Search

Background: The spatial resolution of energy‐integrating diagnostic CT scanners is limited by interpixel reflectors on the detector, which optically isolate pixels but create dead space. Because the width of the reflector cannot easily be decreased, fill factor diminishes as resolution increases. Purpose: We propose loading (or mixing) a high‐Z element into the reflectors, causing the reflectors to be X‐ray fluorescent. Re‐emitted characteristic X‐rays could be detected in adjacent pixels, increasing the effective fill factor and compensating for fill factor loss with higher‐resolution detectors. The purpose of this work is to understand the physical principles of this approach and to analyze its effectiveness using Monte Carlo simulations. Methods: Detector pixels were modeled using the GEANT4 Monte Carlo package. The width of the reflector was kept constant at 0.1 mm throughout, and we considered pixel pitches between 0.5 and 1 mm. The pixelated scintillator material was gadolinium oxysulfide, 3 mm thick. The baseline reflector material was chosen to be acrylic, and varying concentrations of a high‐Z element were loaded into the material. We assumed that the optical characteristics of pixels were ideal (no absorption within pixels, perfect reflection at boundaries). The detector was irradiated uniformly with 10,000 X‐ray photons to estimate its spectral response. The figure of merit was the variance of the detector signal at zero frequency normalized to that of an ideal single‐bin photon‐counting detector with 100% fill factor. Sensitivity analyses were conducted to understand the effect of varying the high‐Z element concentration and the spectrum. Results: Initial simulations suggested that a k‐edge near 50 keV would be ideal. Gd was therefore selected as the high‐Z material. The relative variances for a conventional energy integrating detector without Gd at 1 mm pixel pitch (81% fill factor) and 0.5 mm pixel pitch (64% fill factor) were 1.38 and 1.74, compared to 1.00 for an ideal photon counting detector, implying a 26% variance penalty for 0.5 mm pitch. When 1 g/cm3 Gd was loaded into the interpixel reflector, the relative variance improved to 1.27 and 1.43, respectively, implying that the variance penalty for including Gd together with 0.5 mm pitch is only 4%. Performance was nearly maximized at 1.0 g/cm3 of Gd, but a concentration of 0.5 g/cm3 of Gd showed most of the benefit. Improvements depend weakly on kV, with lower kV associated with higher improvements. An external anti‐scatter grid was not modeled in our simulations and would reduce the expected benefit, depending greatly on the pitch and dimensionality of the anti‐scatter grid. Conclusions: The losses in fill factor associated with smaller pixel pitch can be reduced if Gd or a similar element could be loaded into the interpixel reflector. These improvements in noise efficiency are yet to be verified experimentally. [ABSTRACT FROM AUTHOR]

Background: A common rule of thumb for object detection is the Rose criterion, which states that a signal must be five standard deviations above background to be detectable to a human observer. The validity of the Rose criterion in CT imaging is limited due to the presence of correlated noise. Recent reconstruction and denoising methodologies are also able to restore apparent image quality in very noisy conditions, and the ultimate limits of these methodologies are not yet known. Purpose: To establish a lower bound on the minimum achievable signal‐to‐noise ratio (SNR) for object detection, below which detection performance is poor regardless of reconstruction or denoising methodology. Methods: We consider a numerical observer that operates on projection data and has perfect knowledge of the background and the objects to be detected, and determine the minimum projection SNR that is necessary to achieve predetermined lesion‐level sensitivity and case‐level specificity targets. We define a set of discrete signal objects O$\mathcal{O}$ that encompasses any lesion of interest and could include lesions of different sizes, shapes, and locations. The task is to determine which object of O$\mathcal{O}$ is present, or to state the null hypothesis that no object is present. We constrain each object in O$\mathcal{O}$ to have equivalent projection SNR and use Monte Carlo methods to calculate the required projection SNR necessary. Because our calculations are performed in projection space, they impose an upper limit on the performance possible from reconstructed images. We chose O$\mathcal{O}$ to be a collection of elliptical or circular low contrast metastases and simulated detection of these objects in a parallel beam system with Gaussian statistics. Unless otherwise stated, we assume a target of 80% lesion‐level sensitivity and 80% case‐level specificity and a search field of view that is 6 cm by 6 cm by 10 slices. Results: When O$\mathcal{O}$ contains only a single object, our problem is equivalent to two‐alternative forced choice (2AFC) and the required projection SNR is 1.7. When O$\mathcal{O}$ consists of circular 6‐mm lesions at different locations in space, the required projection SNR is 5.1. When O$\mathcal{O}$ is extended to include ellipses and circles of different sizes, the required projection SNR increases to 5.3. The required SNR increases if the sensitivity target, specificity target, or search field of view increases. Conclusions: Even with perfect knowledge of the background and target objects, the ideal observer still requires an SNR of approximately 5. This is a lower bound on the SNR that would be required in real conditions, where the background and target objects are not known perfectly. Algorithms that denoise lesions with less than 5 projection SNR, regardless of the denoising methodology, are expected to show vanishing effects or false positive lesions. [ABSTRACT FROM AUTHOR]

Purpose: Mobile lung tumors are increasingly being treated with ablative radiotherapy, for which precise motion management is essential. In‐room stereoscopic radiography systems are able to guide ablative radiotherapy for stationary cranial lesions but not optimally for lung tumors unless fiducial markers are inserted. We propose augmenting stereoscopic radiographic systems with multiple small x‐ray sources to provide the capability of imaging with stereoscopic, single frame tomosynthesis. Methods: In single frame tomosynthesis, nine x‐ray sources are placed in a 3 × 3 configuration and energized simultaneously. The beams from these sources are collimated so that they converge on the tumor and then diverge to illuminate nine non‐overlapping sectors on the detector. These nine sector images are averaged together and filtered to create the tomosynthesis effect. Single frame tomosynthesis is intended to be an alternative imaging mode for existing stereoscopic systems with a field of view that is three times smaller and a temporal resolution equal to the frame rate of the detector. We simulated stereoscopic tomosynthesis and radiography using Monte Carlo techniques on 60 patients with early‐stage lung cancer from the NSCLC‐Radiomics dataset. Two board‐certified radiation oncologists reviewed these simulated images and rated them on a 4‐point scale (1: tumor not visible; 2: tumor visible but inadequate for motion management; 3: tumor visible and adequate for motion management; 4: tumor visibility excellent). Each tumor was independently presented four times (two viewing angles from radiography and two viewing angles from tomosynthesis) in a blinded fashion over two reading sessions. Results: The fraction of tumors that were rated as adequate or excellent for motion management (scores 3 or 4) from at least one viewing angle was 53% using radiography and 90% using tomosynthesis. From both viewing angles, the corresponding fractions were 7% for radiography and 48% for tomosynthesis. Readers agreed exactly on 62% of images and within 1 point on 98% of images. The acquisition technique was estimated to be 75 mAs at 120 kVp per treatment fraction assuming one verification image per breath, approximately one order of magnitude less than a standard dose cone beam CT. Conclusions: Stereoscopic tomosynthesis may provide a noninvasive, low dose, intrafraction motion verification technique for lung tumors treated by ablative radiotherapy. The system architecture is compatible with real‐time video capture at 30 frames per second. Simulations suggest that most, but not all, lung tumors can be adequately visualized from at least one viewing angle. [ABSTRACT FROM AUTHOR]

Photon counting detectors (PCDs) are classically described as being either paralyzable or nonparalyzable. When the PCD is paralyzed, it is no longer sensitive to the detection of additional flux. A recent strategy in PCD design has been to compensate for detector paralysis by embedding specialized paralysis compensation electronics into the application-specific integrated circuit (ASIC). One such compensation mechanism is the pileup trigger, which places an additional energy bin at very high energy that is triggered only during pileup. Another compensation mechanism is the retrigger architecture, which converts a paralyzable PCD into a nonparalyzable PCD. We propose a third mechanism that modifies the retrigger architecture using dedicated secondary counters. We studied the incremental benefit of these three paralysis compensation mechanisms in simulation. We modeled the spectral response using Monte Carlo simulations and then estimated the variance in basis material decomposition of a single pixel using the Cramér-Rao lower bound (CRLB). In the absence of paralysis compensation, noise in basis material images shows sharp increases at moderate flux (near the characteristic count rate) due to contrast inversion and again at high flux. The pileup trigger reduces noise at high flux but does not eliminate contrast inversion. The retrigger architecture eliminates contrast inversion but does not reduce noise at high flux. Our proposed retrigger architecture with dedicated secondary counters reduce noise at both moderate and high flux. [ABSTRACT FROM AUTHOR]

The performance of X-ray photon counting detectors (PCDs), especially on spectral tasks, is compromised by charge sharing. Existing mechanisms to compensate for charge sharing, such as charge summing circuitry or larger pixel sizes, increase and aggravate pileup effects. We propose a new mechanism, the coincidence counting bin (CCB), which does not increase pileup and which has implementation similarities to existing energy bins. The CCB is triggered by coincident events in adjacent pixels and provides an estimate of the double counts arising from charge sharing. Unlike charge summing, the CCB does not directly restore corrupted events. Nonetheless, knowledge of the number of coincident counts can be used by the estimator to reduce noise. We simulated a PCD with and without the CCB using Monte Carlo simulations, modeling PCD pixels as instantaneous charge collectors and X-ray energy deposition as producing a Gaussian charge cloud with 75 micron FWHM, independent of energy. With typical operating conditions and at low flux (120 kVp, incident count rate 1% of characteristic count rate, 30 cm object thickness, five energy bins, pixel pitch of 300 microns), the CCB improved dose efficiency of iodine and water basis material decomposition by 70% and 50%, respectively. An improvement of 20% was also seen in an iodine CNR task. These improvements are attenuated as incident flux increases and show moderate dependence on filtration and pixel size. At high flux, the CCB does not provide useful information and is discarded by the estimator. The CCB may be an effective and practical mechanism for charge sharing compensation in PCDs. [ABSTRACT FROM AUTHOR]

Charge sharing, scatter, and fluorescence events in a photon counting detector can result in counting of a single incident photon in multiple neighboring pixels, each at a fraction of the true energy. This causes energy distortion and correlation of data across energy bins in neighboring pixels (spatio-energy correlation), with the severity depending on the detector pixel size and detector material. If a “macro-pixel” is formed by combining the counts from multiple adjacent small pixels, it will exhibit correlations across its energy bins. Understanding these effects can be crucial for detector design and for model-based imaging applications. This paper investigates the impact of these effects in basis material and effective monoenergetic estimates using the Cramér-Rao Lower Bound. To do so, we derive a correlation model for the multi-counting events. CdTe detectors with grids of pixels with side length of $250~\mu \text{m}$ , $500~\mu \text{m}$ , and 1 mm were compared, with binning of $4\times4$ , $2\times2$ , and $1\times1$ pixels, respectively, to keep the same net 1 mm2 aperture constant. The same flux was applied to each. The mean and covariance matrix of measured photon counts were derived analytically using spatio-energy response functions precomputed from Monte Carlo simulations. Our results show that a 1 mm2 macro-pixel with $250\times 250\,\,\mu \text{m}^{\textsf {2}}$ sub-pixels shows 35% higher standard deviation than a single 1 mm2 pixel for material-specific imaging, while the penalty for effective monoenergetic imaging is <10% compared with a single 1 mm $^{\textsf {2}}$ pixel. Potential benefits of sub-pixels (higher spatial resolution and lower pulse pile-up effects) are important but were not investigated here. [ABSTRACT FROM AUTHOR]

Purpose: Photon‐counting detectors using CdTe or CZT substrates are promising candidates for future CT systems but suffer from a number of nonidealities, including charge sharing and pulse pileup. By increasing the pixel size of the detector, the system can improve charge sharing characteristics at the expense of increasing pileup. The purpose of this work is to describe these considerations in the optimization of the detector pixel pitch. Methods: The transport of x rays through the CdTe substrate was simulated in a Monte Carlo fashion using GEANT4. Deposited energy was converted into charges distributed as a Gaussian function with size dependent on interaction depth to capture spreading from diffusion and Coulomb repulsion. The charges were then collected in a pixelated fashion. Pulse pileup was incorporated separately with Monte Carlo simulation. The Cramér–Rao lower bound (CRLB) of the measurement variance was numerically estimated for the basis material projections. Noise in these estimates was propagated into CT images. We simulated pixel pitches of 250, 350, and 450 microns and compared the results to a photon counting detector with pileup but otherwise ideal energy response and an ideal dual‐energy system (80/140 kVp with tin filtration). The modeled CdTe thickness was 2 mm, the incident spectrum was 140 kVp and 500 mA, and the effective dead time was 67 ns. Charge summing circuitry was not modeled. We restricted our simulations to objects of uniform thickness and did not consider the potential advantage of smaller pixels at high spatial frequencies. Results: At very high x‐ray flux, pulse pileup dominates and small pixel sizes perform best. At low flux or for thick objects, charge sharing dominates and large pixel sizes perform best. At low flux and depending on the beam hardness, the CRLB of variance in basis material projections tasks can be 32%–55% higher with a 250 micron pixel pitch compared to a 450 micron pixel pitch. However, both are about four times worse in variance than the ideal photon counting detector. The optimal pixel size depends on a number of factors such as x‐ray technique and object size. At high technique (140 kVp/500 mA), the ratio of variance for a 450 micron pixel compared to a 250 micron pixel size is 2126%, 200%, 97%, and 78% when imaging 10, 15, 20, and 25 cm of water, respectively. If 300 mg/cm2 of iodine is also added to the object, the variance ratio is 117%, 91%, 74%, and 72%, respectively. Nonspectral tasks, such as equivalent monoenergetic imaging, are less sensitive to spectral distortion. Conclusions: The detector pixel size is an important design consideration in CdTe detectors. Smaller pixels allow for improved capabilities at high flux but increase charge sharing, which in turn compromises spectral performance. The optimal pixel size will depend on the specific task and on the charge shaping time. [ABSTRACT FROM AUTHOR]

Purpose: The authors recently proposed a dynamic, prepatient x-ray attenuator capable of producing a piecewise-linear attenuation profile customized to each patient and viewing angle. This attenuator was intended to reduce scatter-to-primary ratio (SPR), dynamic range, and dose by redistributing flux. In this work the authors tested the ability of the attenuator to reduce dose and SPR in simulations. Methods: The authors selected four clinical applications, including routine full field-of-view scans of the thorax and abdomen, and targeted reconstruction tasks for an abdominal aortic aneurysm and the pancreas. Raw data were estimated by forward projection of the image volume datasets. The dynamic attenuator was controlled to reduce dose while maintaining peak variance by solving a convex optimization problem, assuming a priori knowledge of the patient anatomy. In targeted reconstruction tasks, the noise in specific regions was given increased weighting. A system with a standard attenuator (or 'bowtie filter') was used as a reference, and used either convex optimized tube current modulation (TCM) or a standard TCM heuristic. The noise of the scan was determined analytically while the dose was estimated using Monte Carlo simulations. Scatter was also estimated using Monte Carlo simulations. The sensitivity of the dynamic attenuator to patient centering was also examined by shifting the abdomen in 2 cm intervals. Results: Compared to a reference system with optimized TCM, use of the dynamic attenuator reduced dose by about 30% in routine scans and 50% in targeted scans. Compared to the TCM heuristics which are typically used without a priori knowledge, the dose reduction is about 50% for routine scans. The dynamic attenuator gives the ability to redistribute noise and variance and produces more uniform noise profiles than systems with a conventional bowtie filter. The SPR was also modestly reduced by 10% in the thorax and 24% in the abdomen. Imaging with the dynamic attenuator was relatively insensitive to patient centering, showing a 17% increase in peak variance for a 6 cm shift of the abdomen, instead of an 82% increase in peak variance for a fixed bowtie filter. Conclusions: A dynamic prepatient x-ray attenuator consisting of multiple wedges is capable of achieving substantial dose reductions and modest SPR reductions. [ABSTRACT FROM AUTHOR]