O Christianson

A methodology for image quality evaluation of advanced CT systems.

PURPOSE This work involved the development of a phantom-based method to quantify the performance of tube current modulation and iterative reconstruction in modern computed tomography (CT) systems. The quantification included resolution, HU accuracy, noise, and noise texture accounting for the impact of contrast, prescribed dose, reconstruction algorithm, and body size. METHODS A 42-cm-long, 22.5-kg polyethylene phantom was designed to model four body sizes. Each size was represented by a uniform section, for the measurement of the noise-power spectrum (NPS), and a feature section containing various rods, for the measurement of HU and the task-based modulation transfer function (TTF). The phantom was scanned on a clinical CT system (GE, 750HD) using a range of tube current modulation settings (NI levels) and reconstruction methods (FBP and ASIR30). An image quality analysis program was developed to process the phantom data to calculate the targeted image quality metrics as a function of contrast, prescribed dose, and body size. RESULTS The phantom fabrication closely followed the design specifications. In terms of tube current modulation, the tube current and resulting image noise varied as a function of phantom size as expected based on the manufacturer specification: From the 16- to 37-cm section, the HU contrast for each rod was inversely related to phantom size, and noise was relatively constant (<5% change). With iterative reconstruction, the TTF exhibited a contrast dependency with better performance for higher contrast objects. At low noise levels, TTFs of iterative reconstruction were better than those of FBP, but at higher noise, that superiority was not maintained at all contrast levels. Relative to FBP, the NPS of iterative reconstruction exhibited an ~30% decrease in magnitude and a 0.1 mm(-1) shift in the peak frequency. CONCLUSIONS Phantom and image quality analysis software were created for assessing CT image quality over a range of contrasts, doses, and body sizes. The testing platform enabled robust NPS, TTF, HU, and pixel noise measurements as a function of body size capable of characterizing the performance of reconstruction algorithms and tube current modulation techniques.

Automated size-specific CT dose monitoring program: Assessing variability in CT dose

PURPOSE The potential health risks associated with low levels of ionizing radiation have created a movement in the radiology community to optimize computed tomography (CT) imaging protocols to use the lowest radiation dose possible without compromising the diagnostic usefulness of the images. Despite efforts to use appropriate and consistent radiation doses, studies suggest that a great deal of variability in radiation dose exists both within and between institutions for CT imaging. In this context, the authors have developed an automated size-specific radiation dose monitoring program for CT and used this program to assess variability in size-adjusted effective dose from CT imaging. METHODS The authors radiation dose monitoring program operates on an independent health insurance portability and accountability act compliant dosimetry server. Digital imaging and communication in medicine routing software is used to isolate dose report screen captures and scout images for all incoming CT studies. Effective dose conversion factors (k-factors) are determined based on the protocol and optical character recognition is used to extract the CT dose index and dose-length product. The patients thickness is obtained by applying an adaptive thresholding algorithm to the scout images and is used to calculate the size-adjusted effective dose (ED(adj)). The radiation dose monitoring program was used to collect data on 6351 CT studies from three scanner models (GE Lightspeed Pro 16, GE Lightspeed VCT, and GE Definition CT750 HD) and two institutions over a one-month period and to analyze the variability in ED(adj) between scanner models and across institutions. RESULTS No significant difference was found between computer measurements of patient thickness and observer measurements (p = 0.17), and the average difference between the two methods was less than 4%. Applying the size correction resulted in ED(adj) that differed by up to 44% from effective dose estimates that were not adjusted by patient size. Additionally, considerable differences were noted in ED(adj) distributions between scanners, with scanners employing iterative reconstruction exhibiting significantly lower ED(adj) (range: 9%-64%). Finally, a significant difference (up to 59%) in ED(adj) distributions was observed between institutions, indicating the potential for dose reduction. CONCLUSIONS The authors developed a robust automated size-specific radiation dose monitoring program for CT. Using this program, significant differences in ED(adj) were observed between scanner models and across institutions. This new dose monitoring program offers a unique tool for improving quality assurance and standardization both within and across institutions.

Automated Technique to Measure Noise in Clinical CT Examinations

OBJECTIVE The purpose of this study was to develop and validate an automated method to measure noise in clinical CT examinations. MATERIALS AND METHODS An automated algorithm was developed to measure noise in CT images. To assess its validity, the global noise level was compared with image noise measured using an image subtraction technique in an anthropomorphic phantom. The global noise level was further compared with image noise values from clinical patient CT images obtained by an observer study. Finally, the clinical utility of the global noise level was shown by assessing variability of image noise across scanner models for abdominopelvic CT examinations performed in 2358 patients. RESULTS The global noise level agreed well with the phantom-based and clinical image-based noise measurements, with an average difference of 3.4% and 4.7% from each of these measures, respectively. No significant difference was detected between the global noise level and the validation dataset in either case. It further indicated differences across scanners, with the median global noise level varying significantly between different scanner models (15-35%). CONCLUSION The global noise level provides an accurate, robust, and automated method to measure CT noise in clinical examinations for quality assurance programs. The significant difference in noise across scanner models indicates the unexploited potential to efficiently assess and subsequently improve protocol consistency. Combined with other automated characterization of imaging performance (e.g., dose monitoring), the global noise level may offer a promising platform for the standardization and optimization of CT protocols.

Evaluation of Low-Contrast Detectability of Iterative Reconstruction across Multiple Institutions, CT Scanner Manufacturers, and Radiation Exposure Levels.

PURPOSE To compare image resolution from iterative reconstruction with resolution from filtered back projection for low-contrast objects on phantom computed tomographic (CT) images across vendors and exposure levels. MATERIALS AND METHODS Randomized repeat scans of an American College of Radiology CT accreditation phantom (module 2, low contrast) were performed for multiple radiation exposures, vendors, and vendor iterative reconstruction algorithms. Eleven volunteers were presented with 900 images by using a custom-designed graphical user interface to perform a task created specifically for this reader study. Results were analyzed by using statistical graphics and analysis of variance. RESULTS Across three vendors (blinded as A, B, and C) and across three exposure levels, the mean correct classification rate was higher for iterative reconstruction than filtered back projection (P < .01): 87.4% iterative reconstruction and 81.3% filtered back projection at 20 mGy, 70.3% iterative reconstruction and 63.9% filtered back projection at 12 mGy, and 61.0% iterative reconstruction and 56.4% filtered back projection at 7.2 mGy. There was a significant difference in mean correct classification rate between vendor B and the other two vendors. Across all exposure levels, images obtained by using vendor Bs scanner outperformed the other vendors, with a mean correct classification rate of 74.4%, while the mean correct classification rate for vendors A and C was 68.1% and 68.3%, respectively. Across all readers, the mean correct classification rate for iterative reconstruction (73.0%) was higher compared with the mean correct classification rate for filtered back projection (67.0%). CONCLUSION The potential exists to reduce radiation dose without compromising low-contrast detectability by using iterative reconstruction instead of filtered back projection. There is substantial variability across vendor reconstruction algorithms.

DQE of wireless digital detectors: Comparative performance with differing filtration schemes

PURPOSE Wireless flat panel detectors are gaining increased usage in portable medical imaging. Two such detectors were evaluated and compared with a conventional flat-panel detector using the formalism of the International Electrotechnical Commission (IEC 62220-1) for measuring modulation transfer function (MTF), normalized noise power spectrum (NNPS), and detective quantum efficiency (DQE) using two different filtration schemes. METHODS Raw images were acquired for three image receptors (DRX-1C and DRX-1, Carestream Health; Inc., Pixium 4600, Trixell) using a radiographic system with a well-characterized output (Philips Super80 CP, Philips Healthcare). Free in-air exposures were measured using a calibrated radiation meter (Unfors Mult-O-Meter Type 407, Unfors Instruments AB). Additional aluminum filtration and a new alternative combined copper-aluminum filtration were used to conform the x ray output to IEC-specified beam quality definitions RQA5 and RQA9. Using the IEC 62220-1 formalism, each detector was evaluated at XN∕2, XN, and 2XN, where the normal exposure level to the detector surface (XN) was set to 8.73 μGy (1.0 mR). The prescribed edge test device was used to evaluate the MTF, while the NNPS was measured using uniform images. The DQE was then calculated from the MTF and NNPS and compared across detectors, exposures, and filtration schemes. RESULTS The three DR systems had largely comparable MTFs with DRX-1 demonstrating lower values above 1.0 cycles∕mm. At each exposure, DRX-1C and Pixium detectors demonstrated better noise performance than that of DRX-1. Zero-frequency DQEs for DRX-1C, Pixium, and DRX-1 detectors were approximately 74%, 63%, and 38% for RQA5 and 50%, 42%, and 28% for RQA9, respectively. CONCLUSIONS DRX-1C detector exhibited superior DQE performance compared to Pixium and DRX-1. In terms of filtration, the alternative filtration was found to provide comparable performance in terms of rank ordering of different detectors with the added convenience of being less bulky for in-the-field measurements.

Variability in Radiation Dose From Repeat Identical CT Examinations: Longitudinal Analysis of 2851 Patients Undergoing 12,635 Thoracoabdominal CT Scans in an Academic Health System

OBJECTIVE The purpose of this study was to conduct longitudinal analyses of radiation dose data from adult patients undergoing clinically indicated, repeat identical thoracoabdominal CT examinations. MATERIALS AND METHODS Radiation dose data were electronically collected from 2851 subjects undergoing 12,635 repeat identical CT scans (mean number of scans per patient, 4.8; range, 2-33) in one health system. Included CT protocols were chest-abdomen-pelvis with contrast administration (n = 4621 CT studies of 1064 patients), abdomen-pelvis with contrast administration (n = 876 CT studies of 261 patients), renal stone (n = 1053 CT studies of 380 patients), and chest (n = 6085 CT studies of 1146 patients) without contrast administration. A radiation-tracking software infrastructure was adopted to extract data from DICOM headers in PACS. Size-specific dose estimate (SSDE) was calculated. RESULTS A trend was observed toward global reduction in SSDE values with all protocols investigated (chest-abdomen-pelvis slope, -1.78; abdomen-pelvis slope, -0.82; renal stone slope, -0.83; chest slope, -0.47; p < 0.001 for all comparisons). The intraindividual analyses of radiation dose distribution showed widespread variability in SSDE values across the four protocols investigated (chest-abdomen-pelvis mean coefficient of variance, 14.02 mGy; abdomen-pelvis mean coefficient of variance, 10.26 mGy; renal stone mean coefficient of variance, 34.18 mGy; chest mean coefficient of variance, 6.74 mGy). CONCLUSION Although there is a trend toward global reduction in radiation doses, this study showed widespread variability in the radiation dose that each patient undergoing identical repeat thoracoabdominal CT protocols absorbs. These data may provide a foundation for the future development of best-practice guidelines for patient-specific radiation dose monitoring.

Improved Nuclear Medicine Uniformity Assessment with Noise Texture Analysis

Because γ cameras are generally susceptible to environmental conditions and system vulnerabilities, they require routine evaluation of uniformity performance. The metrics for such evaluations are commonly pixel value–based. Although these metrics are typically successful at identifying regional nonuniformities, they often do not adequately reflect subtle periodic structures; therefore, additional visual inspections are required. The goal of this project was to develop, test, and validate a new uniformity analysis metric capable of accurately identifying structures and patterns present in nuclear medicine flood-field uniformity images. Methods: A new uniformity assessment metric, termed the structured noise index (SNI), was based on the 2-dimensional noise power spectrum (NPS). The contribution of quantum noise was subtracted from the NPS of a flood-field uniformity image, resulting in an NPS representing image artifacts. A visual response filter function was then applied to both the original NPS and the artifact NPS. A single quantitative score was calculated on the basis of the magnitude of the artifact. To verify the validity of the SNI, an observer study was performed with 5 expert nuclear medicine physicists. The correlation between the SNI and the visual score was assessed with Spearman rank correlation analysis. The SNI was also compared with pixel value–based assessment metrics modeled on the National Electrical Manufacturers Association standard for integral uniformity in both the useful field of view (UFOV) and the central field of view (CFOV). Results: The SNI outperformed the pixel value–based metrics in terms of its correlation with the visual score (ρ values for the SNI, integral UFOV, and integral CFOV were 0.86, 0.59, and 0.58, respectively). The SNI had 100% sensitivity for identifying both structured and nonstructured nonuniformities; for the integral UFOV and CFOV metrics, the sensitivities were only 62% and 54%, respectively. The overall positive predictive value of the SNI was 87%; for the integral UFOV and CFOV metrics, the positive predictive values were only 67% and 50%, respectively. Conclusion: The SNI accurately identified both structured and nonstructured flood-field nonuniformities and correlated closely with expert visual assessment. Compared with traditional pixel value–based analysis, the SNI showed superior performance in terms of its correlation with visual perception. The SNI method is effective for detecting and quantifying visually apparent nonuniformities and may reduce the need for more subjective visual analyses.

WE-D-18A-02: Performance Evaluation of Automatic Exposure Control (AEC) Across 12 Clinical CT Systems

PURPOSE Automatic exposure control (AEC) is not typically evaluated or monitored in CT quality assurance programs. The purpose of this study was to develop/evaluate a new AEC testing platform for the clinical physics program at our institution, and characterize AEC performance across different CT systems. METHODS The Mercury Phantom comprises three tapered and four uniform regions of polyethylene(16, 23, 30, and 37 cm in diameter); each region includes four inserts: air, Polystyrene, Acrylic, and Teflon. The phantom was imaged using AEC and a fixed tube current technique across 12 clinical CT scanners. Those included five Siemens Somatom Definition Flash, four GE Discovery CT750HD, and three GE Lightspeed VCT systems. A custom MATLAB software package provided MTF, NPS, and detectability indices for each diameter section of the phantom. Detectability indices were used to evaluate the relationship between AEC setting, patient size, and image quality. The magnitude of the power of a best fit exponential curve to the detectability indices and phantom diameter was used as a measure of AEC strength. Results were compared within/across scanner models, and as baseline values for comparison with future system performance testing. RESULTS For each scanner model, the percent difference in expected image quality and AEC setting was under 3%(+/-2%). The average decrease in detectability between the small and large diameter phantom sections for the Siemens Flash, GE CT750, and GE VCT was 99%(+/-10%), 42%(+/-25%), and 33%(+/-41%), respectively. The value signifying AEC strength was 0.051(+/-13%), 0.019(+/-18%), and 0.018(+/-26%), for the Siemens Flash, GE CT750, and GE VCT models, respectively. CONCLUSION This study demonstrated a practical approach to test the AEC performance of clinical CT systems at a large academic medical center. The quantification and evaluation of AEC performance should be included in acceptance testing and in annual physics testing of clinical CT systems.

TH‐E‐217BCD‐07: Quantitative Comparison of Noise Texture Across CT Scanners from Different Vendors

Purpose: To quantitatively compare noise texture across CTscanners from different vendors using the Noise Power Spectrum (NPS). Methods: The American College of Radiology CT accreditation phantom (Gammex 464, Gammex, Inc., Middleton, WI) was imaged on two scanners: Discovery CT 750HD (GE Healthcare, Waukesha, WI), and SOMATOM Definition Flash (Siemens Healthcare, Germany), using a consistent acquisition protocol (120 kVp, 5 mm slice thickness, 250/200 mAs, and 22 cm field of view). Images were reconstructed using filtered backprojection and a wide selection of reconstruction kernels. For each image set, the 2D NPS were estimated from the uniform section of the phantom. The spectra were normalized by their integral value, radially averaged, and filtered by the human visual response function. A systematic kernel‐by‐kernel comparison across vendors was performed by computing the peak locations of the spectra and the root mean square error (RMSE) between the filtered NPS of each kernel. GE and Siemens kernels were compared and kernel pairs that minimized the RMSE and the spectra peak location difference were matched. Results: The RMSE between the NPS of GE and Siemens kernels varied from 0.01 to 0.30 mm2. The GE kernels ‘Soft’, ‘Standard’, ‘Chest’, and ‘Lung’ closely matched the Siemens kernels ‘B35f’, ‘B43f’, ‘B46f’, and ‘B80f’ (RMSE<0.03 mm2, peak difference<0.02 cycles/mm). The GE ‘Bone’, ‘Bone+’, and ‘Edge’ kernels were all matched to Siemens ‘B75f’ kernel but with sizeable RMSE and peak differences up to 0.20 mm2 and 0.47 cycles/mm respectively. These sizeable RMSE and peak differences corresponded to visually perceivable differences in the noise texture of the images. Conclusions: It is possible to use the NPS to quantitatively compare noise texture across CT systems. The degree to which similar texture across scanners could be achieved varies and is limited by the kernels available on each scanner.

Dose Index Analytics: More Than a Low Number

What do we know about CT dose, what is its variability across medical practices, and how can we ensure that our patients are safe, considering the complexities and heterogeneity of clinical care within and across medical facilities? The foundational reason we care about CT dose is radiation risk. The risk to an individual patient from a CT scan is likely to be very small. The exact magnitude has been very difficult to ascertain, as the estimates have been derived from data with large uncertainties. Although this uncertainty has given rise to debates, experts agree that individual patients should be exposed only to the minimum amount of radiation necessary for a clinical diagnosis. This position is consistent with the principal precept of bioethics, “primum non nocere” (first, do no harm). In the face of uncertainty, we are morally obligated to do the safest thing. One may approach minimizing patient dose by prospectively prescribing imaging protocols that deliver low dose. Although this step is necessary, it relies on the assumption that the prescribed dose matches the actual delivered dose. There are numerous documented cases in which this assumption has failed, leading to radiation overdoses causing deterministic effects. Even in the absence of deterministic effects, however, many more instances of overdosing (ie, using doses more than necessary) can go unnoticed. Without a mechanism in place to retrospectively assess actual delivered radiation doses, it is impossible to ensure that dose problems do not occur. Therefore, dose monitoring is an essential element of safe and consistent imaging practice.