Index of content:
Volume 35, Issue 6, June 2008
- Imaging Scientific Session: Room 332
- Cone‐Beam CT and Mammography
TH‐C‐332‐01: Low Dose Flat‐Panel Cone‐Beam CT and Tomosynthesis for Interventional Guidance Via Prior Image Constrained Compressed Sensing (PICCS)35(2008); http://dx.doi.org/10.1118/1.2962866View Description Hide Description
Purpose: Intra‐operative three dimensional imaging via FP‐CBCT (flat‐panel cone‐beam computed tomography) offers high spatial resolution and the ability to image low contrast structures such as soft‐tissue. However, repeated FP‐CBCT scanning requires irradiation of the patient for each subsequent acquisition. The aim of this work is to significantly reduce the delivered dose for monitoring and verification scanning, by using a planning scan to constrain the reconstruction.Method and Materials: An extension to compressed sensing (CS) has recently been proposed wherein a prior image is utilized as a constraint in the image reconstruction procedure (i.e. Prior Image Constrained Compressed Sensing — PICCS). The PICCS framework was used here on a clinical neuro‐interventional C‐arm system. Geometric calibration was incorporated into the algorithm through measured projection matrices of a known phantom. Results: The clinical task of placing a needle into a target was simulated with phantom and cadaver experiments. A full range of tomographic angles was explored from 10 degrees to 200 degrees. In the phantom experiments the delivered dose used for the reconstruction was twenty times less than the planning image. In the cadaver experiments PICCS images were reconstructed with a dose reduction factor of ten. The needle was well visualized in the tomosynthetic planes (e.g. coronal planes for a PA arc) for all acquisitions and the PICCS algorithm enabled improved reconstruction of the background anatomy compared with CS and standard FBP. Conclusion: The PICCS algorithm enabled reconstruction of the needle introduced from few views, while maintaining the background anatomy. If a tomosynthesis acquisition is chosen the orientation of the arc with respect to the object of interest is no less important than with standard image reconstruction. Experimental data indicates that a significant undersampling is acceptable (e.g. an order of magnitude dose savings) when using the PICCS reconstruction algorithm.
TH‐C‐332‐02: First Implementation of High‐Resolution Dual‐Detector Region‐Of‐Interest Cone‐Beam Computed Tomography (ROI‐CBCT) for a Rotating C‐Arm Gantry System35(2008); http://dx.doi.org/10.1118/1.2962867View Description Hide Description
Purpose: Region‐of‐interest cone‐beam computed tomography (ROI‐CBCT) has the potential to reduce integral dose while providing higher‐resolution data in the ROI. Previous work used rotating object geometries. We have implemented an ROI‐CBCT system on a clinical rotating C‐arm gantry system and have developed methods for combining high‐resolution projection data within the ROI and low‐resolution, low‐dose data outside the ROI. Method and Materials: High‐resolution ROI projection data were acquired of a coronary stent placed in a rabbit using a custom‐made, high‐sensitivity, microangiographic‐fluoroscope (HSMAF) detector (35 micron pixels), attached to the C‐arm of a clinical fluoroscopic gantry. Full field‐of‐view (FFOV), low‐resolution data were acquired at the same dose as the HSMAF (standard‐dose) and at a lower dose using a commercial flat‐panel detector (FPD) (194 micron pixels). HSMAF data were spatially registered with FPD data using cross‐correlation techniques. The lower‐dose FPD pixel values were matched to the HSMAF values using linear regression. During reconstruction, HSMAF data were used within the ROI, while corrected lower‐dose FPD data were used outside the ROI. Reconstructions were performed with HSMAF/lower‐dose‐FPD and standard‐dose‐FPD data using geometric calibration data, Parker weights, a Shepp‐Logan filter, and a Feldkamp algorithm, generating 512 × 512 × 512 volumes (25 micron voxels). Results: The dual‐detector ROI‐CBCT reconstruction exhibits greater detail than that of the FPD alone. The full‐width‐at‐half‐maxima of line profiles of stent struts (100 micron diameter) are approximately 150 and 290 microns for the dual‐detector and FPD data, respectively. No truncation artifacts are visible. The integral dose of the dual‐detector system is 54% that of the standard‐dose FPD for this experiment. Conclusion: High‐resolution 3D images of stents implanted in a rabbit are successfully obtained using a new rotating‐gantry dual‐detector ROI‐CBCT system. Resolution is significantly higher and integral dose is lower with ROI‐CBCT compared to the standard FPD approach.
(Support: NIH‐Grants R01‐NS43924, R01‐EB002873, Toshiba‐Medical‐Systems‐Corporation).
35(2008); http://dx.doi.org/10.1118/1.2962868View Description Hide Description
Purpose: To investigate the noise power properties of a cone‐beam breast CT (bCT) system.Method and Materials: Polyethylene cylinders were scanned under different acquisition conditions. Normalized noise power spectra (NNPS) were calculated from difference images by subtracting two identical scans. Multi‐dimensional NNPS were used to evaluate the noise properties of the bCT under different acquisition and reconstruction parameters including kVp, mA, number of projections, cone angle, object size, interpolation method, reconstruction filter, field of view (FOV), matrix size, and slice thickness. Results: Findings from the analysis: The bCT NNPS is rotational symmetric within the coronal plane and its shape is determined by the interpolation method and filter. For a cone angle range from 0° to 14°, the shape of the NPS curve changes slightly. The image variance increases slightly with increasing cone angle. Noise aliasing can be avoided if the sampling frequency of CTimage is beyond a specific threshold. The coronal plane noise power decreases with increasing slice thickness. If the radiation dose (mA) is below a specific threshold, the electronic noise from the detector plays a dominate role and the system is no longer quantum limited. The threshold value decreases with increasing object size. When the radiation dose is beyond this threshold, the system is quantum limited. The image variance has a power law relationship of approximately −1.1 with the total dose. If the dose is increased by linearly increasing the number of projections but not the tube current, the noise power has an inverse linear relationship with the dose. When the dose level is kept constant, the imagenoise also depends on the x‐ray spectrum and the object size. Conclusion: A thorough investigation of the noise power was performed. Quantitative results provide guidance for the bCT system operation, optimization and data reconstruction.
TH‐C‐332‐04: Comparison of Residual Motion Artifacts in Phase‐ and Amplitude‐Binned Onboard 4D‐CBCT Imaging35(2008); http://dx.doi.org/10.1118/1.2962870View Description Hide Description
Purpose: Linac‐mounted 4D‐Cone Beam CT(CBCT)imaging is an important tool for extending 3D IGRT techniques to respiration‐motion synchronized treatments. Our purpose is to evaluate the effects of two different projection binning strategies, phase and amplitude binning, on the image quality of our recently implemented 4D‐CBCT method using the Varian On Board Imager and RPM breathing motion surrogate. Method and Materials: We have implemented a 4D‐CBCT imaging procedure that adjusts the acquisition parameters to the breathing period. The Varian Real‐time Position Management system (RPM) was used to establish temporal coincidence between the respiratory motion and the CBCT projections. Projections were sorted into 10 bins. For the projections sorted into each bin, the variances about the RPM amplitude and phase means was calculated and compared to the image quality achieved in CBCTimages. 4D‐CBCTs were studied with a phantom with irregular motion cycle. The phantom consists of two acrylic spheres and one cylinder on a motion platform to mimic the superior‐inferior tumor motion mostly seen in patients. Results: With breathing cycle period variation but not amplitude variation, we found that amplitude binning produced fewer blurring artifacts which was consistent with RPM analysis which showed significantly smaller amplitude variances for amplitude binning. For the patient subjects, the RPM amplitude but not phase variances were smaller for amplitude binning. However, image quality was significantly better when phase binning was used. We hypothesize that the under‐sampling artifacts play a major role over the motion artifacts which is the case for amplitude binning, while the phase binning has a good control of aliasing artifacts. Residual motion artifacts persist in both phase‐ and amplitude‐ binning 4D‐CBCT images.Conclusion: Optimizing the binning strategies requires understanding complex interplay between residual motion in each bin and other factors such as uniformity and adequacy of angular sampling.
Supported by PPG‐NIH‐P01116602.
35(2008); http://dx.doi.org/10.1118/1.2962871View Description Hide Description
Purpose: To investigate a quantitative method for measuring breast density (the percentage of glandular breast tissue) with dual energy mammography using a dual exposure / dual kVp technique. Method and Materials: An analytical simulation model was developed to determine the mean glandular dose required to quantify breast density to within an accuracy of 1%. The simulation considered stochastic x‐ray and detector noise sources in a dual energy breast density image. The breast was modeled as a semicircle 10 cm in radius with homogenous equal thicknesses of adipose and glandular tissues. Monoenergetic beams were simulated with an ideal detector and polyenergetic spectra were simulated with an energy integrating detector modeled after an existing clinical system. A surrogate breast phantom was constructed to simulate a range of breast densities from 0 to 100%. The phantom was imaged using a full field digital mammographysystem at 25 kVp and 39 kVp with an additional 2 mm Al filter added to the high energy image. Dual energy images were generated with a non linear basis decomposition algorithm to yield image thickness maps of glandular and adipose tissue. Breast density was quantified using the dual energy images.Results: For a 4.2 cm breast of 50% density, optimal energies determined from simulation, for monoenergetic beams and an ideal detector, were 19 keV and 71 keV. For an energy integrating detector and polyenergetic spectra, optimal energies were 20 and 69 kVp. The estimated required dose for either technique was less than 0.1 mGy. For dual energy images, the known (K) and measured (M) breast densities were related by M = 1.007K − 0.006 (r = 0.999). RMS error for all densities was calculated to be less than 2% of the mean density. Conclusion: The results suggest that breast density can be accurately measured with dual energy mammography.
TH‐C‐332‐06: Optimization and Calibration Procedures of a Contrast‐Medium Based Subtraction Technique in Digital Mammography35(2008); http://dx.doi.org/10.1118/1.2962872View Description Hide Description
Purpose: To optimize radiological and clinical parameters for the application of a contrast‐medium‐based subtraction technique in digital mammography and to evaluate its clinical feasibility using a commercial mammography unit without any external or internal hardware modification. Method and Materials: This is a two‐stage project. Firstly, an extension of Lemacks' analytical formalism was implemented in order to maximize contrast‐to‐noise ratios (CNR) in simulated applications of dual‐energy and temporal subtraction modalities, and combinations of them. The formalism was validated by imaging tubular structures with an iodine‐based contrast medium embedded in a PMMA phantom. Once two optimized techniques were defined, calculations were performed to obtain exposure values limiting the total glandular dose to 2.5 mGy in a dynamical study of 1 mask + 3 post‐contrast images. Secondly, these parameters were used to calibrate the gray level in subtracted images as function of contrast medium iodine concentration contained in a multi‐well PMMA‐phantom. Optimized subtraction techniques are being applied in a clinical study which shall include 20 patients. Results: Experimental CNR results surpass Rose's criterion (CNR=5) and validate the predicted advantage of temporal techniques. Dual energy temporal subtraction, with iodine administered in the low energy image arises as the optimum subtraction technique, instead of the predicted advantage of contrast medium administration during the high‐energy acquisition. With respect to the calibration, relations between CNR and iodine concentration were found to be approximately linear beyond 8 mg/ml, although for lower concentrations, CNR is undetectable. Conclusion: The advantage of temporal over dual energy subtraction in terms of CNR has been validated, particularly of dual energy temporal subtraction. Furthermore, concentrations below 8 mg/ml do not satisfy the detection criterion.
35(2008); http://dx.doi.org/10.1118/1.2962873View Description Hide Description
Purpose: To investigate lesion enhancement with iodinated contrast material using dual energy breast computed tomography(CT).Method and Materials: The dual energy cone‐beam breast CT system consisted of a flat panel detector and an x‐ray tube installed on an optical bench. The x‐ray tube voltage and x‐ray beam filtration were switched at 30 Hz between 50 kVp (2.0‐mm Al filter) and 120 kVp (2.0‐mm Al + 0.8‐mm Ag filtration). A cylindrical breast phantom was constructed from polymethylmethacrylate (PMMA) and filled with oil and shredded PMMA. Five known concentrations of iodine (0.5, 1, 2, 4, 8 mgI/ml) were embedded in the phantom. The images were reconstructed with a Feldkamp filtered‐back‐projection algorithm. Dual energy subtraction was used to eliminate the contrast of oil and PMMA to enhance iodine signals. Results: The low energy scan could not distinguish PMMA and iodine concentrations of 0.5 and 1 mgI/ml. The high energy image could not distinguish PMMA and iodine concentration of 8 mgI/ml. After subtracting the PMMA, all five concentrations of iodine could be clearly visualized on the dual energy CTimage. The dual energy iodine signal (CTDE) and the iodine concentration (CI) were related by (R = 0.998 and SEE=15.26) Conclusion: Dual energy contrast‐enhanced breast CT can potentially improve lesion.
35(2008); http://dx.doi.org/10.1118/1.2962874View Description Hide Description
Purpose: To accurately estimate the parameters those describe the geometry of a stationary digital breast tomosynthesis system, namely, the source‐detector distance and the location of the focal spots. Method and Materials: Our novel stationary tomosynthesis system consists of 25 x‐ray sources linearly arranged to cover an angle of 48°. With a limited number of images, the tomosynthesisreconstruction becomes very sensitive to geometric alignment. It is vital that we accurately calibrate the geometry of the system with a suitable phantom. We have designed and built such a phantom that consists of two 6 mm stainless steel balls mounted on a low‐density plastic base. We obtained twelve evenly spaced projection images of the phantom over 360° for each x‐ray source. The motion of the two objects in a circle during data acquisition will cause their projections to trace ellipses on the detector. We calculated the centroids of the two objects in the detector plane and used these to derive the ellipse parameters. Together with the known distance between the two objects, this enabled us to analytically solve for the geometry parameters. Results: The source‐detector distance was calculated to be 69.4 ± 2 mm. The calculated distances between the sources were also in good agreement with the designed values. The parameters were used for the tomosynthesisreconstruction of a sponge phantom and the results showed improved image quality and geometric accuracy. Conclusion: We have designed and tested a calibration phantom that can accurately estimate the geometry of a stationary tomosynthesis system. We are continuing to assess the accuracy of our method to optimize our results. Further, we are looking at ways in which to make the process more efficient by calibrating for only the central source using this phantom and estimating the locations of the other sources using a simpler secondary phantom.
TH‐C‐332‐09: The Effect of Variable Exposure Distribution On Microcalcification Detectability in Tomosynthesis35(2008); http://dx.doi.org/10.1118/1.2962875View Description Hide Description
Purpose: To investigate the fundamental limitation of tomosynthesis acquisition on the detectability of microcalcifications (MCs) and to investigate the effects of employing unequal dose distribution (variable exposure) across the tomosynthesis projections. Method and Materials:Ray tracing was used through a 5‐cm thick homogeneous slab phantom, which contained 150‐, 280‐, and 400‐micron diameter spheres embedded at the center. In this study we accounted for acquisition geometry and x‐ray quantum noise. Detectability was calculated using a nonprewhitening observer. In the sinogram data, detectability was computed by integrating observer response over all projection views. In the reconstructed image, observer response was computed using a 2D matched template at the in‐focus slice of the sphere. To test variable distribution of exposure, 50% of the exposure was concentrated in the center three projection views, and the rest was divided equally among the remaining projection views. Results: We determined that the detectability of MCs is reduced for larger projection angles because of the increased pathlength through the phantom and larger source‐to‐detector distance. The spheres are projected on the detector as ellipses at larger angles, which blurs the MCs and decreases detectability. Detectability of MCs using the variable exposure method is approximately 10–20% higher than detectability in the equal exposure distribution method. Conclusion: The detectability of MC is reduced in a tomosynthesis acquisition because of the acquisition geometry if all projections are made with the same exposure. A variable exposure method can improve MC detectability in DBT systems. In future work we will examine the effect of breast structure noise on the variable exposure method. Conflict of Interest: Research sponsored by Hologic, Inc. and Dexela Ltd.
35(2008); http://dx.doi.org/10.1118/1.2962876View Description Hide Description
Purpose: To present the design and performance of a new automated whole‐breast ultrasoundcomputed tomography (USCT) system that employs 3D inverse‐scatter reconstruction. Method and Materials: The scanner (Techniscan Medical Systems, Inc., Salt Lake City, Utah) is installed at the University of California, San Diego to evaluate clinical performance including ability to detect and analyze breast masses. Patients lie prone on a table while opposing transmitter and receiver transducer arrays rotate 360° around the breast. Ultrasound plane waves (300 kHz–2MHz) are emitted every 2 degrees while the scattered beam is detected by a 960‐element six‐row transducer. In the same plane three B‐mode linear arrays acquire backscatter data from wideband signals (3.6–8.4 MHz). The arrays move up the breast in 2mm increments to scan the entire breast. Discrete frequency domain data is used by a proprietary 3D inverse‐scattering algorithm that incorporates multiple scattering within and between the planes. 2‐D coronal images of the entire breast are reconstructed as accurate quantitative maps of sound speed and attenuation along with aberration‐corrected high‐resolution reflection tomograms. Results: To date, more than 50 subjects were scanned with wide range in age (20–78), breast density and diagnostic outcome. Representative cases will be presented comparing mammography, sonography and multi‐planar USCT images. Included are benigns (cysts, fibroadenomas, fibrocystic disease) and biopsy‐proven malignancies (invasive ductal carcinomas, invasive tubular carcinoma and mixed lobular/ductal carcinoma). System performance and tissue characterization are excellent. Accuracy and linearity of sound speed measurements by USCT is very high (R2=0.988) over the range of 1400 to 1600 m/sec. Conclusion: The USCT system provides rapid, automated scanning with quantitative 3D images of the breast and substantially new information for characterizing breast masses. Conflict of Interest: Research sponsored in part by Techniscan Medical Systems, Inc.
- Dosimetry, Radiation Protection, and Quality Control I
35(2008); http://dx.doi.org/10.1118/1.2962400View Description Hide Description
Purpose: To numerically generate radiographic x‐ray spectra that can be conveniently employed in radiation transport simulations or other radiation detection applications. Method and Materials: Based initially on the Tucker, et al model, we developed and evaluated a new code, DXS (Diagnostic X‐Ray Spectra), to numerically generate spectra for tungsten‐target x‐ray tubes spanning the radiographic energy range. The model parameters in our code were adjusted by comparison with corresponding MCNP5 simulated spectra; we modified the semi‐analytical formulation for the characteristic x‐ray production, a caveat of Tucker's model, by incorporating a factor that better accounts for the dependence of the K‐peaks on the tube potential. Parametric fitting functions are used to model the self‐attenuation in the target and attenuation due to inherent and added filtration (aluminum,beryllium,copper,tantalum are the options implemented in the code), as well as for the tungsten mass stopping power and the Thomson‐Whiddington constant. Comparison with Monte Carlo simulated and published measured spectra were used to validate the new code. Results: Normalized to unit area DXS code‐generated spectra for several tube potentials from 50 to 140 kVp agree well, less than 2% relative difference in nearly all energy bins (2 keV), with corresponding MCNP5 simulated spectra for similar tube parameters. Few exceptions are noted and may be attributed to either poorer statistics in the low and high energy tails of the spectrum, or to insufficient accuracy of the numerical computations for the steepest part of the spectra at high accelerating potentials. Good agreement is seen between the DXS and Bhat et al measured spectra. Conclusion: The DXS code generates the spectra, according to user specified input parameters (tube potential, anode angle, filtratation) and energy intervals, and augments them into any discretized energy group structure. Hence, the code can be of great benefit in radiation transport simulations.
MO‐E‐332‐02: Skin Doses in Interventional Radiology Procedures Associated with Oncology Diagnosis and Treatment — Are There Reviewable Sentinel Events?35(2008); http://dx.doi.org/10.1118/1.2962401View Description Hide Description
Purpose: Fluoroscopic examinations with cumulative dose exceeding 15 Gy to a single field is now considered as a ‘reviewable sentinel event’ according to Joint Commission standards. Guidance from the FDA suggests that the potential for injury be recorded in the patient's record for cumulative absorbed dose of 1 Gy or more. The purpose of this study was to estimate the peak radiation skindoses for interventional radiology procedures performed at a high patient volume cancer center. Method and Materials: A single‐center, IRB‐approved retrospective study was performed using data from an oncologic interventional radiology section. Peak skindoses were estimated from consecutive procedures performed during 2006 in three different fluoroscopic suites equipped for these studies. Of 6598 consecutive procedures, 3966 (60%) had dose‐area‐product (DAP) measurements recorded and were included in the study. Results: The mean estimated peak skindose was 0.19 Gy (range 4.95 microGy to 8.65 Gy) with a maximum individual skindose of 8.65 Gy. No procedures resulted in skindoses >15 Gy and over 95% of the procedures resulted in skindoses <1 Gy. Procedures with specific instances of skindoses >1 Gy included: embolization, biliary drain/stent, IVC filter, nephrostomy, arteriogram, abscess catheters, foreign body retrieval, catheter change, cholecystostomy, and gastronomy tube check. Embolizations, and biliary drain/stent procedures were most likely to result in skindoses >1 Gy. Significant variations in skindose were noted for various instances of the same procedure (e.g. range 0.6 mGy to 8.65 Gy for hepatic embolizations). Conclusion: Even when potential errors in methodology are considered, it is unlikely that any typical case performed in an oncologic interventional radiology practice would exceed the Joint Commission ‘reviewable sentinel event’ level of 15 Gy. Identifying procedures that could have peak skindoses greater than 1 Gy can be useful for informed consent and clinical followup.
35(2008); http://dx.doi.org/10.1118/1.2962402View Description Hide Description
Purpose:Charge coupled devices (CCDs) are being increasingly used in radiation therapy.CCDcameras are becoming the tool of choice for applications such as two‐dimensional dosimetry of scintillator sheets or to read hundreds of miniature scintillation detectors arranged in arrays. However, CCDs are sensitive to stray radiation. This effect induces transient noise. Radiation‐induced noise strongly alters the image and therefore limits its quantitative analysis. The purpose of this work is to characterize the radiation‐induced noise and to develop filtration algorithms to restore image quality. Method and Materials: Two models of CCDcameras (Andor Luca and Apogee U2000) were used for measurements in linac environments. Images were acquired with and without radiation. The structure of the transient noise was first characterized. Then, three methods of noise filtration were compared: median filtering of a time series of identical images, uniform median filtering of single images and an adaptive filter with hard‐switching mechanism. Results: The intensity distribution of noisy pixels was similar in both cameras. However, the spatial distribution of the noise was different: the average noise cluster size was 1.2±0.6 and 3.2±2.7 pixels for the U2000 and the Luca respectively. The median of a time series of image resulted in the best filtration and minimal image distortion. For applications where time series is impractical, an adaptive filter must be used to reduce image distortion. We have implemented a modification to the switch filter in order to handle non‐isolated group of noisy pixels. Conclusion: We have characterized the transient noise produced in CCDcameras by scattered radiation from a linac and have developed an efficient filtration scheme to remove this noise and restore image quality. Use of our filtration scheme allows detailed quantitative analysis of an image even when subjected to scattered radiation.
Supported by the NCI (1R01CA120198‐01A2).
MO‐E‐332‐04: Comparison of Dose‐Area‐Product (DAP) and Fluoroscopy Time Between a Mobile and a Fixed C‐Arm Unit for Electrophysiology (EP) Procedures35(2008); http://dx.doi.org/10.1118/1.2962403View Description Hide Description
Purpose: To determine if there is a difference in dose‐area‐product (DAP) or fluoroscopy time between electrophysiology (EP) procedures performed using a mobile and a fixed C‐arm fluoroscopic unit. Method and Materials: DAP and fluoroscopy‐time data was logged for 800 EP procedures performed using a mobile C‐arm unit from 2003 to 2005. In early 2006, a fixed C‐arm unit was installed and the same data logged for over 200 procedures. The procedures were sorted into five categories: 1. electrophysiology studies, 2. radiofrequency ablations, 3. pacemaker and implantable‐cardiac‐defibrillator implants, 4. biventricular interventions, and 5. lead changes. For each category, the distributions of DAP and time were compared and the average, median and range of values determined. Results: The median fluoroscopy time more than doubled for all procedure categories when using the fixed unit. Median DAP increased significantly for procedures 3 and 4, but remained nearly the same for procedures 1, 2 and 5. In all cases, the procedures were successfully completed without evidence of compromising patient care for either unit. However, the cardiologists were much more conservative in their use of fluoroscopy for the mobile unit due to its heat loading limitations and, also, they worked quicker because of their impression that the older mobile C‐arm gave more radiation dose to themselves and the patient. In addition, fluoroscopy at 15 frames per second was used on the fixed unit versus 7 frames per second on the mobile unit. Conclusion: Although the DAP and fluoroscopy time generally was higher for the fixed installation, it should not be concluded that the mobile unit is to be preferred. Rather, these results point out the importance of physician training and dose monitoring, not only to track patient radiation risk, but also to provide physician feedback.
(*Support: NIH Grant R01‐NS43924).
MO‐E‐332‐05: The Effect of Copper Beam Filtration On the Transmission of Scattered X‐Rays Through a Typical Lead Barrier35(2008); http://dx.doi.org/10.1118/1.2962404View Description Hide Description
Purpose: Standard use of copper beam filtration in modern cardiac catheterization and angiography systems can substantially change the x‐ray beam quality and ultimately have an impact on scattered x‐ray transmission through shielded barriers. A study was performed to investigate these effects. Method and Materials: Scatter was measured using broad‐beam geometry at 50 and 100 cm from the center of an 8″‐thick Lucite phantom with an 1800 cc ionization chamber. A Siemens Axiom Artis system was employed using 60, 81, 102 and 125 kVp with filtrations of 0, 0.1, 0.2, 0.3, 0.6 and 0.9mm of copper. The ionization chamber was wrapped in 1/16″‐thick lead to determine the transmission of scattered x‐rays through a typical shielded barrier. Results: Transmission at 125 kVp varied from 1.7×10−3 to 3.0×10−3 for beams with 0 to 0.9 mm of Cu, respectively. Similarly, transmission varied at 102 kVp from 1.2×10−3 to 2.2×10−3 and at 81 kVp from 7.1×10−4 to 9×10−4. Transmission at 60 kVp was about 7×10−4 for beams with 0 to 0.3 mm of Cu. Transmissions without Cu filtration at 125 and 102 kVp were measured to be about half the theoretical value reported by Simpkin and Dixon, which could result from variations in lead thickness. At 60 and 81 kVp transmissions without Cu filtration are more than an order of magnitude higher than that reported at 70 kVp by Simpkin and Dixon. Conclusion: Additional copper filtration can increase the barrier transmission up to a factor of two over unfiltered beams. However, it is unlikely that this amount of increase in transmission will significantly modify a shielding calculation. Further investigation is needed to determine the changes in typical workloads and scattered air kerma at different angles for these systems to understand the combined effect of these changes on barrier shielding calculations.
35(2008); http://dx.doi.org/10.1118/1.2962405View Description Hide Description
Purpose: To create a series of EPID Monte Carlo dose computation kernels which accounts for observed machine‐to‐machine variations in EPID response. Method and Materials:Field size response of aS500 and aS1000 imagers are measured for several Varian Cl21‐series machines that were dosimetrically matched in a water phantom. Deviations in imager response are attributed to differences in back‐scattering materials beneath the imaging panels. Mono‐energetic convolution kernels with various backscatter thicknesses are simultaneously created by sub‐dividing a thick back‐scattering slab into multiple sub‐slabs and using the EGSnrc LATCH bit to score sub‐slab kernel contributions. Energy‐binned particle fluence incident upon the detector convolved with the imager‐specific kernels are used to compute the EPIDimage.Imager‐specific kernels are determined by matching computed and measured EPID field‐size response, using the number of sub‐slabs as a free parameter. Final kernels are used for Monte Carlo‐based pre‐treatment and in‐treatment EPID dose computations. Results: The EPIDimagers on dosimetrically matched accelerators are found to differ. Most, but not all of the deviations appear to be correlated with the imager mounting arm type. The imager‐specific kernels matched the field‐size response for each imager within 1%, and resulted in dosimetric agreement between measured and computed images for pre‐treatment dosimetric verification of IMRT fields. Conclusion: Dosimetric differences between portal imagers on matched accelerators can be accounted for by using computation kernels with differing amounts of back‐scattering materials. Kernels for multiple different back‐scattering thickness can be efficiently calculated. Resultant imager‐specific kernels may be useful for efficient pre‐treatment and in‐treatment Monte Carlo‐based EPID dose computations. Conflict of Interest: This work was funded in part by Varian Medical Systems.
MO‐E‐332‐07: Instrumentation Noise Equivalent Exposure (INEE) for Routine Quality Assurance: INEE Measurements On a Clinical Flat Panel Detector35(2008); http://dx.doi.org/10.1118/1.2962406View Description Hide Description
Purpose: To measure the instrumentation‐noise equivalent exposure (INEE) of a clinical digital imagingsystem under various modes of operation. Method and Materials: The INEE is defined as the exposure at which the quantum‐noise equals the instrumentation‐noise and is measured from the plot of pixel gray‐level‐value variance versus detector entrance exposure. The intercept of such a plot represents the instrumentation‐noise in gray‐level values and the slope provides the conversion factor from these arbitrary units to equivalent exposure. Sequences of 90 flat‐field images in DA and DSA modes were acquired at 1 frame‐per‐second using a Varian PaxScan 2020 flat panel detector, both with image processing enabled and disabled. Image receptor input exposure was measured using an ionization chamber and effects of uncertainties in exposure calibration on the resulting INEE were investigated. Results: The INEE was observed to vary depending on the mode used and was measured to be 0.9±0.2 and 44±6 μR in DA and DSA modes, respectively, with image processing disabled. Exposure calibration error will result in additional proportionate INEE error. With image processing enabled the variance was highly nonlinear with exposure in DA mode indicating a need to have access to linear, unprocessed data. The difference in integration capacitance between DA and DSA modes, 0.5 and 4 pF respectively, helps explain the change in the INEE between these modes. Conclusion: INEE measurements were done on a clinical digital image receptor and were found to be dependent on the operational mode used. The importance of disabling image processing during INEE measurement was demonstrated. Due to the ease in gathering and analyzing the necessary data, we found the INEE to be a useful quality assurance tool for assessing a system's instrumentation‐noise in terms of a clinically relevant measure of exposure. (Support: NIH Grants R01‐ NS43924, R01‐EB002873, Toshiba Medical Systems Corporation).
- Dosimetry, Radiation Protection, and Quality Control II
35(2008); http://dx.doi.org/10.1118/1.2962793View Description Hide Description
Introduction : Tomographic techniques in diagnostic and interventional radiology or radiation therapy are in rapid development. The CTDI formalism that is currently used to assess dose has been challenged by several groups. The goal of this contribution is to evaluate the required scan length to reach equilibrium on various MDCT units and estimate the average dose delivered within a slice when using systems having a cone beam geometry.
Method and Materials:Measurements were performed on three MDCT systems (GEMS 8, 16 and 64‐row), a flat panel fluoroscopy (Allura — Philips), an IGRT (Synergy, Elekta ltd) and on a tomotherapy (TomoTherapy Inc.) using two standard CTDI phantoms and a home made phantoms (PMMA cylinder of 30 cm filled with water) with a conventional small volume ion chamber (0.6 cc Farmer type) and a standard pencil ion chamber.Dose profiles were also recorded at various positions within the slice to study the impact of scatter as a function of the distance between the centre of the phantom and its periphery. Results: For CT units a theoretical length of 400 mm is required to reach the equilibrium at the center of the phantom. The measurements show that a dose plateau was reached after 420 mm. The average dose within a slice measured in our home made phantom for standard abdominal CT (120 kV, 210 effective mAs) was 15.3 mGy; 26.9 mGy for the fluoro‐CT (122kV, 274mAs, 20.7s); 39.0mGy for IGRT system (120kV, 219mAs, 27×27cm2field size, one rotation) and 13.1 mGy for the Tomotherapy system (3.5 MV, pitch=0.8). Conclusion: The CTDI concept should be replaced by a more generic methodology, such as the one presented here, that could be used in all cone beam geometries. Measurements should be done using water equivalent phantoms.
35(2008); http://dx.doi.org/10.1118/1.2962794View Description Hide Description
Purpose: The O‐arm is a cone beam imaging system designed primarily to support orthopedic surgery as well as for image‐guided and vascular surgery. Using a gantry that can be opened or closed, the O‐arm can function as a 2D fluoroscopy device or collect 3D volumetric imaging data like a CT system. Our clinical applications of the O‐arm in spine surgical procedures, assessment of pedicle screw position, kyphoplasty procedures, and etc show that the O‐arm 3D mode provides enhanced imaging information in the surgical procedure compared to radiographs or fluoroscopy alone. However, the radiation dose of the O‐arm has remained uninvestigated. This study is to investigate patient dose and scatter radiation from an O‐arm and compare the results to those from a CT scanner and a conventional C‐arm. Method and Materials: The patient dose was measured using a 0.6 cc Farmer ion chamber and 30 cm long CT head and body phantoms. Scatter radiation was measured at several locations around the O‐arm, at 1m, 2m and 3m distances from the iso‐center of the O‐arm, in both the 2D fluoroscopic mode and the 3D mode with a Radcal 10×5−180 pancake ion chamber using a 30 cm long CTDI body phantom as the source of scatter. The same measurements were made for an OEC C‐arm and a 64 slice CT scanner, respectively. Results: The results show that under identical technical conditions and with the same scan length, the O‐arm 3D mode delivers radiation dose to patients and scatterdose to personnel that is comparable to that of the 64 slice CT scanner. The O‐arm 2D mode produces similar scatter radiation as a conventional GE OEC fluoroscopic C‐arm system. Conclusion: Our study demonstrated that the O‐arm had comparable radiation dose to patients and radiologists as CT and C‐arm systems.
WE‐E‐332‐03: Evaluating Effective Doses and Image Quality for Multislice CT Using Automatic Tube Current Modulation35(2008); http://dx.doi.org/10.1118/1.2962795View Description Hide Description
Purpose: Automatic tube current modulation (ATCM) is one of innovative techniques to lower radiation doses in computed tomography(CT) scanning. The purpose of this study is to estimate the image quality, organ doses and effective doses in clinical multi‐detector CT examinations using ATCM techniques. Method and materials: An anthropomorphic phantom was scanned on a Siemens Sensation 64 multi‐detector CT scanner with CARE Dose4D software which offered nine combinations of ATCM. The scanning regions included head, neck, chest, abdomen and pelvis. The image quality was derived from noise distributions in each image slice using MatLab. Organ does and effective doses were measured by high sensitivity thermoluminescent dosimeters (LiF: Mg, Cu, P) which were inserted into the anthropomorphic phantom at the selected organs.Results: CARE Dose 4D software tends to trigger increasing intensity modulation in the parts of shoulder (upper lung) and lower lung, but to trigger decreasing intensity modulation in the parts of head, neck, middle lung, abdomen and pelvis. Effective doses from ATCM were lower than those from fix tube current except neck examinations. The dose decreases vary from 36% to 68.5% in head examinations, 12.0% to 38.9% in chest examinations, 31.4% to 43.4% in abdomen examinations, and 33% to 55.5% in pelvis examination, but the dose increases are from 18.4% to 48% in neck examinations. The noise increase ratios, which were derived from the images using ATCM comparing with those using fix tube current, are from −1.9% to 2.5%, −15.0% to −19.9%, −13.1% to 0.7%, 7.6% to 15.7%, and 16.2% to 28.0% in head, neck, chest, abdomen and pelvis examinations, respectively. Conclusion: ATCM techniques with appropriate use could reduce effective doses but slightly increase imagenoises. ATCM is an efficient method to reduce patients' radiation doses during CT scanning if there is no influence on the diagnostic acceptability.