Xinhui Duan

Dose Reduction to Anterior Surfaces With Organ-Based Tube-Current Modulation: Evaluation of Performance in a Phantom Study

OBJECTIVE The purpose of this study was to evaluate in phantoms the dose reduction to anterior surfaces and image quality with organ-based tube-current modulation in head and thoracic CT. MATERIALS AND METHODS Organ-based tube-current modulation is designed to reduce radiation dose to superficial radiosensitive organs, such as the lens of the eye, thyroid, and breast, by decreasing the tube current when the tube passes closest to these organs. Dose and image quality were evaluated in phantoms for clinical head and thorax examination protocols with and without organ-based tube-current modulation. Surface dose reduction as a function of position was measured using a 32-cm CT dose index (CTDI) phantom, an anthropomorphic adult phantom, and ion chambers. Surface dose reduction as a function of patient size was investigated using three semianthropomorphic phantoms with posteroanterior dimensions of 14, 25, and 31 cm. Image noise (the SD of CT numbers in regions of interest) was evaluated for the anthropomorphic and the semianthropomorphic phantoms. RESULTS For equivalent scanner output (volume CTDI), the dose to the midline of the anterior surface was reduced by 27-50%, depending on the anatomic region (head or thorax) and phantom size, and the dose to the posterior surface was correspondingly increased. Image noise was not significantly different between scans with and without organ-based tube-current modulation (p = 0.85). CONCLUSION Organ-based tube-current modulation can reduce the dose to the anterior surface of patients without increasing image noise by commensurately increasing the dose to the posterior surface. This technique can reduce the dose to anterior radiosensitive organs for head and thoracic CT scans.

Radiation dose reduction to the breast in thoracic CT: Comparison of bismuth shielding, organ‐based tube current modulation, and use of a globally decreased tube current

PURPOSE The purpose of this work was to evaluate dose performance and image quality in thoracic CT using three techniques to reduce dose to the breast: bismuth shielding, organ-based tube current modulation (TCM) and global tube current reduction. METHODS Semi-anthropomorphic thorax phantoms of four different sizes (15, 30, 35, and 40 cm lateral width) were used for dose measurement and image quality assessment. Four scans were performed on each phantom using 100 or 120 kV with a clinical CT scanner: (1) reference scan; (2) scan with bismuth breast shield of an appropriate thickness; (3) scan with organ-based TCM; and (4) scan with a global reduction in tube current chosen to match the dose reduction from bismuth shielding. Dose to the breast was measured with an ion chamber on the surface of the phantom. Image quality was evaluated by measuring the mean and standard deviation of CT numbers within the lung and heart regions. RESULTS Compared to the reference scan, dose to the breast region was decreased by about 21% for the 15-cm phantom with a pediatric (2-ply) shield and by about 37% for the 30, 35, and 40-cm phantoms with adult (4-ply) shields. Organ-based TCM decreased the dose by 12% for the 15-cm phantom, and 34-39% for the 30, 35, and 40-cm phantoms. Global lowering of the tube current reduced breast dose by 23% for the 15-cm phantom and 39% for the 30, 35, and 40-cm phantoms. In phantoms of all four sizes, image noise was increased in both the lung and heart regions with bismuth shielding. No significant increase in noise was observed with organ-based TCM. Decreasing tube current globally led to similar noise increases as bismuth shielding. Streak and beam hardening artifacts, and a resulting artifactual increase in CT numbers, were observed for scans with bismuth shields, but not for organ-based TCM or global tube current reduction. CONCLUSIONS Organ-based TCM produces dose reduction to the breast similar to that achieved with bismuth shielding for both pediatric and adult phantoms. However, organ-based TCM does not affect image noise or CT number accuracy, both of which are adversely affected by bismuth shielding. Alternatively, globally decreasing the tube current can produce the same dose reduction to the breast as bismuth shielding, with a similar noise increase, yet without the streak artifacts and CT number errors caused by the bismuth shields. Moreover, globally decreasing the tube current reduces the dose to all tissues scanned, not simply to the breast.

Bismuth Shielding, Organ-based Tube Current Modulation, and Global Reduction of Tube Current for Dose Reduction to the Eye at Head CT

PURPOSE To compare the dose and image quality of three methods for reducing the radiation dose to the eye at head computed tomography (CT): bismuth shielding, organ-based tube current modulation (TCM), and global reduction of the tube current. MATERIALS AND METHODS An anthropomorphic head phantom was scanned under six conditions: (a) without any dose reduction techniques (reference scanning); (b) with one bismuth eye shield; (c) with organ-based TCM; (d) with reduced tube current to yield the same dose reduction as one bismuth shield; (e) with two layers of bismuth shields; and (f) with organ-based TCM and one bismuth shield. Dose to the eye, image noise, and CT numbers in the brain region were measured and compared. The effect of increasing distance between the bismuth shield and eye lens was also investigated. RESULTS Relative to the reference scan, the dose to the eye was reduced by 26.4% with one bismuth shield, 30.4% with organ-based TCM, and 30.2% with a global reduction in tube current. A combination of organ-based TCM with one bismuth shield reduced the dose by 47.0%. Image noise in the brain region was slightly increased for all dose reduction methods. CT numbers were increased whenever the bismuth shield was used. Increasing the distance between the bismuth shield and the eye lens helped reduce CT number errors, but the increase in noise remained. CONCLUSION Organ-based TCM provided superior image quality to that with bismuth shielding while similarly reducing dose to the eye. Simply reducing tube current globally by about 30% provides the same dose reduction to the eye as bismuth shielding; however, CT number accuracy is maintained and dose is reduced to all parts of the head.

CT scanner x‐ray spectrum estimation from transmission measurements

PURPOSE In diagnostic CT imaging, multiple important applications depend on the knowledge of the x-ray spectrum, including Monte Carlo dose calculations and dual-energy material decomposition analysis. Due to the high photon flux involved, it is difficult to directly measure spectra from the x-ray tube of a CT scanner. One potential method for indirect measurement involves estimating the spectrum from transmission measurements. The expectation maximization (EM) method is an accurate and robust method to solve this problem. In this article, this method was evaluated in a commercial CT scanner. METHODS Two step-wedges (polycarbonate and aluminum) were used to produce different attenuation levels. Transmission measurements were performed on the scanner and the measured data from the scanner were exported to an external computer to calculate the spectra. The EM method was applied to solve the equations that represent the attenuation processes of polychromatic x-ray photons. Estimated spectra were compared to the spectra simulated using a software provided by the manufacturer of the scanner. To test the accuracy of the spectra, a verification experiment was performed using a phantom containing different depths of water. The measured transmission data were compared to the transmission values calculated using the estimated spectra. RESULTS Spectra of 80, 100, 120, and 140 kVp from a dual-source CT scanner were estimated. The estimated and simulated spectra were well matched. The differences of mean energies were less than 1 keV. In the verification experiment, the measured and calculated transmission values were in excellent agreement. CONCLUSIONS Spectrum estimation using transmission data and the EM method is a quantitatively accurate and robust technique to estimate the spectrum of a CT system. This method could benefit studies relying on accurate knowledge of the x-ray spectra from CT scanner.

Electronic Noise in CT Detectors: Impact on Image Noise and Artifacts

OBJECTIVE The objective of our study was to evaluate in phantoms the differences in CT image noise and artifact level between two types of commercial CT detectors: one with distributed electronics (conventional) and one with integrated electronics intended to decrease system electronic noise. MATERIALS AND METHODS Cylindric water phantoms of 20, 30, and 40 cm in diameter were scanned using two CT scanners, one equipped with integrated detector electronics and one with distributed detector electronics. All other scanning parameters were identical. Scans were acquired at four tube potentials and 10 tube currents. Semianthropomorphic phantoms were scanned to mimic the shoulder and abdominal regions. Images of two patients were also selected to show the clinical values of the integrated detector. RESULTS Reduction of image noise with the integrated detector depended on phantom size, tube potential, and tube current. Scans that had low detected signal had the greatest reductions in noise, up to 40% for a 30-cm phantom scanned using 80 kV. This noise reduction translated into up to 50% in dose reduction to achieve equivalent image noise. Streak artifacts through regions of high attenuation were reduced by up to 45% on scans obtained using the integrated detector. Patient images also showed superior image quality for the integrated detector. CONCLUSION For the same applied radiation level, the use of integrated electronics in a CT detector showed a substantially reduced level of electronic noise, resulting in reductions in image noise and artifacts, compared with detectors having distributed electronics.

Metal artifact reduction in CT images by sinogram TV inpainting

In this paper, a total variation (TV) inpainting method is proposed for metal artifact reduction in medical computed tomography (CT). Digital inpainting is an image processing method to fill in the lost image information in a consistent way. In our work, projection data with metal projection region (MPR) are treated as a damaged image, and TV inpainting is applied to “inpaint” the information missing region. Compared to conventional interpolation methods, the advantage of our algorithm lies in dealing with complicate cases such as an image with multiple metal objects. In numerical experiments, both TV inpainting and linear interpolation method are performed on noise-free and additive noisy projection of a modified Shepp-Logan phantom. Results show that the algorithm proposed fills metal projection gaps more smoothly and accurately than linear interpolation, hence produces images of superior quality after reconstruction. Relevant practical issues including the limitation of the algorithm and possible improvements for future work are discussed.

Attenuation‐based estimation of patient size for the purpose of size specific dose estimation in CT. Part II. Implementation on abdomen and thorax phantoms using cross sectional CT images and scanned projection radiograph images

PURPOSE To estimate attenuation using cross sectional CT images and scanned projection radiograph (SPR) images in a series of thorax and abdomen phantoms. METHODS Attenuation was quantified in terms of a water cylinder with cross sectional area of A(w) from both the CT and SPR images of abdomen and thorax phantoms, where A(w) is the area of a water cylinder that would absorb the same dose as the specified phantom. SPR and axial CT images were acquired using a dual-source CT scanner operated at 120 kV in single-source mode. To use the SPR image for estimating A(w), the pixel values of a SPR image were calibrated to physical water attenuation using a series of water phantoms. A(w) and the corresponding diameter D(w) were calculated using the derived attenuation-based methods (from either CT or SPR image). A(w) was also calculated using only geometrical dimensions of the phantoms (anterior-posterior and lateral dimensions or cross sectional area). RESULTS For abdomen phantoms, the geometry-based and attenuation-based methods gave similar results for D(w). Using only geometric parameters, an overestimation of D(w) ranging from 4.3% to 21.5% was found for thorax phantoms. Results for D(w) using the CT image and SPR based methods agreed with each other within 4% on average in both thorax and abdomen phantoms. CONCLUSIONS Either the cross sectional CT or SPR images can be used to estimate patient attenuation in CT. Both are more accurate than use of only geometrical information for the task of quantifying patient attenuation. The SPR based method requires calibration of SPR pixel values to physical water attenuation and this calibration would be best performed by the scanner manufacturer.

Attenuation-based estimation of patient size for the purpose of size specific dose estimation in CT. Part I. Development and validation of methods using the CT image

PURPOSE For the purpose of size-specific dose estimation, information regarding patient attenuation is required. The purpose of this work is to describe a method for measuring patient attenuation and expressing the results in terms of a water cylinder, with cross sectional area A(w), which would absorb the same average dose as the irradiated patient. The ability to calculate A(w) directly from the CT image was validated with Monte Carlo simulations and an analytical model. METHODS A series of virtual cylinders were created with diameters ranging from 10 to 40 cm and lengths of 40 cm. The cylinders were given an atomic number equal to that of water; the density of the cylinders was varied from 0.26 to 1.2 g∕cm(3). The average dose to the cylinders from an axial scan at the longitudinal center position was calculated using Monte Carlo simulation and an analytical model. The relationship between phantom cross sectional area and calculated dose was determined for each density value to determine the dependence of A(w) on object attenuation. In addition, A(w) was estimated from the virtual CT images based on two derived models expressing the potential dependence of A(w) on object attenuation, one model assuming a linear dependence and the other assuming a quadratic dependence. Model results were compared with those from the Monte Carlo simulation and the analytical dose calculation approach. Virtual thorax and abdomen phantoms of adult and pediatric sizes were created, and A(w) was estimated using geometrical size parameters or the derived models. The accuracy of each approach for estimating A(w) was determined by comparing the average dose to the virtual phantom calculated using Monte Carlo simulation to the average dose to a water equivalent phantom of cross sectional area A(w). RESULTS In the absence of a bowtie filter, both the Monte Carlo simulation and analytical model showed that (A(w)∕A) had a quadratic dependence on (μ∕μ(w)). However, including a bowtie filter in the Monte Carlo simulation altered the relationship, such that A(w)∕A was linearly dependent on μ∕μ(w). Using this relationship, the dose absorbed by a water cylinder of area A(w) agreed with the dose absorbed by adult and pediatric, thorax and abdomen phantoms to within 6% (mean difference = 0.5 ± 4.8%). Estimates of A(w) (or the water equivalent diameter D(w)) using only anterior-posterior and lateral phantom dimensions led to dose estimates that agreed with Monte Carlo-derived dose values within 3% and 6% for the abdomen adult and pediatric phantoms, respectively. However, because of density differences between lung and tissue, larger differences in dose relative to Monte Carlo-derived values were observed in the thorax adult and pediatric phantoms (15% and 11%, respectively) when only geometrical parameters were used to estimate D(w). CONCLUSIONS Patient attenuation can be quantified in terms of the diameter of a water cylinder that absorbs same average dose as the irradiated cross section of the patient. The linear dependence of A(w) on object attenuation makes it straightforward to calculate A(w) from a CT image on most operator consoles or clinical workstations.

Few-View Projection Reconstruction With an Iterative Reconstruction-Reprojection Algorithm and TV Constraint

In applications of tomographic imaging, insufficient data problems can take various forms, such as few-view projection imaging which enables rapid scanning with lower X-ray dose. In this work, an iterative reconstruction-reprojection (IRR) algorithm with total variation (TV) constraint is developed for few-view projections. The IRR algorithm is used to estimate the missing projection data by iterative extrapolation between projection and image space. TV minimization is a popular image restoration method with edge preserving. In recent studies, it has been successfully used for reconstructing images from sparse samplings, such as few-view projections. Our method is derived from this work. The combination of IRR and TV achieves both estimation in projection space and regularization in image space, which accelerates the convergence of the iterations. To improve the quality of the image reconstructed from few-view fan-beam projections, a short-scan type IRR is also approached to reduce the redundancy of projection data. An improved weighting function is proposed for few-view short-scan projection reconstruction by the filtered backprojection (FBP) algorithm. Numerical simulations show that the IRR-TV algorithm is effective for the few-view problem of reconstructing sparse-gradient images.

Size-specific Dose Estimates for Chest, Abdominal, and Pelvic CT: Effect of Intrapatient Variability in Water-equivalent Diameter

PURPOSE To develop software to automatically calculate size-specific dose estimates (SSDEs) and to assess the effect of variations in water-equivalent diameter (Dw) along the z-axis on SSDE for computed tomographic (CT) examinations of the torso. MATERIALS AND METHODS In this institutional review board-approved, HIPAA-compliant, retrospective study, a software program was used to calculate Dw at each image position in 102 consecutive CT examinations of the combined chest, abdomen, and pelvis. SSDE was calculated by multiplying the size-dependent conversion factor and volume CT dose index (CTDIvol) at each image position. The variations in Dw along the z-axis were determined for six hypothetical scanning ranges: chest alone; abdomen alone; pelvis alone; chest and abdomen; abdomen and pelvis; and chest, abdomen, and pelvis. Mean SSDE was calculated in two ways: (a) from the SSDE at each position and (b) from the mean CTDIvol over each scan range and the conversion factor corresponding to Dw at the middle of the scan range. Linear regression analysis was performed to determine the correlation between SSDE values calculated in these two ways. RESULTS Across patients, for scan ranges 1-6, the mean of the difference between maximal and minimal Dw within a given patient was 5.2, 4.9, 2.5, 6.0, 5.6, and 6.5 cm, respectively. The mean SSDE values calculated by using the two methods were in close agreement, with root mean square differences of 0.9, 0.5, 0.5, 1.4, 1.0, and 1.1 mGy or 6%, 3%, 2%, 9%, 4%, and 6%, for the scan ranges of chest; abdomen; pelvis; chest and abdomen; abdomen and pelvis; and chest, abdomen, and pelvis, respectively. CONCLUSION Using the mean CTDIvol from the whole scan range and Dw from the image at the center of the scan range provided an easily obtained estimate of SSDE for the whole scan range that agreed well with values from an image-by-image approach, with a root mean square difference less than 1.4 mGy (9%).