Optimization of Hardware and Image Processing for Improved Image Quality in X-Ray Phase Contrast Imaging.
Abstract not provided.
Dagel, Amber Lynn↗
SEARCH · Engineering Papers
Results for “image quality”
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Image quality and diagnostic accuracy of complex-averaged high b value images in diffusion-weighted MRI of prostate cancer
Purpose: To evaluate the impact of complex-averaging on image quality (IQ) and diagnostic accuracy of acquired and calculated high b value (aHBV, cHBV) images in diffusion-weighted prostate MRI. Materials and methods: This retrospective study included 84 patients who underwent multiparametric prostate MRI at 3 Tesla without endorectal coil. DWIs were acquired at three different b values which included two lower b values (b = 50,900 s/mm{sup 2}) and one higher b value (aHBV at 2000 s/mm{sup 2}). The acquired data were postprocessed to generate two different types of trace-weighted images—using conventional magnitude-averaging and complex-averaging. Using lower b values (b = 50,900 s/mm{sup 2}) from both conventional and complex-averaged image sets, cHBV images (b = 2000 s/mm{sup 2}) and ADC maps were derived. All image sets were reviewed by two radiologists in different reading sessions to assess image quality and PIRADS. The diagnostic accuracy of different image sets for the detection of prostate lesions was performed by correlating PIRADS and Gleason scores. Results: Complex-averaging did not impact ADC values of the prostate lesions compared to magnitude-averaging (P = 0.08). Complex-averaging improved image quality of acquired high b value and calculated high b value images (P < 0.0001). Complex-averaging also improved the level of confidence (LOC) of the acquired high b value for both readers (P < 0.0001, P < 0.05), but only for reader A in calculated high b value (P < 0.0001). The image quality of calculated high b value images was not significantly different than acquired high b value images. The dataset combining complex-averaging and calculated high b value provided the highest diagnostic accuracy (but not statistically significant) for detection of the significant prostate lesion compared to the magnitude-averaged acquired high b value (79.55% vs. 72.73%; P = 0.317). The mean acquisition time for b = 2000 s/mm{sup 2} sequence (aHBV) was 6 min 30 s (± 1 min 16 s) out of a total of 28 min 31 s (± 4 min 26 s) for the entire mp-MRI protocol (approximately 25% of total scan time). Conclusion: Complex-averaging provides better image quality and level of confidence without significant impact on ADC values and diagnostic accuracy for detection of the significant prostate lesions . The calculated high b value images are also comparable to (and can substitute) the acquired high b value images which can help in reducing the imaging time.
62 RADIOLOGY AND NUCLEAR MEDICINE↗
Rapid kVp switching dual-energy CT in the assessment of urolithiasis in patients with large body habitus: preliminary observations on image quality and stone characterization
Purpose: The purpose of this study was to investigate the image quality (IQ) considerations of rapid kVp switching dual-energy CT (rsDECT) in the assessment of urolithiasis in patients with large body habitus and to evaluate whether it allows stone characterization. Materials and methods: In this IRB-approved, HIPAA compliant retrospective study, 93 consecutive patients (M/F = 72/21, mean age 56.9 years, range 23–83 years) with large body habitus (> 90 kg/198 lbs) who underwent dual-energy (DE) stone protocol CT on a rapid kVp switching DECT scanner between January 2013 and December 2016 were included. Scan acquisition protocol included an initial unenhanced single-energy CT (SECT) scan of KUB followed by targeted DECT in the region of stones. Two readers evaluated both CT data sets (axial 5 mm 120 kVp/140 kVp QC/70 keV monoenergetic, material density water/iodine images and coronal/sagittal 3 mm images) for the assessment of image quality (Scores: 1–4) and characterization of stone composition (reference standard: crystallography). Results: One hundred and five CT examinations were performed in 93 patients (mean body weight 105.12 ± 13.53 kg, range 91–154 kg), and a total of 321 urinary tract calculi (mean size-4.8 ± 3.2 mm, range 1.2–22 mm) were detected. Both SECT and targeted monoenergetic images were of acceptable image quality (mean IQ: 3.77 and 3.83, kappa 0.79 and 0.87 respectively). Material density water and iodine images had lower IQ scores (mean IQ: 2.97 and 3.09 respectively) with image quality deterioration due to severe photon starvation/streak artifacts in 20% (21/105) and 17% (18/105) scans, respectively. Characterization of stone composition into uric acid/non-uric acid stones was achieved in 93.14% (299/321) of calculi (mean size: 4.99 ± 3.3 mm, range 1.2–22 mm), while 7% (22/321) stones could not be characterized (mean size 3.03 ± 1.16 mm, range 1.6–6.4 mm) (p < 0.001). Most common reason for non-characterization was image quality deterioration of the material density iodine images due to severe photon starvation artifacts. On multivariate regression, stone size and patient weight were predictors of stone composition determination on DECT (p < 0.05). The transverse diameter had a weak negative correlation with stone composition determination, but it was not statistically significant. Stone characterization into uric acid vs. non-uric acid stones was accurate in 95% (n = 38/40) of stones in comparison with crystallography. Conclusion: In patients with large body habitus, rsDECT allowed characterization of most calculi (93%) despite image quality deterioration due to photon starvation/streak artifacts in up to 20% of material density images. Stone size and patient weight were predictors of stone composition determination on DECT, and small calculi in very large patients may not be characterized.
62 RADIOLOGY AND NUCLEAR MEDICINE↗
Metal Artifact Reduction for Orthopedic Prosthesis in Lower Extremity CT Venography: Evaluation of Image Quality and Vessel Conspicuity
Purpose: This study evaluates the image quality of lower extremity CT venography reconstructed with orthopedic metal artifact reduction (O-MAR) in patients with unilateral or bilateral metallic prostheses in the hip or knee. Methods: This retrospective study was approved by our institutional review board, and informed consent was waived. Twenty-nine patients of lower extremity CT with 51 metallic hip or knee prostheses were reconstructed to both standard CT images and O-MAR images. The subjective image quality and vessel conspicuity for both images were evaluated by two readers using five-point scales (0–4). Vessel conspicuity scores of 3 or 4 were considered diagnostically acceptable. Image noise was measured in the air and subcutaneous fat. Results: O-MAR images showed significantly higher scores of subjective image quality (p < .001) and vessel conspicuity (p = .002) than standard CT images. Diagnostic acceptance of vessel conspicuity was not significantly different between O-MAR images and standard CT images (p = 1.000). O-MAR images showed significantly less image noise than conventional CT images (p < .001 for both air and subcutaneous fat). Conclusion: O-MAR may be an effective solution for the metal artifacts in lower extremity CT venography; however, the distance between prostheses and vessels affects the diagnostic acceptance in patients with metallic hip or knee prostheses. Level of Evidence: Level 4, Case Series.
62 RADIOLOGY AND NUCLEAR MEDICINE↗
Effect of low tube voltage and low iodine concentration abdominal CT on image quality and radiation dose in children: preliminary study
Purpose: To evaluate the image quality of a double-low protocol (low tube voltage and low iodine concentration) for abdominal CT in children. Materials and methods: The double-low protocol was compared to the conventional protocol in pediatric patients weighing less than 40 kg from May 2016 to December 2016. Double-low protocol (Group A, n = 18): tube voltage, 70 kVp; and iodine concentration,: 250 mgI/mL versus Conventional protocol (Group B, n = 13): tube voltage, 80–100 kVp; and iodine concentration, 350 mgI/mL. Mean attenuation, noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were compared between the two groups. Image contrast, noise, beam-hardening artifacts, and overall image quality were subjectively scored. Reader performance for correctly differentiating two groups by visual assessment was evaluated. Radiation dose and total iodine load were recorded. Results: The mean attenuations of the portal vein and liver and the mean image noise in Group A were higher than in Group B (p = 0.04, 0.03, 0.004, respectively). The mean SNR and CNR of the main portal vein and liver were lower in Group A without any statistically significant difference. There were no statistically significant differences between the two groups in qualitative analysis (image contrast, image noise, and overall image quality) with substantial agreement between the reviewers (weighted kappa values; 0.59–0.76). Significantly diminished radiation dose and iodine load were observed in Group A compared with Group B (25.0%, 36.8% reduction; p = 0.007, 0.006, respectively). Conclusion: The double-low protocol was feasible for pediatric abdominal CT and reduced both radiation dose and iodine load, while maintaining image quality.
62 RADIOLOGY AND NUCLEAR MEDICINE↗
Weight-adapted ultra-low-dose pancreatic perfusion CT: radiation dose, image quality, and perfusion parameters
Purpose: We evaluate the reliability and feasibility of weight-adapted ultra-low-dose pancreatic perfusion CT. Methods: A total of 100 (47 men, 53 women) patients were enrolled prospectively and were assigned to five groups (A, B, C, D, and E) with different combination of tube voltage and tube current according to their body weight. Radiation dose parameters including volume CT dose index (CTDI) and dose-length product (DLP) were recorded. Image quality was evaluated both subjectively and objectively (noise, signal-to-noise ratio, contrast-to-noise ratio). Perfusion parameters including blood flow (BF), blood volume (BV), and permeability (PMB) were measured. The dose, image quality measurements, and perfusion parameters were compared between the five groups using one-way analysis of variance (ANOVA). Results: Radiation dose reached 8.7 mSv in patients under 50 kg and was 18.9 mSv in patients above 80 kg. The mean subjective image quality score was above 4.45 on a 5-point scale with good agreement between two radiologists. Groups A–D had equivalent performance on objective image quality (P > 0.05), while Group E performed even better (P < 0.05). No significant differences emerged in comparison with perfusion parameters (BF, BV, PMB) of normal pancreas parenchyma between the five groups. Conclusion: Weight-adapted ultra-low-dose pancreatic perfusion CT can effectively reduce radiation dose without prejudice to image quality, and the perfusion parameters of normal parenchyma are accurate and reliable.
62 RADIOLOGY AND NUCLEAR MEDICINE↗
Evaluating Machine Learning-Based MRI Reconstruction Using Digital Image Quality Phantoms
Quantitative and objective evaluation tools are essential for assessing the performance of machine learning (ML)-based magnetic resonance imaging (MRI) reconstruction methods. However, the commonly used fidelity metrics, such as mean squared error (MSE), structural similarity (SSIM), and peak signal-to-noise ratio (PSNR), often fail to capture fundamental and clinically relevant MR image quality aspects. To address this, we propose evaluation of ML-based MRI reconstruction using digital image quality phantoms and automated evaluation methods. Our phantoms are based upon the American College of Radiology (ACR) large physical phantom but created in k-space to simulate their MR images, and they can vary in object size, signal-to-noise ratio, resolution, and image contrast. Our evaluation pipeline incorporates evaluation metrics of geometric accuracy, intensity uniformity, percentage ghosting, sharpness, signal-to-noise ratio, resolution, and low-contrast detectability. We demonstrate the utility of our proposed pipeline by assessing an example ML-based reconstruction model across various training and testing scenarios. The performance results indicate that training data acquired with a lower undersampling factor and coils of larger anatomical coverage yield a better performing model. The comprehensive and standardized pipeline introduced in this study can help to facilitate a better understanding of the performance and guide future development and advancement of ML-based reconstruction algorithms.
47 OTHER INSTRUMENTATION↗
Comparison of image noise and image quality between full-dose abdominal computed tomography scans reconstructed with weighted filtered back projection and half-dose scans reconstructed with improved sinogram-affirmed iterative reconstruction (SAFIRE*)
Purpose: To retrospectively compare the image noise, signal-to-noise ratio (SNR), and subjective image quality between CT images acquired with a dual-source, split-dose imaging protocol reconstructed at full and half doses with weighted filtered back projection (wFBP) and an improved sinogram-affirmed iterative reconstruction algorithm (SAFIRE*). Methods: Fifty-three consecutive patients underwent contrast-enhanced CT of the abdomen using a standardized dual-source, single energy CT protocol. Half-dose images were retrospectively generated using data from one detector only. Full-dose datasets were reconstructed with wFBP, while half-dose datasets were reconstructed with wFBP and SAFIRE* strengths 1–5. Region of interest analysis was performed to assess SNR and noise. Diagnostic acceptability, subjective noise, and spatial resolution were graded on a 10-point scale by two readers. Statistical analysis was carried out with repeated measures analysis of variance, Wilcoxon signed rank test, and Cohen’s κ test. Results: With the increasing strengths of SAFIRE*, a progressive reduction in noise and increase in SNR (p < 0.01) was observed. There was a statistically significant decrease in objective noise and increase in SNR in half-dose SAFIRE* strength 4 and 5 reconstructions compared to full-dose reconstructions using wFBP (p < 0.01). Qualitative analysis revealed a progressive increase in diagnostic acceptability, decrease in subjective noise and increase in spatial resolution for half-dose images reconstructed with the increasing strengths of SAFIRE* (p < 0.01). Conclusions: Half-dose CT images reconstructed with SAFIRE* at strength 4 and 5 have superior image quality compared to full-dose images reconstructed with wFBP. SAFIRE* potentially allows dose reductions in the order of 50% over wFBP.
62 RADIOLOGY AND NUCLEAR MEDICINE↗
Maximizing the Radar Generalized Image Quality Equation for Bistatic SAR Using Waveform Frequency Agility
The radar generalized image quality equation (RGIQE) is a metric used to measure both monostatic and bistatic synthetic aperture radar (BSAR) image quality, it is a function of signal-to-noise ratio (SNR) and 2-D bandwidth. The 2-D bandwidth is equal to the area of the transfer function’s (TF) passband region. With the exception of side-looking monostatic geometries, almost all monostatic and bistatic geometries have skewed passband shapes when waveform frequency parameters remain unchanged from pulse to pulse. Most synthetic aperture radar (SAR) applications require a rectangular-shaped passband region, this is achieved by inscribing a rectangular region within the skewed intrinsic passband region. Increasing skewness results in less inscription area reducing 2-D bandwidth, image SNR, and thus RGIQE capacity. In this article, a waveform with frequency agility is used to rectify the skewness that degrades RGIQE capacity. By changing the waveform’s center frequency and instantaneous bandwidth from pulse to pulse in a particular manner, the intrinsic passband region can be de-skewed. The de-skewed shape maximizes the inscription area thus maximizing 2-D bandwidth, image SNR, and RGIQE capacity. Here, three examples are given in this article, one monostatic geometry, and two bistatic geometries. RGIQE capacity is increased by 52.02%, 44.42%, and 79.09% for the three examples.
47 OTHER INSTRUMENTATION↗
Quantitative Comparisons of Image Quality for Flash X-Ray Detectors
Due to X-rays’ ability to penetrate materials, flash X-ray radiography can be used for high-speed measurements where direct optical access is not possible. Choice of detector has a pronounced impact on resulting image quality. Four different detector systems were evaluated with a 450kVp flash source to quantitatively compare image quality metrics. The scintillating digital detector had less image noise than the three different storage phosphor computed radiography detectors across all transmission levels, but lacked the spatial resolution of the computed radiography detectors. For the screens tested here, the HPX-DR digital system had the highest signal to noise ratio of 68.24 and contrast to noise ratio of 35.53, but had the lowest spatial resolution, resolving 2.5 line pairs per millimeter at 1.78% contrast. At a value of 37.59, the Flex GP imaging plate had a signal to noise value above its storage phosphor counterparts under a 450kVp flash source. For radiographic setups typically used for dynamic experiments, the Flex XL Blue and Flex HR detectors had signal to noise ratios of 18.44 and 26.56 respectively. The highest resolved spatial frequencies of the Flex GP, Flex XL Blue, and Flex HR with the flash source are 3.85, 5.00, and 3.85 line pairs per millimeter, respectively. In conclusion, the Flex GP detector had the best combination of signal to noise ratio, contrast to noise ratio, and spatial resolution under a flash source.
computed radiography↗
Optical array for high-quality imaging in harsh environments
Methods and apparatus are disclosed for producing high quality images in uncontrolled or impaired environments. In some examples of the disclosed technology, groups of cameras for high dynamic range (HDR), polarization diversity, and optional other diversity modes are arranged to concurrently image a common scene. For example, in a vehicle checkpoint application, HDR provides discernment of dark objects inside a vehicle, while polarization diversity aids in rejecting glare. Spectral diversity, infrared imaging, and active illumination can be applied for better imaging through a windshield. Preprocessed single-camera images are registered and fused. Faces or other features of interest can be detected in the fused image and identified in a library. Impairments can include weather, insufficient or interfering lighting, shadows, reflections, window glass, occlusions, or moving objects.
42 ENGINEERING↗
Optical array for high-quality imaging in harsh environments
Methods and apparatus are disclosed for producing high quality images in uncontrolled or impaired environments. In some examples of the disclosed technology, groups of cameras for high dynamic range (HDR), polarization diversity, and optional other diversity modes are arranged to concurrently image a common scene. For example, in a vehicle checkpoint application, HDR provides discernment of dark objects inside a vehicle, while polarization diversity aids in rejecting glare. Spectral diversity, infrared imaging, and active illumination can be applied for better imaging through a windshield. Preprocessed single-camera images are registered and fused. Faces or other features of interest can be detected in the fused image and identified in a library. Impairments can include weather, insufficient or interfering lighting, shadows, reflections, window glass, occlusions, or moving objects.
Baba, Justin S.↗
Notes on Synthetic Aperture Radar Image Quality
Synthetic Aperture Radar (SAR) creates an image of a target scene by coherently processing radar echo returns collected along a flightpath. The quality of the SAR image is inextricably linked to the utility of the image for exploitation supporting the task at hand. Aspects of quality include the fidelity with which it can render the scene being imaged, to include the system’s Impulse Response (IPR) and underlying noise levels/characteristics. Other factors also impact utility.
47 OTHER INSTRUMENTATION↗
Application of a Metal Artifact Reduction Algorithm for C-Arm Cone-Beam CT: Impact on Image Quality and Diagnostic Confidence for Bronchial Artery Embolization
Objective: The objective of this study was to evaluate the potential benefit of a dedicated cone-beam-CT streak metal artifact removal technique (SMART) in terms of both image quality and diagnostic confidence in patients undergoing bronchial artery embolization. Methods: A total of 17 patients were included in this retrospective study. The SMART algorithm was applied to images containing streak artifacts generated by a radiopaque intra-arterial catheter tip. Quantitative evaluation of artifact severity was performed via measurement of the Hounsfield units along a closed loop surrounding the catheter tip and was conducted in the frequency domain following the application of the discrete Fourier transform to the measured data. A high proportion of power in the low frequencies of the resulting spectrum indicated a high level of streak artifacts. Qualitative evaluation of diagnostic confidence was performed using a 4-point Likert scale. Results: Both quantitative and qualitative evaluation demonstrated a significant reduction in artifact severity using the SMART algorithm. Quantitative evaluation demonstrated a mean artifact reduction of 22.5% using SMART compared to non-SMART images (p < 0.001). Qualitative evaluation demonstrated the greatest artifact reduction at the inner and outer aortic curvature, as well as immediately surrounding the tip of the catheter. In 6 of 17 cases, the use of the SMART algorithm yielded additional clinical information, increasing mean diagnostic confidence from 3.17 to 3.78 (p < 0.001). Conclusion: The SMART algorithm allows for efficient reduction of metal artifacts introduced by radiopaque catheter tips during cone-beam CT. Using this algorithm, diagnostic images of the aortic arch were significantly improved both quantitatively and qualitatively, yielding clinically relevant levels of enhanced diagnostic confidence. These results demonstrate that the SMART algorithm improves diagnostic and clinical characterization of the course of bronchial arteries on CBCT images, potentially improving the accuracy and clinical efficacy of bronchial artery embolization. Level of Evidence: 3.
62 RADIOLOGY AND NUCLEAR MEDICINE↗
Image quality indicator and dechromator setup for static LANSCE proton radiography experiment
This document describes a static geometry consisting of a sphere of gold and Ti-6Al-4V alloy, along with an energy-loss matching dechromator composed of polycarbonate, to field with the X3 electromagnetic magnifying lens at LANSCE during the next beam cycle. This static configuration will allow us to: (1) characterize the system’s resolution at the material interface between the gold and titanium alloy, and test the image quality for fine features associated with the screws used to fasten the titanium alloy hemispheres; (2) tune our models of proton scattering and imaging for these materials; and (3) test a procedure to align the polycarbonate dechromator and the sphere within the proton beamline. The data will also be suitable for validation of the software options available for proton radiography simulations.
46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
Does time-of-flight improve image quality in the heart?
No abstract prepared.
62 RADIOLOGY AND NUCLEAR MEDICINE↗
Geometrical-Based Generative Adversarial Network to Enhance Digital Rock Image Quality
X-ray microcomputed tomography (micro-CT) is a common tool for the study of porous media structures and properties. High-quality micro-CT data are required to accurately capture pore structures. Acquiring high-quality micro-CT data, however, is not always possible, owing to application limitations and experimental constraints. Therefore, we propose a geometrical-based generative adversarial network (GAN) to rapidly restore noisy micro-CT images to their clean counterparts. The training data and related ground-truth (GT) data are scanned for 7 min and 9.5 h, respectively. To evaluate the performance of the geometrical-based GAN, a 6003 voxel image that has never been used for training is reconstructed and compared with the corresponding GT image. Histogram matching and linear normalization are implemented to adjust the histogram of the reconstructed image to that of the GT image. A watershed-based segmentation method is then applied to delineate pore and solid phases. Lastly, we measure the Minkowski functionals and petrophysical properties, including absolute permeability, pore size distribution, drainage capillary pressure-saturation curve, and imbibition relative permeability, to estimate the physical accuracy of the denoised image. The results show that the proposed geometrical-based GAN can accurately restore noisy micro-CT data. By reducing the scanning time from 9.5 h to 7 min, the expenditure of collecting micro-CT can be decreased significantly. This is particularly important for applications where time-lapse images of a dynamic process are required, high-throughput imaging is necessary for real-time data analysis, or where the quantification of large sample volumes is required.
58 GEOSCIENCES↗
Deep learning accelerated topology optimization with inherent control of image quality
Not provided.
Computer Science↗