible with values used in current and widely implemented internal dosimetry software.
skeletal dosimetry; bone dosimetry; hematopoietic stem cells; osteoprogenitor cells; reference adult male; electron transport
In many studies, the estimation of the apparent diffusion coefficient (ADC) of lesions in visceral organs in diffusion-weighted (DW) magnetic resonance images requires an accurate lesion-segmentation algorithm. To evaluate these lesion-segmentation algorithms, region-overlap measures are used currently. However, the end task from the DW images is accurate ADC estimation, and the region-overlap measures do not evaluate the segmentation algorithms on this task. Moreover, these measures rely on the existence of gold-standard segmentation of the lesion, which is typically unavailable. In this paper, we study the problem of task-based evaluation of segmentation algorithms in DW imaging in the absence of a gold standard. We first show that using manual segmentations instead of gold-standard segmentations for this task-based evaluation is unreliable. We then propose a method to compare the segmentation algorithms that does not require gold-standard or manual segmentation results. The no-gold-standard method estimates the bias and the variance of the error between the true ADC values and the ADC values estimated using the automated segmentation algorithm. The method can be used to rank the segmentation algorithms on the basis of both accuracy and precision. We also propose consistency checks for this evaluation technique.
Task-based evaluation; Segmentation; No-gold-standard; Apparent diffusion coefficient; Diffusion-weighted magnetic resonance imaging
Uncertainties in the estimated mean excitation energies (I-values) needed for calculating proton stopping powers can be in the order of 10–15%, which introduces a fundamental limitation in the accuracy of proton range determination. Previous efforts have quantified shifts in proton depth dose distributions due to I-value uncertainties in water and homogenous tissue phantoms. This study is the first to quantify the clinical impact of I-value uncertainties on proton dose distributions within patient geometries. A previously developed Geant4 based Monte Carlo code was used to simulate a proton treatment plan for three patients (prostate, pancreases, and liver) with varying tissue I-values. A uniform variation study was conducted in which the tissue I-values were varied by ±5% and ±10% of the nominal values as well as a probabilistic variation study in which the I-values were randomly sampled according to a normal distribution with the mean equal to the nominal I-value and a standard deviation of 5 and 10% of the nominal values. Modification of tissue I-values impacted both the proton range and SOBP width. R90 range shifts up to 7.7 mm (4.4.%) and R80 range shifts up to 4.8 mm (1.9%) from the nominal range were recorded. Modulating the tissue I-values by 10% the nominal value resulted in up to a 3.5% difference mean dose in the target volumes and organs at risk (OARs) compared to the nominal case. The range and dose differences were the largest for the deeper-seated prostate and pancreas cases. The treatments that were simulated with randomly sampled I-values resulted in range and dose differences that were generally within the upper and lower bounds set by the 10% uniform variations. This study demonstrated the impact of I-value uncertainties on patient dose distributions. Clearly, sub-millimeter precision in proton therapy would necessitate a reduction in I-value uncertainties to ensure an efficacious clinical outcome.
Pediatric patients who received radiation therapy are at risk of developing side effects like radiogenic second cancer. We compared proton and photon therapies in terms of the predicted risk of second cancers for a 4-year-old medulloblastoma patient receiving craniospinal irradiation (CSI). Two CSI treatment plans with 23.4 Gy or Gy (RBE) prescribed dose were computed: a three-field 6-MV photon therapy plan and a four-field proton therapy plan. The primary doses for both plans were determined using a commercial treatment planning system. Stray radiation doses for proton therapy were determined from Monte Carlo simulations, and stray radiation doses for photon therapy were determined from measured data. Dose-risk models based on the Biological Effects of Ionization Radiation VII report were used to estimate risk of second cancer in eight tissues/organs. Baseline predictions of the relative risk for each organ were always less for proton CSI than for photon CSI at all attained ages. The total lifetime attributable risks of the incidence of second cancer considered after proton CSI and photon CSI were 7.7% and 92%, respectively, and the ratio of lifetime risk was 0.083. Uncertainty analysis revealed that the qualitative findings of this study were insensitive to any plausible changes of dose-risk models and mean radiation weighting factor for neutrons. Proton therapy confers lower predicted risk of second cancer than photon therapy for the pediatric medulloblastoma patient.
Proton therapy; second cancer; medulloblastoma; craniospinal irradiation; comparative treatment planning
We introduce a method for denoising dynamic PET data, spatio-temporal expectation-maximization (STEM) filtering, that combines 4-dimensional Gaussian filtering with EM deconvolution. The initial Gaussian filter suppresses noise at a broad range of spatial and temporal frequencies and EM deconvolution quickly restores the frequencies most important to the signal. We aim to demonstrate that STEM filtering can improve variance in both individual time frames and in parametric images without introducing significant bias.
We evaluate STEM filtering with a dynamic phantom study, and with simulated and human dynamic PET studies of a tracer with reversible binding behaviour, [C-11]raclopride, and a tracer with irreversible binding behaviour, [F-18]FDOPA. STEM filtering is compared to a number of established 3 and 4-dimensional denoising methods. STEM filtering provides substantial improvements in variance in both individual time frames and in parametric images generated with a number of kinetic analysis techniques while introducing little bias. STEM filtering does bias early frames, but this does not affect quantitative parameter estimates. STEM filtering is shown to be superior to the other simple denoising methods studied. STEM filtering is a simple and effective denoising method that could be valuable for a wide range of dynamic PET applications.
Dynamic positron emission tomography; 4-dimensional denoising; kinetic analysis; parametric image
Accurate understanding and modeling of respiration-induced uncertainties is essential in image-guided radiotherapy. Explicit modeling of overall lung motion and interaction among different organs promises to be a useful approach. Recently, preliminary studies on 3D fluoroscopic treatment imaging and tumor localization based on Principal Component Analysis (PCA) motion models and cost function optimization have shown encouraging results. However, the performance of this technique for varying breathing parameters and under realistic conditions remains unclear and thus warrants further investigation. In this work, we present a systematic evaluation of a 3D fluoroscopic image generation algorithm via two different approaches. In the first approach the model’s accuracy is tested for changing parameters for sinusoidal breathing. These parameters included changing respiratory motion amplitude, period, and baseline shift. The effects of setup error, imaging noise and different tumor sizes are also examined. In the second approach, we test the model for anthropomorphic images obtained from a modified XCAT phantom. This set of experiments is important as all the underlying breathing parameters are simultaneously tested, as in realistic clinical conditions. Based on our simulation results for more than 250 seconds of breathing data for 8 different lung patients, the overall tumor localization accuracy of the model in left-right (LR), anterior-posterior (AP) and superior-inferior (SI) directions are 0.1 ± 0.1 mm, 0.5 ± 0.5 mm and 0.8 ± 0.8 mm respectively. 3D tumor centroid localization accuracy is 1.0 ± 0.9 mm.
Age-related macular degeneration is a leading cause of vision loss for the elderly population of industrialized nations. The IRay® Radiotherapy System, developed by Oraya® Therapeutics, Inc., is a stereotactic low-voltage irradiation system designed to treat the wet form of the disease. The IRay System uses three robotically positioned 100 kVp collimated photon beams to deliver an absorbed dose of up to 24 Gy to the macula. The present study uses the Monte Carlo radiation transport code MCNPX to assess absorbed dose to six non-targeted tissues within the eye - total lens, radiosensitive tissues of the lens, optic nerve, distal tip of the central retinal artery, non-targeted portion of the retina, and the ciliary body – all as a function of eye size and beam entry angle. The ocular axial length was ranged from 20 to 28 mm in 2 mm increments, with the polar entry angle of the delivery system varied from 18 to 34 degrees in 2 degree increments. The resulting data showed insignificant variations in dose for all eye sizes. Slight variations in the dose to the optic nerve and the distal tip of the central retinal artery were noted as the polar beam angle changed. An increase in non-targeted retinal dose was noted as the entry angle increased, while the dose to the lens, sensitive volume of the lens, and ciliary body decreased as the treatment polar angle increased. Polar angles of 26 degrees or greater resulted in no portion of the sensitive volume of the lens receiving an absorbed dose of 0.5 Gy or greater. All doses to non-targeted structures reported in this study were less than accepted thresholds for post-procedure complications.
Age-related macular degeneration (AMD); stereotactic radiosurgery (SRS); non-targeted tissue complication probability (NTCP); dose volume histogram (DVH)