Specific absorbed fractions from the image-based VIP-Man body model and EGS4-VLSI Monte Carlo code: internal electron emitters

VIP-Man is a whole-body anatomical model newly developed at Rensselaer from the high-resolution colour images of the National Library of Medicine's Visible Human Project. This paper summarizes the use of VIP-Man and the Monte Carlo method to calculate specific absorbed fractions from internal electron emitters. A specially designed EGS4 user code, named EGS4-VLSI, was developed to use the extremely large number of image data contained in the VIP-Man. Monoenergetic and isotropic electron emitters with energies from 100 keV to 4 MeV are considered to be uniformly distributed in 26 organs. This paper presents, for the first time, results of internal electron exposures based on a realistic whole-body tomographic model. Because VIP-Man has many organs and tissues that were previously not well defined (or not available) in other models, the efforts at Rensselaer and elsewhere bring an unprecedented opportunity to significantly improve the internal dosimetry.     

Image covariance and lesion detectability in direct fan-beam x-ray computed tomography

We consider noise in computed tomography images that are reconstructed using the classical direct fan-beam filtered backprojection algorithm, from both full- and short-scan data. A new, accurate method for computing image covariance is presented. The utility of the new covariance method is demonstrated by its application to the implementation of a channelized Hotelling observer for a lesion detection task. Results from the new covariance method and its application to the channelized Hotelling observer are compared with results from Monte Carlo simulations. In addition, the impact of a bowtie filter and x-ray tube current modulation on reconstruction noise and lesion detectability are explored for full-scan reconstruction.     

Fluence-to-dose conversion coefficients from monoenergetic neutrons below 20 MeV based on the VIP-man anatomical model

A new set of fluence-to-absorbed dose and fluence-to-effective dose conversion coefficients have been calculated for neutrons below 20 MeV using a whole-body anatomical model, VIP-Man, developed from the high-resolution transverse colour photographic images of the National Library of Medicine's Visible Human Project. Organ dose calculations were performed using the Monte Carlo code MCNP for 20 monoenergetic neutron beams between 1 x 10(-9) MeV and 20 MeV under six different irradiation geometries: anterior-posterior, posterior-anterior, right lateral, left lateral, rotational and isotropic. The absorbed dose for 24 major organs and effective dose results based on the realistic VIP-Man are presented and compared with those based on the simplified MIRD-based phantoms reported in the literature. Effective doses from VIP-Man are not significantly different from earlier results for neutrons in the energy range studied. There are, however, remarkable deviations in organ doses due to the anatomical differences between the image-based and the earlier mathematical models.     

Rapid calculation of detectability in Bayesian single photon emission computed tomography

We consider the calculation of lesion detectability using a mathematical model observer, the channelized Hotelling observer (CHO), in a signal-known-exactly/background-known-exactly detection task for single photon emission computed tomography (SPECT). We focus on SPECT images reconstructed with Bayesian maximum a posteriori methods. While model observers are designed to replace time-consuming studies using human observers, the calculation of CHO detectability is usually accomplished using a large number of sample images, which is still time consuming. We develop theoretical expressions for a measure of detectability, the signal-to-noise-ratio (SNR) of a CHO observer, that can be very rapidly evaluated. Key to our expressions are approximations to the reconstructed image covariance. In these approximations, we use methods developed in the PET literature, but modify them to reflect the different nature of attenuation and distance-dependent blur in SPECT. We validate our expressions with Monte Carlo methods. We show that reasonably accurate estimates of the SNR can be obtained at a computational expense equivalent to approximately two projection operations, and that evaluating SNR for subsequent lesion locations requires negligible additional computation.     

Optimization of a tomosynthesis system for the detection of lung nodules

Mathematical observers that track human performance can be used to reduce the number of human observer studies needed to optimize imaging systems. The performance of human observers for the detection of a 3.6 mm lung nodule in anatomical backgrounds was measured as a function of varying tomosynthetic angle and compared with mathematical observers. The human observer results showed a dramatic increase in the percent of correct responses, from 80% in the projection images to 96% in the projection images with a tomosynthetic angle of just 3 degrees. This result suggests the potential usefulness of the scanned beam digital x-ray system for this application. Given the small number of images (40) used per tomosynthetic angle and the highly nonstationary statistical nature of the backgrounds, the nonprewhitening eye observer achieved a higher performance than the channelized Hotelling observer using a Laguerre-Gauss basis. The channelized Hotelling observer with internal noise and the eye filter matched to the projection data were shown to track human performance as the tomosynthetic angle changed. The validation of these mathematical observers extends their applicability to the optimization of tomosynthesis systems.     

X-ray properties of an anthropomorphic breast phantom for MRI and x-ray imaging

The purpose of this study is to characterize the x-ray properties of a dual-modality, anthropomorphic breast phantom whose MRI properties have been previously evaluated. The goal of this phantom is to provide a platform for optimization and standardization of two- and three-dimensional x-ray and MRI breast imaging modalities for the purpose of lesion detection and discrimination. The phantom is constructed using a mixture of lard and egg whites, resulting in a variable, tissue-mimicking structure with separate adipose- and glandular-mimicking components. The phantom can be produced with either a compressed or uncompressed shape. Mass attenuation coefficients of the phantom materials were estimated using elemental compositions from the USDA National Nutrient Database for Standard Reference and the atomic interaction models from the Monte Carlo code PENELOPE and compared with human values from the literature. The image structure was examined quantitatively by calculating and comparing spatial covariance matrices of the phantom and patient mammography images. Finally, a computerized version of the phantom was created by segmenting a computed tomography scan and used to simulate x-ray scatter of the phantom in a mammography geometry. Mass attenuation coefficients of the phantom materials were within 20% and 15% of the values for adipose and glandular tissues, respectively, which is within the estimation error of these values. Matching was improved at higher energies (>20 keV). Tissue structures in the phantom have a size similar to those in the patient data, but are slightly larger on average. Correlations in the patient data appear to be longer than those in the phantom data in the anterior-posterior direction; however, they are within the error bars of the measurement. Simulated scatter-to-primary ratio values of the phantom images were as high as 85% in some areas and were strongly affected by the heterogeneous nature of the phantom. Key physical x-ray properties of the phantom have been quantitatively evaluated and shown to be comparable to those of breast tissue. Since the MRI properties of the phantom have been previously evaluated, we believe it is a useful tool for quantitative evaluation of two- and three-dimensional x-ray and MRI breast imaging modalities for the purpose of lesion detection and characterization.     

Evaluation of internal noise methods for Hotelling observer models

The inclusion of internal noise in model observers is a common method to allow for quantitative comparisons between human and model observer performance in visual detection tasks. In this article, we studied two different strategies for inserting internal noise into Hotelling model observers. In the first strategy, internal noise was added to the output of individual channels: (a) Independent nonuniform channel noise, (b) independent uniform channel noise. In the second strategy, internal noise was added to the decision variable arising from the combination of channel responses. The standard deviation of the zero mean internal noise was either constant or proportional to: (a) the decision variable's standard deviation due to the external noise, (b) the decision variable's variance caused by the external noise, (c) the decision variable magnitude on a trial to trial basis. We tested three model observers: square window Hotelling observer (HO), channelized Hotelling observer (CHO), and Laguerre-Gauss Hotelling observer (LGHO) using a four alternative forced choice (4AFC) signal known exactly but variable task with a simulated signal embedded in real x-ray coronary angiogram backgrounds. The results showed that the internal noise method that led to the best prediction of human performance differed across the studied model observers. The CHO model best predicted human observer performance with the channel internal noise. The HO and LGHO best predicted human observer performance with the decision variable internal noise. The present results might guide researchers with the choice of methods to include internal noise into Hotelling model observers when evaluating and optimizing medical image quality.     

Effect of x-ray energy dispersion in digital subtraction imaging at the iodine K-edge--a Monte Carlo study

The effect of the energy dispersion of a quasi-monochromatic x-ray beam on the performance of a dual-energy x-ray imaging system is studied by means of Monte Carlo simulations using MCNPX (Monte Carlo N-Particle eXtended) version 2.6.0. In particular, the case of subtraction imaging at the iodine K-edge, suitable for angiographic imaging application, is investigated. The average energies of the two beams bracketing the iodine K-edge are set to the values of 31.2 and 35.6 keV corresponding to the ones obtained with a compact source based on a conventional x-ray tube and a mosaic crystal monochromator. The energy dispersion of the two beams is varied between 0 and 10 keV of full width at half-maximum (FWHM). The signal and signal-to-noise ratio produced in the simulated images by iodine-filled cavities (simulating patient vessels) drilled in a PMMA phantom are studied as a function of the x-ray energy dispersion. The obtained results show that, for the considered energy separation of 4.4 keV, no dramatic deterioration of the image quality is observed with increasing x-ray energy dispersion up to a FWHM of about 2.35 keV. The case of different beam energies is also investigated by means of fast simulations of the phantom absorption.     

Some properties of photon scattering in water phantoms in diagnostic radiology

Monte Carlo simulation was applied to study the histories of photon interactions in a soft-tissue-equivalent medium under diagnostic imaging conditions. We examined the dependence on incident x-ray energy and phantom thickness of the basic properties of photon scattering, including the probabilities of occurrence of the various interaction processes, and the frequency distributions of scattering events. We investigated the properties of scattered radiation for monoenergetic incident x rays, which provide a basis for deriving the physical properties of scattered radiation for any polyenergetic incident beam. We also included four incident x-ray beams with broad spectra; these represented the incident x rays typically used for diagnostic imaging.     

A high-resolution anthropomorphic voxel-based tomographic phantom for proton therapy of the eye

Proton therapy is increasingly used in medical treatments for cancer patients due to the sharp dose conformity offered by the characteristic Bragg peak. Proton beam interactions with the eye will be simulated using the MCNPX Monte Carlo code and available nuclear cross-section data to calculate the dose distribution in the eye gel and surrounding organs. A high-resolution eye model will be employed using a 3D geometrical voxel-based anthropomorphic head phantom obtained from the Visible Human Project (female data). Manual segmentation of the eye, carried out by the Medical Physics group at the University of Surrey resulted in 15 identified structures. This work emphasizes the use of a realistic phantom for accurately predicting dose deposition by protons.     

Physical characteristics of scattered radiation in diagnostic radiology: Monte Carlo simulation studies

We applied Monte Carlo methods for the simulation of x-ray scattering in water phantoms. The phantom thickness was varied from 5 to 20 cm, and the monoenergetic incident x rays were varied from 15 to 100 keV. Eight screen pairs and a total absorption system were used as x-ray receptors. We determined the angular, spectral, and spatial distributions of the scattered radiation and the scatter fractions recorded in the image plane. The dependence of these properties on the incident x-ray energy, the phantom thickness, and the energy response of the recording system was examined. The results of this study provide useful information for the development of antiscatter techniques and for the evaluation of radiographic procedures.     

Penalized maximum-likelihood image reconstruction for lesion detection

Detecting cancerous lesions is one major application in emission tomography. In this paper, we study penalized maximum-likelihood image reconstruction for this important clinical task. Compared to analytical reconstruction methods, statistical approaches can improve the image quality by accurately modelling the photon detection process and measurement noise in imaging systems. To explore the full potential of penalized maximum-likelihood image reconstruction for lesion detection, we derived simplified theoretical expressions that allow fast evaluation of the detectability of a random lesion. The theoretical results are used to design the regularization parameters to improve lesion detectability. We conducted computer-based Monte Carlo simulations to compare the proposed penalty function, conventional penalty function, and a penalty function for isotropic point spread function. The lesion detectability is measured by a channelized Hotelling observer. The results show that the proposed penalty function outperforms the other penalty functions for lesion detection. The relative improvement is dependent on the size of the lesion. However, we found that the penalty function optimized for a 5 mm lesion still outperforms the other two penalty functions for detecting a 14 mm lesion. Therefore, it is feasible to use the penalty function designed for small lesions in image reconstruction, because detection of large lesions is relatively easy.     

Signal detectability in digital radiography: spatial domain figures of merit

The usefulness of Fourier-based measures of imaging performance has come into question for the evaluation of digital imaging systems. Figures of merit such as detective quantum efficiency (DQE) based on Fourier domain parameters are relevant for linear, shift-invariant systems with stationary noise. However, no digital imaging system is shift invariant, and realistic images do not satisfy the stationarity condition. Our methods for the task-based evaluation of imaging systems, based on signal detectability in the spatial domain, do not require such assumptions. We have computed the performance of ideal and quasi-ideal observers for the task of signal detection in digital radiography. Signal detectability in terms of an observer signal-to-noise-ratio (SNR) has been compared to results obtained from a Monte Carlo simulation of the digital image-acquisition process. The simulation incorporates the effects of random amplification and secondary quantum blur, integration over pixel area, and electronic noise. The observer figures of merit that have been previously shown to bracket human performance directly specify the usefulness of the images for the stated diagnostic task. In addition, the observer figures of merit give a task-dependent measure of imaging system efficiency in terms of the ratio of an output SNR2 to an input SNR2. Thus, the concept of "detective quantum efficiency" reappears in a natural way but based in the spatial domain and not dependent on shift invariance and stationarity assumptions. With respect to the optimum amount of system blur, our simulations indicate that under certain task-dependent conditions, large signals are fairly insensitive to blur in the x-ray transducer, while an optimum blur is found for small signals.     

A three-dimensional x-ray scattering system for multi-parameter imaging of the human head

This work examines the suitability of a non-rotating one-side 3D x-ray scatter system for imaging the human head. The system simultaneously produces images of the x-ray attenuation coefficients at two photon energies, as well as an image of the electron density. The system relies on measuring the scattered radiation at two directions orthogonal to an incident beam that scans the object from one side, in addition to the traditionally recorded transmitted radiation. Algorithms for this multi-parameter imaging process are presented, and their numerical viability is demonstrated, using both idealized detector responses and those independently estimated from Monte Carlo simulations. The absorbed radiation dose is also calculated, and was shown to be about one quarter of that of conventional CT systems, for 5 mm spatial-resolution images. The introduced system can therefore be useful in radiotherapy planning, and in post-treatment imaging.     

Sensitivity study of proton radiography and comparison with kV and MV x-ray imaging using GEANT4 Monte Carlo simulations

The imaging sensitivity of proton radiography has been studied and compared with kV and MV x-ray imaging using Monte Carlo simulations. A phantom was specifically modeled using 21 different material inserts with densities ranging from 0.001 to 1.92 g?cm(-3). These simulations were run using the MGH double scattered proton beam, scanned pencil proton beams from 200 to 490 MeV, as well as pure 50 keV, 100 keV, 1 MeV and 2 MeV gamma x-ray beams. In order to compare the physics implied in both proton and photon radiography without being biased by the current state of the art in detector technology, the detectors were considered perfect. Along with spatial resolution, the contrast-to-noise ratio was evaluated and compared for each material. These analyses were performed using radiographic images that took into account the following: only primary protons, both primary and secondary protons, and both contributions while performing angular and energetic cuts. Additionally, tissue-to-tissue contrasts in an actual lung cancer patient case were studied for simulated proton radiographs and compared against the original kV x-ray image which corresponds to the current patient set-up image in the proton clinic. This study highlights the poorer spatial resolution of protons versus x-rays for radiographic imaging purposes, and the excellent density resolution of proton radiography. Contrasts around the tumor are higher using protons in a lung cancer patient case. The high-density resolution of proton radiography is of great importance for specific tumor diagnostics, such as in lung cancer, where x-ray radiography operates poorly. Furthermore, the use of daily proton radiography prior to proton therapy would ameliorate patient set-up while reducing the absorbed dose delivered through imaging.     

Digital mammography image simulation using Monte Carlo

Monte Carlo simulations of digital images of the contrast detail phantom and the ACR phantom are presented for two different x-ray digital mammography modalities: a synchrotron mammography system and a next-generation scanning slot clinical system. A combination of variance reduction methods made it possible to simulate accurate images using real pixel dimensions within reasonable computation times. The complete method of image simulation, including a simple detector response model, a simple noise model, and the incorporation of system effects (MTF), is presented. The simulated images of the phantoms show good agreement with images measured on the two systems.     

Characterization of scattered radiation in kV CBCT images using Monte Carlo simulations

Kilovoltage (kV) cone beam computed tomography (CBCT) images suffer from a substantial scatter contribution. In this study, Monte Carlo (MC) simulations are used to evaluate the scattered radiation present in projection images. These predicted scatter distributions are also used as a scatter correction technique. Images were acquired using a kV CBCT bench top system. The EGSnrc MC code was used to model the flat panel imager, the phantoms, and the x-ray source. The x-ray source model was validated using first and second half-value layers (HVL) and profile measurements. The HVLs and the profile were found to agree within 3% and 6%, respectively. MC simulated and measured projection images for a cylindrical water phantom and for an anthropomorphic head phantom agreed within 8% and 10%. A modified version of the DOSXYZnrc MC code was used to score phase space files with identified scattered and primary particles behind the phantoms. The cone angle, the source-to-detector distance, the phantom geometry, and the energy were varied to determine their effect on the scattered radiation distribution. A scatter correction technique was developed in which the MC predicted scatter distribution is subtracted from the projections prior to reconstruction. Preliminary testing of the procedure was done with an anthropomorphic head phantom and a contrast phantom. Contrast and profile measurements were obtained for the scatter corrected and noncorrected images. An improvement of 3% for contrast between solid water and a liver insert and 11% between solid water and a Teflon insert were obtained and a significant reduction in cupping and streaking artifacts was observed.     

Optimal kvp selection for dual-energy imaging of the chest: evaluation by task-specific observer preference tests

Human observer performance tests were conducted to identify optimal imaging techniques in dual-energy (DE) imaging of the chest with respect to a variety of visualization tasks for soft and bony tissue. Specifically, the effect of kVp selection in low- and high-energy projection pairs was investigated. DE images of an anthropomorphic chest phantom formed the basis for observer studies, decomposed from low-energy and high-energy projections in the range 60-90 kVp and 120-150 kVp, respectively, with total dose for the DE image equivalent to that of a single chest radiograph. Five expert radiologists participated in observer preference tests to evaluate differences in image quality among the DE images. For visualization of soft-tissue structures in the lung, the [60/130] kVp pair provided optimal image quality, whereas [60/140] kVp proved optimal for delineation of the descending aorta in the retrocardiac region. Such soft-tissue detectability tasks exhibited a strong dependence on the low-kVp selection (with 60 kVp providing maximum soft-tissue conspicuity) and a weaker dependence on the high-kVp selection (typically highest at 130-140 kVp). Qualitative examination of DE bone-only images suggests optimal bony visualization at a similar technique, viz., [60/140] kVp. Observer preference was largely consistent with quantitative analysis of contrast, noise, and contrast-to-noise ratio, with subtle differences likely related to the imaging task and spatial-frequency characteristics of the noise. Observer preference tests offered practical, semiquantitative identification of optimal, task-specific imaging techniques and will provide useful guidance toward clinical implementation of high-performance DE imaging systems.     

Monte Carlo simulation of the image formation process in portal imaging

We have written Monte Carlo programs to simulate the formation of radiological images. Our code is used to propagate a simulated x-ray fluence through each component of an existing video-based portal imaging system. This simulated fluence consists of a 512 x 512 pixel image containing both contrast-detail patterns as well as checker patterns to assess spatial resolution of the simulated portal imager. All of the components of the portal imaging system were modeled as a cascade of eight linear stages. Using this code, one can assess the visual impact of changing components in the imaging chain by changing the appropriate probability density function. Virtual experiments were performed to assess the visual impact of replacing the lens and TV camera by an amorphous silicon array, and the effect of scattered radiation on portal images.     

Performance of a PSPMT based detector for scintimammography

In breast scintigraphy, compact detectors with high intrinsic spatial resolution and small inactive peripheries can provide improvements in extrinsic spatial resolution, efficiency and contrast for small lesions relative to larger conventional cameras. We are developing a pixelated small field-of-view gamma camera for scintimammography. Extensive measurements of the imaging properties of a prototype system have been made, including spatial resolution, sensitivity, uniformity of response, geometric linearity and energy resolution. An anthropomorphic torso phantom providing a realistic breast exit gamma spectrum has been used in a qualitative study of lesion detectability. A new type of breast imaging system that combines scintimammography and digital mammography in a single upright unit has also been developed. The system provides automatic co-registration between the scintigram and the digital mammogram, obtained with the breast in a single configuration. Intrinsic spatial resolution was evaluated via calculation of the phase-dependent modulation transfer function (MTF). Measurements of extrinsic spatial resolution, sensitivity and uniformity of response were made for two types of parallel hole collimator using NEMA (National Electrical Manufacturers Association) protocols. Geometric linearity was quantified using a line input and least squares analysis of the measured line shape. Energy resolution was measured for seven different crystal types, and the effectiveness of optical grease coupling was assessed. Exit gamma spectra were obtained using a cadmium zinc telluride based spectrometer. These were used to identify appropriate radioisotope concentrations for the various regions of an anthropomorphic torso phantom, such that realistic scatter conditions could be obtained during phantom measurements. For prone scintimammography, a special imaging table was constructed that permits simultaneous imaging of both breasts, as well as craniocaudal views. A dedicated breast imaging system was also developed that permits simultaneous acquisition and superposition of planar gamma images and digital x-ray images. The intrinsic MTF is nonstationary, and is dependent on the phase relationship between the signal and the crystal array matrix. Averaged over all phases, the MTF is approximately 0.75, 0.57 and 0.40 at spatial frequencies of 1.0, 1.5 and 2.0 cycles per cm, respectively. The phase averaged line spread function (LSF) has a FWHM value of 2.6 mm. Following uniformity corrections, the RMS deviations in flood images are only slightly greater than is predicted from counting statistics. Across an 80 mm section of the active area, the differential linearity is 0.83 mm and the absolute linearity 2.0 mm. Using an anthropomorphic torso phantom with detachable breasts, scatter radiation similar to that observed exiting the breast of scintimammography patients was observed. It was observed that scattered gamma rays can constitute the majority of the radiation incident on the detector, but that the scatter-to-primary ratio varies significantly across the field of view, being greatest in the caudal portion of the breast, where scatter from the liver is high. Using a lesion-to-breast concentration ratio of 6:1, a 1.0 cm3 simulated breast lesion was detectable in lateral images obtained with both the developmental camera and with a clinical camera, while a 0.35 cm3 lesion was detectable in neither. Utilization of the dual x-ray transmission, gamma emission breast imaging system greatly increases the conspicuity of scintimammographic lesions relative to prone imaging, as well as greatly facilitating the localization and identification of structures in the gamma image. The prototype imaging gamma detector exhibits spatial resolution superior to that of conventional cameras, and comparable uniformity of response and geometric linearity. Because of light losses in the crystals, the energy resolution is inferior to that of single crystal NaI(Tl) came