Scatter correction for cone-beam CT in radiation therapy

Department of Radiation Oncology, Stanford University, Stanford, California 94305, USA.
Medical Physics (Impact Factor: 3.01). 07/2009; 36(6):2258-68. DOI: 10.1118/1.3130047
Source: PubMed

ABSTRACT Cone-beam CT (CBCT) is being increasingly used in modern radiation therapy for patient setup and adaptive replanning. However, due to the large volume of x-ray illumination, scatter becomes a rather serious problem and is considered as one of the fundamental limitations of CBCT image quality. Many scatter correction algorithms have been proposed in literature, while a standard practical solution still remains elusive. In radiation therapy, the same patient is scanned repetitively during a course of treatment, a natural question to ask is whether one can obtain the scatter distribution on the first day of treatment and then use the data for scatter correction in the subsequent scans on different days. To realize this scatter removal scheme, two technical pieces must be in place: (i) A strategy to obtain the scatter distribution in on-board CBCT imaging and (ii) a method to spatially match a prior scatter distribution with the on-treatment CBCT projection data for scatter subtraction. In this work, simple solutions to the two problems are provided. A partially blocked CBCT is used to extract the scatter distribution. The x-ray beam blocker has a strip pattern, such that partial volume can still be accurately reconstructed and the whole-field scatter distribution can be estimated from the detected signals in the shadow regions using interpolation/extrapolation. In the subsequent scans, the patient transformation is determined using a rigid registration of the conventional CBCT and the prior partial CBCT. From the derived patient transformation, the measured scatter is then modified to adapt the new on-treatment patient geometry for scatter correction. The proposed method is evaluated using physical experiments on a clinical CBCT system. On the Catphan 600 phantom, the errors in Hounsfield unit (HU) in the selected regions of interest are reduced from about 350 to below 50 HU; on an anthropomorphic phantom, the error is reduced from 15.7% to 5.4%. The proposed method is attractive in applications where a high CBCT image quality is critical, for example, dose calculation in adaptive radiation therapy.

  • [Show abstract] [Hide abstract]
    ABSTRACT: A beam stop array (BSA) is widely used to estimate the scatter distribution of cone-beam computed tomography (CT) experimentally. A rectangular BSA, which is the most generally used, produces a penumbra region at the edge of the grid on the projection image. The penumbra region can be the cause of an inaccurate scatter estimate and loss of image data. In this study, two types of BSAs have been designed and developed, a linear BSA (l-BSA), which is a simple rectangular type, and a curved BSA (c-BSA) to avoid the penumbra effect. A Monte Carlo simulation was performed to determine the thickness and the material of beam stop arrays. The particle tracking technique of MCNP5 code was used to analyze the accuracy of the scatter distribution acquired by the beam stop arrays. The improvement in the image quality was analyzed by using an in-house phantom, a simple type CT phantom, and a humanoid phantom. How the curved beam stop array improved the penumbra effect was also analyzed in a large field of view. The additional dose caused by a dual scan was analyzed by using a Monte Carlo simulation and experiment. The accuracies of the scatter estimates were 83.4% for the c-BSA and 90.4% for the l-BSA. The c-BSA compensated the penumbra effect effectively while the l-BSA had a severe penumbra region in the partial projection image in a large field of view. Contrasts were improved by 24.8 (air-PMMA), 55.8 (Teflon-PMMA), and 81.7% (water-PMMA) for the c-BSA and 11.1 (air-PMMA), 44.1 (Teflon-PMMA), and 82.2% (water-PMMA) for the l-BSA by scatter correction. After noise suppression, the contrast-to-noise ratios were improved by 2.7–4.1 times for the c-BSA and 3.5–5.3 times for the l-BSA. Uniformities were also improved by up to 10 times for the c-BSA and 12 times for the l-BSA. The quality of the humanoid phantom image was also improved after the scatter correction. Consequently, the c-BSA can improve the image quality of cone-beam CT and compensate for the penumbra effect without an additional dose of radiation.
    Journal- Korean Physical Society 04/2014; 64(8):1220-1229. DOI:10.3938/jkps.64.1220 · 0.43 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Abstract Objective: To evaluate the impact of tube voltage, tube current, pulse number, and magnification factor on the image quality of a novel experimental set-up and the corresponding radiation. Materials and Methods: Six human temporal bones with cochlear implant were imaged using various tube voltages, tube currents, pulse numbers, and magnification. The effect of radiation was evaluated using a metaloxide semiconductor field-effect transistor (MOSFET) dosimeter device on an anthropomorphic RANDO RAN102 male head phantom. A copper and aluminum combination filter was used for hardware filtration. Results: Overall, 900 frames, 11 mA, and 88 kV provided the best image quality. In temporal bones imaged with the optimized parameters, the cochlea, osseous spiral lamina, modiolus, stapes, round window niche, and oval window landmarks were demonstrated with anatomic structures still fully assessable in all parts and acceptable image quality. The most dominant contributor to the effective dose was bone marrow (36%-37 %) followed by brain (34%-36%), remainder tissues (12%), extra-thoracic airways (7%), and oral mucosa (5%). Conclusions: By increasing the number of frames, the image quality of the inner ear details obtained using the novel cone-beam computed tomography improved.
    The Annals of otology, rhinology, and laryngology 12/2014; DOI:10.1177/0003489414565004 · 1.05 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: CT is the frontline imaging modality for diagnosis of acute traumatic brain injury (TBI), involving the detection of fresh blood in the brain (contrast of 30–50 HU, detail size down to 1 mm) in a non-contrast-enhanced exam. A dedicated point-of-care imaging system based on cone-beam CT (CBCT) could benefit early detection of TBI and improve direction to appropriate therapy. However, flat-panel detector (FPD) CBCT is challenged by artifacts that degrade contrast resolution and limit application in soft-tissue imaging. We present and evaluate a fairly comprehensive framework for artifact correction to enable soft-tissue brain imaging with FPD CBCT. The framework includes a fast Monte Carlo (MC)-based scatter estimation method complemented by corrections for detector lag, veiling glare, and beam hardening.The fast MC scatter estimation combines GPU acceleration, variance reduction, and simulation with a low number of photon histories and reduced number of projection angles (sparse MC) augmented by kernel de-noising to yield a runtime of ~4 min per scan. Scatter correction is combined with two-pass beam hardening correction. Detector lag correction is based on temporal deconvolution of the measured lag response function. The effects of detector veiling glare are reduced by deconvolution of the glare response function representing the long range tails of the detector point-spread function. The performance of the correction framework is quantified in experiments using a realistic head phantom on a testbench for FPD CBCT.Uncorrected reconstructions were non-diagnostic for soft-tissue imaging tasks in the brain. After processing with the artifact correction framework, image uniformity was substantially improved, and artifacts were reduced to a level that enabled visualization of ~3 mm simulated bleeds throughout the brain. Non-uniformity (cupping) was reduced by a factor of 5, and contrast of simulated bleeds was improved from ~7 to 49.7 HU, in good agreement with the nominal blood contrast of 50 HU. Although noise was amplified by the corrections, the contrast-to-noise ratio (CNR) of simulated bleeds was improved by nearly a factor of 3.5 (CNR = 0.54 without corrections and 1.91 after correction). The resulting image quality motivates further development and translation of the FPD-CBCT system for imaging of acute TBI.
    Physics in Medicine and Biology 02/2015; 60(4). DOI:10.1088/0031-9155/60/4/1415 · 2.92 Impact Factor

Full-text (2 Sources)

Available from
Aug 22, 2014