MANTIS: combined x-ray, electron and optical Monte Carlo simulations of indirect radiation imaging systems

We describe MANTIS (Monte carlo x-rAy electroN opTical Imaging Simulation), a tool for simulating imaging systems that tracks x-rays, electrons and optical photons in arbitrary materials and complex geometries. The x-ray and electron transport and involved physics models are from the PENELOPE package, and the optical transport and corresponding physics models are from DETECT-II and include Fresnel refraction and reflection at material boundaries, bulk absorption and scattering. Complex geometries can be handled with the aid of the geometry routines included in PENELOPE. When x-rays or electrons interact and deposit energy in the scintillator, the code generates a number of optical quanta according to a user-selected model for the conversion process. The optical photons are then tracked until they reach an absorption event, which in some cases contributes to the output signal, or escape from the geometry. We demonstrate the capabilities of this new tool with respect to the statistics of the optical signal detected and to the three-dimensional point-response functions corresponding to columnar phosphor screens.     

Modeling granular phosphor screens by Monte Carlo methods

The intrinsic phosphor properties are of significant importance for the performance of phosphor screens used in medical imaging systems. In previous analytical-theoretical and Monte Carlo studies on granular phosphor materials, values of optical properties, and light interaction cross sections were found by fitting to experimental data. These values were then employed for the assessment of phosphor screen imaging performance. However, it was found that, depending on the experimental technique and fitting methodology, the optical parameters of a specific phosphor material varied within a wide range of values, i.e., variations of light scattering with respect to light absorption coefficients were often observed for the same phosphor material. In this study, x-ray and light transport within granular phosphor materials was studied by developing a computational model using Monte Carlo methods. The model was based on the intrinsic physical characteristics of the phosphor. Input values required to feed the model can be easily obtained from tabulated data. The complex refractive index was introduced and microscopic probabilities for light interactions were produced, using Mie scattering theory. Model validation was carried out by comparing model results on x-ray and light parameters (x-ray absorption, statistical fluctuations in the x-ray to light conversion process, number of emitted light photons, output light spatial distribution) with previous published experimental data on Gd2O2S: Tb phosphor material (Kodak Min-R screen). Results showed the dependence of the modulation transfer function (MTF) on phosphor grain size and material packing density. It was predicted that granular Gd2O2S: Tb screens of high packing density and small grain size may exhibit considerably better resolution and light emission properties than the conventional Gd2O2S: Tb screens, under similar conditions (x-ray incident energy, screen thickness).     

Monte Carlo simulations of the imaging performance of metal plate/phosphor screens used in radiotherapy.

The imaging performance of metal plate/phosphor screens which are used for the creation of portal images in radiotherapy is investigated by using Monte Carlo simulations. To this end the modulation transfer function, the noise power spectrum and the detective quantum efficiency [DQE(f)] are calculated for different metals and phosphors and different thicknesses of metal and phosphor for a range of spatial resolutions. The interaction of x-rays with the metal plate/phosphor screen is modeled with the EGS4 electron gamma shower code. Optical transport in the phosphor is modeled by simulating scattering and reabsorption events of individual optical photons. It is shown that metals with a high atomic number perform better than lighter metals in maximizing the DQE(f). It is furthermore shown that the DQE(f) for the metal plate/phosphor screen alone is nearly x-ray quantum absorption limited up to spatial frequencies of 0.4 cycles/mm. In addition, it is argued that the secondary quantum sink of optical photons imposed by the optical chain (mirror, lenses and video camera) leads to a significant degradation of the signal-to-noise ratio at spatial frequencies which are most important for successful registration of portal images. Therefore, the conclusion is that a replacement of the optical chain by a flat array of photodiodes placed directly under the phosphor will lead to a substantial improvement in image quality of portal images.     

hybridMANTIS: a CPU-GPU Monte Carlo method for modeling indirect x-ray detectors with columnar scintillators

The computational modeling of medical imaging systems often requires obtaining a large number of simulated images with low statistical uncertainty which translates into prohibitive computing times. We describe a novel hybrid approach for Monte Carlo simulations that maximizes utilization of CPUs and GPUs in modern workstations. We apply the method to the modeling of indirect x-ray detectors using a new and improved version of the code MANTIS, an open source software tool used for the Monte Carlo simulations of indirect x-ray imagers. We first describe a GPU implementation of the physics and geometry models in fastDETECT2 (the optical transport model) and a serial CPU version of the same code. We discuss its new features like on-the-fly column geometry and columnar crosstalk in relation to the MANTIS code, and point out areas where our model provides more flexibility for the modeling of realistic columnar structures in large area detectors. Second, we modify PENELOPE (the open source software package that handles the x-ray and electron transport in MANTIS) to allow direct output of location and energy deposited during x-ray and electron interactions occurring within the scintillator. This information is then handled by optical transport routines in fastDETECT2. A load balancer dynamically allocates optical transport showers to the GPU and CPU computing cores. Our hybridMANTIS approach achieves a significant speed-up factor of 627 when compared to MANTIS and of 35 when compared to the same code running only in a CPU instead of a GPU. Using hybridMANTIS, we successfully hide hours of optical transport time by running it in parallel with the x-ray and electron transport, thus shifting the computational bottleneck from optical tox-ray transport. The new code requires much less memory than MANTIS and, asa result, allows us to efficiently simulate large area detectors.     

Integrated model for estimating phosphor signal and noise transfer characteristics on medical images: application to CdPO<Subscript>3</Subscript>Cl:Mn phosphor screens

An integrated model describing the signal and noise transfer characteristics of the objective image quality and information content in phosphor-produced images is presented. In the context of this model, important imaging parameters, namely optical gain, modulation transfer function, noise transfer function, detective quantum efficiency and information capacity were experimentally evaluated using seven laboratory-prepared CdPO₃Cl:Mn test phosphor screens of varying coating thickness. This phosphor has been previously shown to exhibit high spectral compatibility properties with the films and optical sensors used in digital imaging systems. Experiments were performed using 50–120 kVp X-rays produced by a medical X-ray unit. Results showed that, for thick screens, optical gain attained peak values close to 200 optical photons per incident X-ray at 50 kVp. The noise transfer function was higher than the modulation transfer function. For the thin screen of 21 mgcm⁻², the modulation transfer function was 0.25 at 100 line pairs mm⁻¹, and the corresponding noise transfer function was 0.4. The detection quantum efficiency peak value was 0.22 at 50 kVp. These values are within acceptable performance limits, and, given the phosphor material's high spectral compatibility and medium temporal response, CdPO₃Cl:Mn could be considered for use in X-ray detectors of static radiography imaging.     

A method for determining the information capacity of x-ray imaging scintillator detectors by means of luminescence and modulation transfer function measurements

A method to determine the information capacity of x-ray phosphor screens used in the detectors of medical imaging systems is described. Information capacity was determined via x-ray luminescence efficiency (XLE), modulation transfer function (MTF) and emission spectrum measurements. The method was applied to laboratory prepared screens from commonly employed phosphor materials. The screen coating weight varied from 50 mg cm⁻² to 140 mg cm⁻². Results indicated that information capacity decreased with screen coating thickness but also depended on intrinsic phosphor properties (density, effective atomic number, intrinsic conversion efficiency, light wavelength). The Gd₂O₂S:Tb phosphor, exhibiting high density and effective atomic number, was found to be superior to La₂O₂S:Tb and Y₂O₂S:Tb.     

Development of high quantum efficiency flat panel detectors for portal imaging: intrinsic spatial resolution

Recently developed flat panel detectors have been proven to have a much better image quality than conventional electronic portal imaging devices (EPIDs). They are, however, not yet the ideal systems for portal imaging application due to the low x-ray absorption, i.e., low quantum efficiency (QE), which is typically on the order of 2-4% as compared to the theoretical limit of 100%. The QE of current flat panel systems can be improved by significantly increasing the thickness of the energy conversion layer (i.e., amorphous selenium or phosphor screen). This, however, will be at the expense of a decrease in spatial resolution mainly due to x-ray scatter in the conversion layer (and also the spread of optical photons in the case of phosphor screen). In this paper, we investigate theoretically the intrinsic spatial resolution of a high QE flat panel detector with a new energy conversion layer that is much denser and thicker than that of current flat panel systems. The modulation transfer function (MTF) of the system is calculated based on a theoretical model using a novel approach, which uses an analytical expression for absorbed dose. It is found that if appropriate materials are used for the conversion layer, then the intrinsic MTF of the high QE flat panel is better than that of current EPIDs, and in addition they have a high QE (e.g., approximately 60%). Some general rules for the design of the conversion layer to achieve both high QE and high resolution as well as high DQE are also discussed.     

A comparison of the imaging characteristics of the new Kodak Hyper Speed G film with the current T-MAT G/RA film and the CR 9000 system

Three standard radiation qualities (RQA 3, RQA 5 and RQA 9) and two screens, Kodak Lanex Regular and Insight Skeletal, were used to compare the imaging performance and dose requirements of the new Kodak Hyper Speed G and the current Kodak T-MAT G/RA medical x-ray films. The noise equivalent quanta (NEQ) and detective quantum efficiencies (DQE) of the four screen-film combinations were measured at three gross optical densities and compared with the characteristics for the Kodak CR 9000 system with GP (general purpose) and HR (high resolution) phosphor plates. The new Hyper Speed G film has double the intrinsic sensitivity of the T-MAT G/RA film and a higher contrast in the high optical density range for comparable exposure latitude. By providing both high sensitivity and high spatial resolution, the new film significantly improves the compromise between dose and image quality. As expected, the new film has a higher noise level and a lower signal-to-noise ratio than the standard film, although in the high frequency range this is compensated for by a better resolution, giving better DQE results--especially at high optical density. Both screen-film systems outperform the phosphor plates in terms of MTF and DQE for standard imaging conditions (Regular screen at RQA 5 and RQA 9 beam qualities). At low energy (RQA 3), the CR system has a comparable low-frequency DQE to screen-film systems when used with a fine screen at low and middle optical densities, and a superior low-frequency DQE at high optical density.     

The physics of computed radiography: measurements of pulse height spectra of photostimulable phosphor screens using prompt luminescence

Computed radiography (CR) is a digital technology that employs reusable photostimulable phosphor (PSP) imaging plates (IP) to acquire radiographic images. In CR, the x-ray attenuation pattern of the imaged object is temporarily stored as a latent charge image within the PSP. The latent image is optically readout as photostimulated luminescence (PSL) when the phosphor is subsequently stimulated using a scanning laser. The multiple stages necessary to create a CR image make it difficult to investigate either experimentally or theoretically. In order to examine the performance of the CR system at a fundamental level separate measurements of the processes involved are desirable. Here pulse height spectroscopy is used to study the prompt violet light emission or prompt luminescence (PL) from commercial PSP screens. Since the mechanism by which light escapes from the phosphor is identical for PL and PSL, observations and conclusions based on the pulse height spectra (PHS) of PL are relevant to the understanding of the behavior of the PSL light emission that outputs the radiographic image in CR. The PL PHS of screens of different thickness and optical properties were measured and compared with the PHS of conventional phosphors. A new method for calibration of the PHS in terms of the absolute number of optical photons per x-ray is introduced and compared to previously established methods.     

A comparison of x-ray detectors for mouse CT imaging

There is significant interest in using computed tomography (CT) for in vivo imaging applications in mouse models of disease. Most commercially available mouse x-ray CT scanners utilize a charge-coupled device (CCD) detector coupled via fibre optic taper to a phosphor screen. However, there has been little research to determine if this is the optimum detector for the specific task of in vivo mouse imaging. To investigate this issue, we have evaluated four detectors, including an amorphous selenium (a-Se) detector, an amorphous silicon (a-Si) detector with a gadolinium oxysulphide (GOS) screen, a CCD with a 3:1 fibre taper and a GOS screen, and a CCD with a 2:1 fibre taper and both GOS and thallium-doped caesium iodide (CsI:Tl) screens. The detectors were evaluated by measuring the modulation transfer function (MTF), noise power spectrum (NPS), detective quantum efficiency (DQE), stability over multiple exposures, and noise in reconstructed CT images. The a-Se detector had the best MTF and the highest DQE (0.6 at 0 lp mm(-1)) but had the worst stability (45% reduction after 2000 exposure frames). The a-Si detector and the CCD with the 3:1 fibre, both of which used the GOS screen, had very similar performance with a DQE of approximately 0.30 at 0 lp mm(-1). For the CCD with the 2:1 fibre, the CsI:Tl screen resulted in a nearly two-fold improvement in DQE over the GOS screen (0.4 versus 0.24 at 0 lp mm(-1)). The CCDs both had the best stability, with less than a 1% change in pixel values over multiple exposures. The pixel values of the a-Si detector increased 5% over multiple exposures due to the effects of image lag. Despite the higher DQE of the a-Se detector, the reconstructed CT images acquired with the a-Si detector had lower noise levels, likely due to the blurring effects from the phosphor screen.     

Segmented phosphors: MEMS-based high quantum efficiency detectors for megavoltage x-ray imaging

Current electronic portal imaging devices (EPIDs) based on active matrix flat panel imager (AMFPI) technology use a metal plate+phosphor screen combination for x-ray conversion. As a result, these devices face a severe trade-off between x-ray quantum efficiency (QE) and spatial resolution, thus, significantly limiting their imaging performance. In this work, we present a novel detector design for indirect detection-based AMFPI EPIDs that aims to circumvent this trade-off. The detectors were developed using micro-electro-mechanical system (MEMS)-based fabrication techniques and consist of a grid of up to approximately 2 mm tall, optically isolated cells of a photoresist material, SU-8. The cells are dimensionally matched to the pixels of the AMFPI array, and packed with a scintillating phosphor. In this paper, various design considerations for such detectors are examined. An empirical evaluation of three small-area (approximately 7 x 7 cm2) prototype detectors is performed in order to study the effects of two design parameters--cell height and phosphor packing density, both of which are important determinants of the imaging performance. Measurements of the x-ray sensitivity, modulation transfer function (MTF) and noise power spectrum (NPS) were performed under radiotherapy conditions (6 MV), and the detective quantum efficiency (DQE) was determined for each prototype SU-8 detector. In addition, theoretical calculations using Monte Carlo simulations were performed to determine the QE of each detector, as well as the inherent spatial resolution due to the spread of absorbed energy. The results of the present studies were compared with corresponding measurements published in an earlier study using a Lanex Fast-B phosphor screen coupled to an indirect detection array of the same design. The SU-8 detectors exhibit up to 3 times higher QE, while achieving spatial resolution comparable or superior to Lanex Fast-B. However, the DQE performance of these early prototypes is significantly lower than expected due to high levels of optical Swank noise. Consequently, the SU-8 detectors presently exhibit DQE values comparable to Lanex Fast-B at zero spatial frequency and significantly lower than Fast-B at higher frequencies. Finally, strategies for reducing Swank noise are discussed and theoretical calculations, based on the cascaded systems model, are presented in order to estimate the performance improvement that can be achieved through such noise reduction.     

Resolution degradation due to K-characteristic photon absorption in dual phosphor image receptors

Image receptors based on the absorption of x rays by phosphor screens have a system resolution described predominantly by the modulation transfer function (MTF) of the phosphor screens utilized. This ideal MTF is modified by the presence of secondary effects within the imaging receptor such as scatter from the receptor cover, visible light crossover (in the case of screen film cassettes) and scatter generated within the phosphor itself, presumed to be almost entirely due to photoelectric interactions. In this paper the MTF characteristics resulting from the absorption of K-characteristic x rays emitted by one screen of a pair in an image receptor and absorbed by the adjacent screen (referred to as cross-fluorescence hereafter) are examined. This effect has been previously modeled [Med. Phys. 23, 1253 (1996)] and recently a computer program has been written to calculate intensity of the radial component of a point spread function (PSF) describing the above process. From this, the MTF of cross-fluorescence under varying conditions can be calculated. The model was applied to a dual screen experiment described in the literature [Med. Phys. 23, 871 (1996)], from which it was possible to show good agreement between calculated results and those derived from the measurement. The effect of screen thickness, screen separation, and phosphor material on the MTF of cross-fluorescence was investigated using simulated x-ray beam spectra in association with phosphor screen parameters. Results showed that the calculated MTF falls to below 0.1 at 1 cy/mm in all cases examined. This MTF was reduced further as the separation between the screens increased from that commonly used in a screen film cassette. Beam energy, phosphor thickness, and phosphor material had a minimal or no effect on MTF. The effect of cross-fluorescence on total image receptor MTF is to reduce the high-frequency components of that function by an amount that is in proportion to the scatter fraction of the effect.     

Resolution at oblique incidence angles of a flat panel imager for breast tomosynthesis

Oblique incidence of x rays on an imaging detector causes blurring that reduces spatial resolution. For simple projection imaging this effect is small and often ignored. However, for breast tomosynthesis, the incidence angle can be larger (>20 degrees), leading to increased blur for some of the projections. The modulation transfer function (MTF) is measured for a typical phosphor-coupled flat-panel detector versus angular incidence of the x-ray beam for two x-ray spectra: 26 kV Mo/Mo and 40 kV Rh/Al. At an incidence angle of 40 degrees the MTF at 5 mm(-1) falls by 35% and 40% for each spectrum, respectively (and 65%/80% at 8 mm(-1)). Increasing the detector absorber thickness to achieve improved quantum efficiency will cause the blurring effect due to beam obliquity to become greater. The impact of this blur is likely to cause misregistration and increased relative noise in tomosynthesis reconstructed images.     

X-ray imaging performance of scintillator-filled silicon pore arrays

The need for fine detail visibility in various applications such as dental imaging, mammography, but also neurology and cardiology, is the driver for intensive efforts in the development of new x-ray detectors. The spatial resolution of current scintillator layers is limited by optical diffusion. This limitation can be overcome by a pixelation, which prevents optical photons from crossing the interface between two neighboring pixels. In this work, an array of pores was etched in a silicon wafer with a pixel pitch of 50 microm. A very high aspect ratio was achieved with wall thicknesses of 4-7 microm and pore depths of about 400 microm. Subsequently, the pores were filled with Tl-doped cesium iodide (CsI:Tl) as a scintillator in a special process, which includes powder melting and solidification of the CsI. From the sample geometry and x-ray absorption measurement the pore fill grade was determined to be 75%. The scintillator-filled samples have a circular active area of 16 mm diameter. They are coupled with an optical sensor binned to the same pixel pitch in order to measure the x-ray imaging performance. The x-ray sensitivity, i.e., the light output per absorbed x-ray dose, is found to be only 2.5%-4.5% of a commercial CsI-layer of similar thickness, thus very low. The efficiency of the pores to transport the generated light to the photodiode is estimated to be in the best case 6.5%. The modulation transfer function is 40% at 4 lp/mm and 10%-20% at 8 lp/mm. It is limited most likely by the optical gap between scintillator and sensor and by K-escape quanta. The detective quantum efficiency (DQE) is determined at different beam qualities and dose settings. The maximum DQE(0) is 0.28, while the x-ray absorption with the given thickness and fill factor is 0.57. High Swank noise is suspected to be the reason, mainly caused by optical scatter inside the CsI-filled pores. The results are compared to Monte Carlo simulations of the photon transport inside the pore array structure. In addition, some x-ray images of technical and anatomical phantoms are shown. This work shows that scintillator-filled pore arrays can provide x-ray imaging with high spatial resolution, but are not suitable in their current state for most of the applications in medical imaging, where increasing the x-ray doses cannot be tolerated.     

Studies of x-ray energy absorption and quantum noise properties of x-ray screens by use of Monte Carlo simulation

The imaging properties of the phosphor layer in fluorescent screens or image intensifiers are related to its x-ray absorption characteristics. In this study, we applied Monte Carlo methods for the simulation of x-ray photon diffusion in a phosphor layer. The K-reabsorption factor, absorbed x-ray energy, quantum absorption efficiency, statistical factor, and noise-equivalent absorption were determined as a function of the incident energy and angle of the x rays for eight commonly used phosphor layers. These basic physical quantities will be useful for the prediction of the information transfer properties of a phosphor layer.     

Screen optics effects on detective quantum efficiency in digital radiography: zero-frequency effects

Indirect flat panel imagers have been developed for digital radiography, fluoroscopy and mammography, and are now in clinical use. Screens made from columnar structured cesium iodide (CsI) scintillators doped with thallium have been used extensively in these detectors. The purpose of this article is to investigate the effect of screen optics, e.g., light escape efficiency versus depth, on gain fluctuation noise, expressed as the Swank factor. Our goal is to obtain results useful in optimizing screens for digital radiography systems. Experimental measurements from structured CsI samples were used to derive their screen optics properties, and the same methods can also be applied to powder screens. CsI screens, all of the same thickness but with different optical designs and manufacturing techniques, were obtained from Hamamatsu Photonics Corporation. The pulse height spectra (PHS) of the screens were measured at different x-ray energies. A theoretical model was developed for the light escape efficiency and a method for deriving light escape efficiency versus depth from experimental PHS measurements was implemented and applied to the CsI screens. The results showed that the light escape efficiency varies essentially linearly as a function of depth in the CsI samples, and that the magnitude of variation is relatively small, leading to a high Swank factor.     

A comparison of the imaging properties of CCD-based devices used for small field digital mammography

The objective imaging characteristics of three systems that use charge coupled devices (CCD) for small-field digital mammography (SFDM) have been compared in terms of spatial resolution and signal to noise ratio. The results indicate that although they are designed for nominally the same tasks of stereotactic localization and spot imaging these detectors have significantly differing physical imaging properties. Imaging system design parameters such as the phosphor screen type and thickness, screen configuration and method of optically coupling the phosphor to the CCD have significant effects on the imaging performance of the detectors.