Extremity radiographs derived from low-dose ultra-high-resolution CT: a phantom study.

OBJECTIVE: To demonstrate the potential of low-dose ultra-high-resolution CT (UHRCT) images to generate high-quality radiographic images on extremity phantoms and to estimate the radiation dose required for this. MATERIALS AND METHODS: A hand and knee phantom containing real human bones was imaged on an UHRCT scanner at full-dose, half-dose, and quarter-dose levels using a high-resolution extremity protocol. The raw data was reconstructed using both filtered back projection (FBP) and an iterative reconstruction algorithm (AIDR3D). Using custom designed software, each CT volume data set was converted to attenuation coefficients, and then a synthesized radiograph (synDX) was generated by forward projecting the volume data sets from a point source onto a 2D synthetic detector. The signal-to-noise ratio (SNR) was measured in the synDXs across all dose levels and the root-mean-squared error (RMSE) was computed with the FD synDXs as the reference. RESULTS: The proposed workflow generates high-quality synDXs at any arbitrary angle. For FBP, the SNR largely tracked with the radiation dose levels for both the knee and hand phantoms. For the knee phantom, iterative reconstruction provided a 6.1% higher SNR when compared to FBP. The RMSE was overall higher for the lowest dose levels and monotonically decreased with increasing dose. No substantial differences were observed qualitatively in the visualization of skeletal detail of the phantoms. CONCLUSION: The fine detail provided by UHRCT acquisitions of extremities facilitates the ability to generate quality radiographs, potentially eliminating the need for additional scanning on a conventional digital radiography system.

CT Phantom Evaluation of 67,392 American College of Radiology Accreditation Examinations: Implications for Opportunistic Screening of Osteoporosis Using CT.

OBJECTIVE. The purpose of this study was to investigate whether systematic bias in attenuation measurements occurs among CT scanners made by four major manufacturers and the relevance of this bias regarding opportunistic screening for osteoporosis. MATERIALS AND METHODS. Data on attenuation measurement accuracy were acquired using the American College of Radiology (ACR) accreditation phantom and were evaluated in a blinded fashion for four CT manufacturers (8500 accreditation submissions for manufacturer A; 18,575 for manufacturer B; 8278 for manufacturer C; and 32,039 for manufacturer D). The attenuation value for water, acrylic (surrogate for trabecular bone), and Teflon (surrogate for cortical bone; Chemours) materials for an adult abdominal CT technique (120 kV, 240 mA, standard reconstruction algorithm) was used in the analysis. Differences in attenuation value across all manufacturers were assessed using the Kruskal-Wallis test followed by a post hoc test for pairwise comparisons. RESULTS. The mean attenuation value for water ranged from -0.3 to 2.7 HU, with highly significant differences among all manufacturers (p < 0.001). For the trabecular bone surrogate, differences in attenuation values across all manufacturers were also highly significant (p < 0.001), with mean values of 120.9 (SD, 3.5), 124.6 (3.3), 126.9 (4.4), and 123.9 (3.4) HU for manufacturers A, B, C, and D, respectively. For the cortical bone surrogate, differences in attenuation values across all manufacturers were also highly significant (p < 0.001), with mean values of 939.0 (14.2), 874.3 (13.3), 897.6 (11.3), and 912.7 (13.4) HU for manufacturers A, B, C, and D, respectively. CONCLUSION. CT scanners made by different manufacturers show systematic offsets in attenuation measurement when compared with each other. Knowledge of these off-sets is useful for optimizing the accuracy of opportunistic diagnosis of osteoporosis.

Opportunistic Screening for Osteoporosis Using Computed Tomography: State of the Art and Argument for Paradigm Shift.

PURPOSE OF REVIEW: Osteoporosis is disproportionately common in rheumatology patients. For the past three decades, the diagnosis of osteoporosis has benefited from well-established practice guidelines that emphasized the use of dual x-ray absorptiometry (DXA). Despite these guidelines and the wide availability of DXA, approximately two thirds of eligible patients do not undergo testing. One strategy to improve osteoporosis testing is to employ computed tomography (CT) examinations obtained as part of routine patient care to "opportunistically" screen for osteoporosis, without additional cost or radiation exposure to patients. This review examines the role of opportunistic CT in the evaluation of osteoporosis. RECENT FINDINGS: Recent evidence suggests that opportunistic measurement of bone attenuation (radiodensity) using CT has sensitivity comparable to DXA. More importantly, such an approach has been shown to predict osteoporotic fractures. The paradigm shift of using CTs obtained for other reasons to opportunistically screen for osteoporosis promises to substantially improve patient care.

Predictors of CT Radiation Dose and Their Effect on Patient Care: A Comprehensive Analysis Using Automated Data.

Purpose To determine patient, vendor, and institutional factors that influence computed tomography (CT) radiation dose. Materials and Methods The relevant institutional review boards approved this HIPAA-compliant study, with waiver of informed consent. Volume CT dose index (CTDIvol) and effective dose in 274 124 head, chest, and abdominal CT examinations performed in adult patients at 12 facilities in 2013 were collected prospectively. Patient, vendor, and institutional characteristics that could be used to predict (a) median dose by using linear regression after log transformation of doses and (b) high-dose examinations (top 25% of dose within anatomic strata) by using modified Poisson regression were assessed. Results There was wide variation in dose within and across medical centers. For chest CTDIvol, overall median dose across all institutions was 11 mGy, and institutional median dose was 7-16 mGy. Models including patient, vendor, and institutional factors were good for prediction of median doses (R2 = 0.31-0.61). The specific institution where the examination was performed (reflecting the specific protocols used) accounted for a moderate to large proportion of dose variation. For chest CTDIvol, unadjusted median CTDIvol was 16.5 mGy at one institution and 6.7 mGy at another (adjusted relative median dose, 2.6 mGy [95% confidence interval: 2.5, 2.7]). Several variables were important predictors that a patient would undergo high-dose CT. These included patient size, the specific institution where CT was performed, and the use of multiphase scanning. For example, while 49% of patients (21 411 of 43 696) who underwent multiphase abdominal CT had a high-dose examination, 8% of patients (4977 of 62 212) who underwent single-phase CT had a high-dose examination (adjusted relative risk, 6.20 [95% CI: 6.17, 6.23]). If all patients had been examined with single-phase CT, 69% (18 208 of 26 388) of high-dose examinations would have been eliminated. Patient size, institutional-specific protocols, and multiphase scanning were the most important predictors of dose (change in R2 = 8%-32%), followed by manufacturer and iterative reconstruction (change in R2, 0.2%-15.0%). Conclusion CT doses vary considerably within and across facilities. The primary factors that influenced dose variation were multiphase scanning and institutional protocol choices. It is unknown if the variation in these factors influenced diagnostic accuracy. © RSNA, 2016.

Phantom-based standardization of CT angiography images for spot sign detection.

PURPOSE: The CT angiography (CTA) spot sign is a strong predictor of hematoma expansion in intracerebral hemorrhage (ICH). However, CTA parameters vary widely across centers and may negatively impact spot sign accuracy in predicting ICH expansion. We developed a CT iodine calibration phantom that was scanned at different institutions in a large multicenter ICH clinical trial to determine the effect of image standardization on spot sign detection and performance. METHODS: A custom phantom containing known concentrations of iodine was designed and scanned using the stroke CT protocol at each institution. Custom software was developed to read the CT volume datasets and calculate the Hounsfield unit as a function of iodine concentration for each phantom scan. CTA images obtained within 8 h from symptom onset were analyzed by two trained readers comparing the calibrated vs. uncalibrated density cutoffs for spot sign identification. ICH expansion was defined as hematoma volume growth >33%. RESULTS: A total of 90 subjects qualified for the study, of whom 17/83 (20.5%) experienced ICH expansion. The number of spot sign positive scans was higher in the calibrated analysis (67.8 vs 38.9% p < 0.001). All spot signs identified in the non-calibrated analysis remained positive after calibration. Calibrated CTA images had higher sensitivity for ICH expansion (76 vs 52%) but inferior specificity (35 vs 63%) compared with uncalibrated images. CONCLUSION: Normalization of CTA images using phantom data is a feasible strategy to obtain consistent image quantification for spot sign analysis across different sites and may improve sensitivity for identification of ICH expansion.

JOURNAL CLUB: Quantification of Fetal Dose Reduction if Abdominal CT Is Limited to the Top of the Iliac Crests in Pregnant Patients With Trauma.

OBJECTIVE: The purposes of this study were to correlate fetal z-axis location within the maternal abdomen on CT with gestational age and estimate fetal dose reduction of a study limited to the abdomen only, with its lower aspect at the top of the iliac crests, compared with full abdominopelvic CT in pregnant trauma patients. MATERIALS AND METHODS: We performed a study of pregnant patients who underwent CT of the abdomen and pelvis for trauma at a single institution over a 10-year period. The inferior aspect of maternal liver, spleen, gallbladder, pancreas, adrenals, and kidneys was recorded as above or below the iliac crests. The distance from the iliac crest to the top of the fetus or gestational sac was determined. The CT images of the limited and full scanning studies were independently reviewed by two blinded radiologists to identify traumatic injuries. Fetal dose profiles, including both scatter and primary radiation, were computed analytically along the central axis of the patient to estimate fetal dose reduction. Linear regression analysis was performed between gestational age and distance of the fetus to the iliac crests. RESULTS: Thirty-five patients were included (mean age, 26.2 years). Gestational age ranged from 5 to 38 weeks, with 5, 19, and 11 gestations in the first, second, and third trimesters, respectively. All solid organs were above the iliac crests in all patients. In three of six patients, traumatic findings in the pelvis would have been missed with the limited study. There was high correlation between gestational age and distance of the fetus to the iliac crests (R(2) = 0.84). The mean gestational age at which the top of the fetus was at the iliac crest was 17.3 weeks. Using the limited scanning study, fetuses at 5, 20, and 40 weeks of gestation would receive an estimated 4.3%, 26.2%, and 59.9% of the dose, respectively, compared with the dose for the full scanning study. CONCLUSION: In pregnant patients in our series with a history of trauma, CT of the abdomen only was an effective technique to reduce fetal radiation exposure compared with full abdomen and pelvis CT.

Radiation Doses in Consecutive CT Examinations from Five University of California Medical Centers.

PURPOSE: To summarize data on computed tomographic (CT) radiation doses collected from consecutive CT examinations performed at 12 facilities that can contribute to the creation of reference levels. MATERIALS AND METHODS: The study was approved by the institutional review boards of the collaborating institutions and was compliant with HIPAA. Radiation dose metrics were prospectively and electronically collected from 199 656 consecutive CT examinations in 83 181 adults and 3871 consecutive CT examinations in 2609 children at the five University of California medical centers during 2013. The median volume CT dose index (CTDIvol), dose-length product (DLP), and effective dose, along with the interquartile range (IQR), were calculated separately for adults and children and stratified according to anatomic region. Distributions for DLP and effective dose are reported for single-phase examinations, multiphase examinations, and all examinations. RESULTS: For adults, the median CTDIvol was 50 mGy (IQR, 37-62 mGy) for the head, 12 mGy (IQR, 7-17 mGy) for the chest, and 12 mGy (IQR, 8-17 mGy) for the abdomen. The median DLPs for single-phase, multiphase, and all examinations, respectively, were as follows: head, 880 mGy · cm (IQR, 640-1120 mGy · cm), 1550 mGy · cm (IQR, 1150-2130 mGy · cm), and 960 mGy · cm (IQR, 690-1300 mGy · cm); chest, 420 mGy · cm (IQR, 260-610 mGy · cm), 880 mGy · cm (IQR, 570-1430 mGy · cm), and 550 mGy · cm (IQR 320-830 mGy · cm); and abdomen, 580 mGy · cm (IQR, 360-860 mGy · cm), 1220 mGy · cm (IQR, 850-1790 mGy · cm), and 960 mGy · cm (IQR, 600-1460 mGy · cm). Median effective doses for single-phase, multiphase, and all examinations, respectively, were as follows: head, 2 mSv (IQR, 1-3 mSv), 4 mSv (IQR, 3-8 mSv), and 2 mSv (IQR, 2-3 mSv); chest, 9 mSv (IQR, 5-13 mSv), 18 mSv (IQR, 12-29 mSv), and 11 mSv (IQR, 6-18 mSv); and abdomen, 10 mSv (IQR, 6-16 mSv), 22 mSv (IQR, 15-32 mSv), and 17 mSv (IQR, 11-26 mSv). In general, values for children were approximately 50% those for adults in the head and 25% those for adults in the chest and abdomen. CONCLUSION: These summary dose data provide a starting point for institutional evaluation of CT radiation doses.

CT Hounsfield numbers of soft tissues on unenhanced abdominal CT scans: variability between two different manufacturers' MDCT scanners.

OBJECTIVE: The purpose of this study is to determine whether Hounsfield numbers of soft tissues on unenhanced abdominal CT of the same patient vary on repeat scans done on two different manufacturers' MDCT scanners. MATERIALS AND METHODS: A database search was performed to identify patients older than 18 years who underwent unenhanced CT of the abdomen and pelvis performed both on a Volume CT (GE Healthcare) and a Definition AS Plus (Siemens Healthcare) 64-MDCT scanner within 12 months of each other. After excluding those patients for whom Hounsfield unit measurements would be affected by mitigating factors, 48 patients (mean age, 58.8 years) were identified. Hounsfield unit measurements were obtained in nine different soft-tissue anatomic locations on each scan, and the location of these sites was kept identical on each scan pair. Data were analyzed to evaluate Hounsfield unit differences between these scanners. RESULTS: In general, there was a low consistency in the Hounsfield unit measurements for each of these sites on scans obtained by the two scanners, with the subcutaneous fat in the left posterolateral flank showing the lowest correlation (intraclass correlation coefficient, 0.198). There were differences in the Hounsfield unit measurements obtained in all anatomic sites on scans obtained by both scanners. Mean Hounsfield unit measurements obtained on the Definition AS Plus scanner were lower than those obtained on the Volume CT scanner, with the intriguing exception of the anterior midline subcutaneous fat Hounsfield unit measurements, which were higher on the Definition AS Plus scanner. All differences were statistically significant (p < 0.05). CONCLUSION: Hounsfield unit measurements for unenhanced abdominal soft tissues of the same patient vary between scanners of two common MDCT manufacturers.

Level set segmentation of breast masses in contrast-enhanced dedicated breast CT and evaluation of stopping criteria.

Dedicated breast CT (bCT) produces high-resolution 3D tomographic images of the breast, fully resolving fibroglandular tissue structures within the breast and allowing for breast lesion detection and assessment in 3D. In order to enable quantitative analysis, such as volumetrics, automated lesion segmentation on bCT is highly desirable. In addition, accurate output from CAD (computer-aided detection/diagnosis) methods depends on sufficient segmentation of lesions. Thus, in this study, we present a 3D lesion segmentation method for breast masses in contrast-enhanced bCT images. The segmentation algorithm follows a two-step approach. First, 3D radial-gradient index segmentation is used to obtain a crude initial contour, which is then refined by a 3D level set-based active contour algorithm. The data set included contrast-enhanced bCT images from 33 patients containing 38 masses (25 malignant, 13 benign). The mass centers served as input to the algorithm. In this study, three criteria for stopping the contour evolution were compared, based on (1) the change of region volume, (2) the average intensity in the segmented region increase at each iteration, and (3) the rate of change of the average intensity inside and outside the segmented region. Lesion segmentation was evaluated by computing the overlap ratio between computer segmentations and manually drawn lesion outlines. For each lesion, the overlap ratio was averaged across coronal, sagittal, and axial planes. The average overlap ratios for the three stopping criteria ranged from 0.66 to 0.68 (dice coefficient of 0.80 to 0.81), indicating that the proposed segmentation procedure is promising for use in quantitative dedicated bCT analyses.

Microcalcification detectability in breast CT images using CNN observers.

BACKGROUND: Breast computed tomography (CT) is an emerging breast imaging modality, and ongoing developments aim to improve breast CT's ability to detect microcalcifications. To understand the effects of different parameters on microcalcification detectability, a virtual clinical trial study was conducted using hybrid images and convolutional neural network (CNN)-based model observers. Mathematically generated microcalcifications were embedded into breast CT data sets acquired at our institution, and parameters related to calcification size, calcification contrast, cluster diameter, cluster density, and image display method (i.e., single slices, slice averaging, and maximum-intensity projections) were evaluated for their influence on microcalcification detectability. PURPOSE: To investigate the individual effects and the interplay of parameters affecting microcalcification detectability in breast CT. METHODS: Spherical microcalcifications of varying diameters (0.04, 0.10, 0.15, 0.20, 0.25, 0.30, 0.35, 0.40 mm) and native intensities were computer simulated to portray the partial volume effects of the imaging system. Calcifications were mathematically embedded into 109 patient breast CT volume data sets as individual calcifications or as clusters of calcifications. Six numbers of calcifications (1, 3, 5, 7, 10, 15) distributed within six cluster diameters (1, 3, 5, 6, 8, 10 mm) were simulated to study the effect of cluster density. To study the role of image display method, 2D regions of interest (ROIs) and 3D volumes of interest (VOIs) were generated using single slice extraction, slice averaging, and maximum-intensity projection (MIP). 2D and 3D CNNs were trained on the ROIs and VOIs, and receiver operating characteristic (ROC) curve analysis was used to evaluate detection performance. The area under the ROC curve (AUC) was used as the primary performance metric. RESULTS: Detection performance decreased with increasing section thickness, and peak detection performance occurred using the native section thickness (0.2 mm) and MIP display. The MIP display method, despite using a single slice, yielded comparable performance to the native section thickness, which employed 50 slices. Reduction in slices did not sacrifice detection accuracy and provided significant computational advantages over multi-slice image volumes. Larger cluster diameters resulted in reduced overall detectability, while smaller cluster diameters led to increased detectability. Additionally, we observed that the presence of more calcifications within a cluster improved the overall detectability, while fewer calcifications decreased it. CONCLUSIONS: As breast CT is still a relatively new breast imaging modality, there is an ongoing need to identify optimal imaging protocols. This work demonstrated the utility of MIP presentation for displaying image volumes containing microcalcification clusters. It is likely that human observers may also benefit from viewing MIPs compared to individual slices. The results of this investigation begin to elucidate how model observers interact with microcalcification clusters in a 3D volume, and will be useful for future studies investigating a broader set of parameters related to breast CT.

Objective image quality assurance in cone-beam CT: Test methods, analysis, and workflow in longitudinal studies.

BACKGROUND: Standards for image quality evaluation in multi-detector CT (MDCT) and cone-beam CT (CBCT) are evolving to keep pace with technological advances. A clear need is emerging for methods that facilitate rigorous quality assurance (QA) with up-to-date metrology and streamlined workflow suitable to a range of MDCT and CBCT systems. PURPOSE: To evaluate the feasibility and workflow associated with image quality (IQ) assessment in longitudinal studies for MDCT and CBCT with a single test phantom and semiautomated analysis of objective, quantitative IQ metrology. METHODS: A test phantom (CorgiTM Phantom, The Phantom Lab, Greenwich, New York, USA) was used in monthly IQ testing over the course of 1 year for three MDCT scanners (one of which presented helical and volumetric scan modes) and four CBCT scanners. Semiautomated software analyzed image uniformity, linearity, contrast, noise, contrast-to-noise ratio (CNR), 3D noise-power spectrum (NPS), modulation transfer function (MTF) in axial and oblique directions, and cone-beam artifact magnitude. The workflow was evaluated using methods adapted from systems/industrial engineering, including value stream process modeling (VSPM), standard work layout (SWL), and standard work control charts (SWCT) to quantify and optimize test methodology in routine practice. The completeness and consistency of DICOM data from each system was also evaluated. RESULTS: Quantitative IQ metrology provided valuable insight in longitudinal quality assurance (QA), with metrics such as NPS and MTF providing insight on root cause for various forms of system failure-for example, detector calibration and geometric calibration. Monthly constancy testing showed variations in IQ test metrics owing to system performance as well as phantom setup and provided initial estimates of upper and lower control limits appropriate to QA action levels. Rigorous evaluation of QA workflow identified methods to reduce total cycle time to ∼10 min for each system-viz., use of a single phantom configuration appropriate to all scanners and Head or Body scan protocols. Numerous gaps in the completeness and consistency of DICOM data were observed for CBCT systems. CONCLUSION: An IQ phantom and test methodology was found to be suitable to QA of MDCT and CBCT systems with streamlined workflow appropriate to busy clinical settings.

Correcting for asymmetry of the proximal tibial epiphysis is warranted to determine postoperative alignment deviations in kinematic alignment from planned alignment of the tibial component on the native tibia.

BACKGROUND: In total knee arthroplasty, unrestricted kinematic alignment aims to restore pre-arthritic lower limb alignment and joint lines. Joint line orientations of the contralateral healthy proximal tibia might be used to evaluate accuracy of tibial component alignment post-operatively if asymmetry is minimal. Our objective was to evaluate left-to-right asymmetry of the proximal tibial epiphysis in posterior tibial slope and varus-valgus orientation as related to unrestricted kinematic alignment principles. METHODS: High resolution CT images (0.5 mm slice thickness) were acquired from bilateral lower limbs of 11 skeletally mature subjects with no skeletal abnormalities. Images were segmented to generate 3D tibia models. Asymmetry was quantified by differences in orientations required to shape-match the proximal epiphysis of the mirror 3D tibia model to the proximal epiphysis of the contralateral 3D tibia model. FINDINGS: Systematic and random differences (i.e. mean ± standard deviation) in tibial slope and varus-valgus orientation were - 0.8° ± 1.2° and - 0.2° ± 0.8°, respectively. Ninety five percent confidence intervals on the means included 0° indicating that systematic differences were minimal. INTERPRETATION: Since random differences due to asymmetry are substantial in relation to random surgical deviations from pre-arthritic joint lines previously reported, post-operative computer tomograms of the contralateral healthy tibia should not be used to directly assess accuracy of tibial component alignment on a group level without correcting for differences in tibial slope and varus-valgus orientation due to asymmetry.

Joint AAPM Task Group 282/EFOMP Working Group Report: Breast dosimetry for standard and contrast-enhanced mammography and breast tomosynthesis.

Currently, there are multiple breast dosimetry estimation methods for mammography and its variants in use throughout the world. This fact alone introduces uncertainty, since it is often impossible to distinguish which model is internally used by a specific imaging system. In addition, all current models are hampered by various limitations, in terms of overly simplified models of the breast and its composition, as well as simplistic models of the imaging system. Many of these simplifications were necessary, for the most part, due to the need to limit the computational cost of obtaining the required dose conversion coefficients decades ago, when these models were first implemented. With the advancements in computational power, and to address most of the known limitations of previous breast dosimetry methods, a new breast dosimetry method, based on new breast models, has been developed, implemented, and tested. This model, developed jointly by the American Association of Physicists in Medicine and the European Federation for Organizations of Medical Physics, is applicable to standard mammography, digital breast tomosynthesis, and their contrast-enhanced variants. In addition, it includes models of the breast in both the cranio-caudal and the medio-lateral oblique views. Special emphasis was placed on the breast and system models used being based on evidence, either by analysis of large sets of patient data or by performing measurements on imaging devices from a range of manufacturers. Due to the vast number of dose conversion coefficients resulting from the developed model, and the relative complexity of the calculations needed to apply it, a software program has been made available for download or online use, free of charge, to apply the developed breast dosimetry method. The program is available for download or it can be used directly online. A separate User's Guide is provided with the software.

Clogging evaluation of open graded friction course pavements tested under rainfall and heavy vehicle simulators.

In this study a new procedure is developed to obtain core samples from field sections to assess clogging mechanisms of open graded friction course (OGFC) pavements using X-ray computed tomography (CT) imaging. The approach compared X-ray computed tomography (CT) images taken before and after: (1) rainfall simulations without trafficking to investigate particle-related clogging and (2) full-scale accelerated pavement rutting tests (APT) to investigate deformation related clogging of OGFC layers. Rainfall simulations were performed with runoff water of known total suspended solids (TSS) and particle size distributions (PSDs). Full-scale accelerated rutting tests were performed under controlled temperature and loads. Both investigations were performed for three different OGFC pavements with different layer thicknesses and mix types. The clogging of rutting test sections were also evaluated by comparing the surface permeability measurements performed before and after APT testing. The results of X-ray CT image processing revealed a significant reduction in air-void content of core samples after APT rutting tests. The highest air-void reduction was concentrated at the bottom of the OGFC layers. Permeability measurements also showed a 40%-90% reduction in permeability after APT trafficking. X-ray CT image processing of core samples tested under simulated rainfall showed that air void content reduction is concentrated in the lower part (2-6 mm from the bottom) of the OGFC layers as a result of particle accumulation. Small changes in air void contents were observed in the upper part of the OGFC layers (10-15 mm) while these reductions in air void contents were not significant to cause surface overflow and hence it is expected that the tested OGFC pavements will have sufficient permeability to infiltrate water during most average storm events.

Pre-whitened matched filter and convolutional neural network based model observer performance for mass lesion detection in non-contrast breast CT.

BACKGROUND: Mathematical model observers have been shown to reasonably predict human observer performance and are useful when human observer studies are infeasible. Recently, convolutional neural networks (CNNs) have also been used as substitutes for human observers, and studies have shown their utility as an optimal observer. In this study, a CNN model observer is compared to the pre-whitened matched filter (PWMF) model observer in detecting simulated mass lesions inserted into 253 acquired breast computed tomography (bCT) images from patients imaged at our institution. PURPOSE: To compare CNN and PWMF model observers for detecting signal-known-exactly (SKE) location-known-exactly (LKE) simulated lesions in bCT images with real anatomical backgrounds, and to use these model observers collectively to optimize parameters and understand trends in performance with breast CT. METHODS: Spherical lesions with different diameters (1, 3, 5, 9 mm) were mathematically inserted into reconstructed patient bCT image data sets to mimic 3D mass lesions in the breast. 2D images were generated by extracting the center slice along the axial dimension or by slice averaging across adjacent slices to model thicker sections (0.4, 1.2, 2.0, 6.0, 12.4, 20.4 mm). The role of breast density was retrospectively studied using the range of breast densities intrinsic to the patient bCT data sets. In addition, mass lesions were mathematically inserted into Gaussian images matched to the mean and noise power spectrum of the bCT images to better understand the performance of the CNN in the context of a known ideal observer (the PWMF). The simulated Gaussian and bCT images were divided into training and testing data sets. Each training data set consisted of 91 600 images, and each testing data set consisted of 96 000 images. A CNN and PWMF was trained on the Gaussian training images, and a different CNN and PWMF was trained on the bCT training images. The trained model observers were tested, and receiver operating characteristic (ROC) curve analysis was used to evaluate detection performance. The area under the ROC curve (AUC) was the primary performance metric used to compare the model observers. RESULTS: In the Gaussian background, the CNN performed essentially identically to the PWMF across lesion sizes and section thicknesses. In the bCT background, the CNN outperformed the PWMF across lesion size, breast density, and most section thicknesses. These findings suggest that there are higher-order features in bCT images that are harnessed by the CNN observer but are inaccessible to the PWMF. CONCLUSIONS: The CNN performed equivalently to the ideal observer in Gaussian textures. In bCT background, the CNN captures more diagnostic information than the PWMF and may be a more pertinent observer when conducting optimal performance studies in breast CT images.

Cone-beam CT sampling incompleteness: analytical and empirical studies of emerging systems and source-detector orbits.

Purpose: Motivated by emerging cone-beam computed tomography (CBCT) systems and scan orbits, we aim to quantitatively assess the completeness of data for 3D image reconstruction-in turn, related to "cone-beam artifacts." Fundamental principles of cone-beam sampling incompleteness are considered with respect to an analytical figure-of-merit [FOM, denoted tan(ψmin)] and related to an empirical FOM (denoted zmod) for measurement of cone-beam artifact magnitude in a test phantom. Approach: A previously proposed analytical FOM [tan(ψmin), defined as the minimum angle between a point in the 3D image reconstruction and the x-ray source over the scan orbit] was analyzed for a variety of CBCT geometries. A physical test phantom was configured with parallel disk pairs (perpendicular to the z-axis) at various locations throughout the field of view, quantifying cone-beam artifact magnitude in terms of zmod (the relative signal modulation between the disks). Two CBCT systems were considered: an interventional C-arm (Cios Spin 3D; Siemens Healthineers, Forcheim Germany) and a musculoskeletal extremity scanner; Onsight3D, Carestream Health, Rochester, United States)]. Simulations and physical experiments were conducted for various source-detector orbits: (a) a conventional 360 deg circular orbit, (b) tilted and untilted semi-circular (196 deg) orbits, (c) multi-source (three x-ray sources distributed along the z axis) semi-circular orbits, and (d) a non-circular (sine-on-sphere, SoS) orbit. The incompleteness of sampling [tan(ψmin)] and magnitude of cone-beam artifacts (zmod) were evaluated for each system and orbit. Results: The results show visually and quantitatively the effect of system geometry and scan orbit on cone-beam sampling effects, demonstrating the relationship between analytical tan(ψmin) and empirical zmod. Advanced source-detector orbits (e.g., three-source and SoS orbits) exhibited superior sampling completeness as quantified by both the analytical and the empirical FOMs. The test phantom and zmod metric were sensitive to variations in CBCT system geometry and scan orbit and provided a surrogate measure of underlying sampling completeness. Conclusion: For a given system geometry and source-detector orbit, cone-beam sampling completeness can be quantified analytically (in terms arising from Tuy's condition) and/or empirically (using a test phantom for quantification of cone-beam artifacts). Such analysis provides theoretical and practical insight on sampling effects and the completeness of data for emerging CBCT systems and scan trajectories.

Model observer performance in contrast-enhanced lesions in breast CT: The influence of contrast concentration on detectability.

BACKGROUND: The use of iodine-based contrast agent for better delineation of tumors in breast CT (bCT) has been shown to be compelling, similar to the tumor enhancement in contrast-enhanced breast MRI. Contrast-enhanced bCT (CE-bCT) is a relatively new tool, and a structured evaluation of different imaging parameters at play has yet to be conducted. In this investigation, data sets of acquired bCT images from 253 patients imaged at our institution were used in concert with simulated mathematically inserted spherical contrast-enhanced lesions to study the role of contrast enhancement on detectability. PURPOSE: To quantitatively evaluate the improvement in lesion detectability due to contrast enhancement across lesion diameter, section thickness, view plane, and breast density using a pre-whitened matched filter (PWMF) model observer. METHODS: The relationship between iodine concentration and Hounsfield units (HU) was measured using spectral modeling. The lesion enhancement from clinical CE-bCT images in 22 patients was evaluated, and the average contrast enhancement (ΔHU) was determined. Mathematically generated spherical mass lesions of varying diameters (1, 3, 5, 9, 11, 15 mm) and contrast enhancement levels (0, 0.25, 0.50, 0.75, 1) were inserted at random locations in 253 actual patient bCT datasets. Images with varying thicknesses (0.4-19.8 mm) were generated by slice averaging, and the role of view plane (coronal and axial planes) was studied. A PWMF was used to generate receiver operating characteristic (ROC) curves across parameters of lesion diameter, contrast enhancement, section thickness, view plane, and breast density. The area under the ROC curve (AUC) was used as the primary performance metric, generated from over 90,000 simulated lesions. RESULTS: An average 20% improvement (ΔAUC = 0.1) in lesion detectability due to contrast enhancement was observed across lesion diameter, section thickness, breast density, and view plane. A larger improvement was observed when stratifying patients based on breast density. For patients with VGF ≤ 40%, detection performance improved up to 20% (until AUC →1), and for patients with denser breasts (VGF > 40%), detection performance improved more drastically, ranging from 20% to 80% for 1- and 5-mm lesions. For the 1 mm lesion, detection performance raised slightly at the 1.2 mm section thickness before falling off as thickness increased. For larger lesions, detection performance was generally unaffected as section thickness increased up until it reached 5.8 mm, where performance began to decline. Detection performance was higher in the axial plane compared to the coronal plane for smaller lesions and thicker sections. CONCLUSIONS: For emerging diagnostic tools like CE-bCT, it is important to optimize imaging protocols for lesion detection. In this study, we found that intravenous contrast can be used to detect small lesions in dense breasts. Optimal section thickness for detectability has dependencies on breast density and lesion size, therefore, display thickness should be adjusted in real-time using display software. These findings may be useful for the development of CE-bCT as well as other x-ray-based breast imaging modalities.

Performance of high-resolution CT for detection and discrimination tasks related to stenotic lesions - A phantom study using model observers.

BACKGROUND: Accurate detection and grading of atheromatous stenotic lesions within the cardiac, renal, and intracranial vasculature is imperative for early recognition of disease and guiding treatment strategies. PURPOSE: In this work, a stenotic lesion phantom was used to compare high resolution and normal resolution modes on the same CT scanner in terms of detection and size discrimination performance. MATERIALS AND METHODS: The phantom is comprised of three acrylic cylinders (each 15.0 cm in diameter and 1.3 cm thick) with a matching array of holes in each module. The outer two modules contain holes that are slightly larger than the corresponding hole in the central module to simulate stenotic narrowing in vasculature. The stack of modules was submerged in an iodine solution simulating contrast-enhanced stenotic lesions with a range of lumen diameters (1.32-10.08 mm) and stenosis severity (0%, 50%, 60%, 70%, and 80%). The phantom was imaged on the Canon Aquilion Precision high-resolution CT scanner in high-resolution (HR) mode (0.25 mm × 0.50 mm detector element size) and normal-resolution (NR) mode (0.50 mm × 0.50 mm) using 120 kV and two dose levels (14 and 21 mGy SSDE) with 30 repeat scans acquired for each combination. Filtered back-projection (FBP) and a hybrid-iterative reconstruction (AIDR) were used with the FC18 kernel, as well as a deep learning algorithm (AiCE) which is only available for HR. A non-prewhitening model observer with an eye filter was implemented to quantify performance for detection and size discrimination tasks in the axial plane. RESULTS: Detection performance improved with increasing diameter, dose, and for AIDR in comparison to FBP for a fixed resolution mode. Performance in the HR mode was generally higher than NR for the smaller lumen diameters (1-5 mm) with decreasing differences as the diameter increased. Performance in NR mode surpassed HR mode for lumen diameters greater than ∼4 mm and ∼5 mm for 14 mGy and 21 mGy, respectively. AiCE provided consistently higher detection performance compared with AIDR-FC18 (48% higher for a 6 mm lumen diameter). Discrimination performance increased with increasing nominal diameter, dose, and for larger differences in stenosis severity. When comparing discrimination performance in HR to NR modes, the largest relative differences occur at the smallest nominal diameters and smallest differences in stenosis severity. The AiCE reconstruction algorithm produced the highest overall discrimination performance values, and these were significantly higher than AIDR-FC18 for nominal diameters of 7.14 and 10.08 mm. CONCLUSIONS: HR mode outperforms NR for detection up to a specific diameter and the results improve with AiCE and for higher dose levels. For the task of size discrimination, HR mode consistently outperforms NR if AIDR-FC18 is used for dose levels of at least 21 mGy, and the results improve with AiCE and for the smallest differences in stenosis severity investigated (50% vs. 60%). High-resolution CT appears to be beneficial for detecting smaller simulated lumen diameters (<5 mm) and is generally advantageous for discrimination tasks related to stenotic lesions, which inherently contain information at higher frequencies, given the right reconstruction algorithm and dose level.

Deep learning for x-ray scatter correction in dedicated breast CT.

BACKGROUND: Accurate correction of x-ray scatter in dedicated breast computed tomography (bCT) imaging may result in improved visual interpretation and is crucial to achieve quantitative accuracy during image reconstruction and analysis. PURPOSE: To develop a deep learning (DL) model to correct for x-ray scatter in bCT projection images. METHODS: A total of 115 patient scans acquired with a bCT clinical system were segmented into the major breast tissue types (skin, adipose, and fibroglandular tissue). The resulting breast phantoms were divided into training (n = 110) and internal validation cohort (n = 5). Training phantoms were augmented by a factor of four by random translation of the breast in the image field of view. Using a previously validated Monte Carlo (MC) simulation algorithm, 12 primary and scatter bCT projection images with a 30-degree step were generated from each phantom. For each projection, the thickness map and breast location in the field of view were also calculated. A U-Net based DL model was developed to estimate the scatter signal based on the total input simulated image and trained single-projection-wise, with the thickness map and breast location provided as additional inputs. The model was internally validated using MC-simulated projections and tested using an external data set of 10 phantoms derived from images acquired with a different bCT system. For this purpose, the mean relative difference (MRD) and mean absolute error (MAE) were calculated. To test for accuracy in reconstructed images, a full bCT acquisition was mimicked with MC-simulations and then assessed by calculating the MAE and the structural similarity (SSIM). Subsequently, scatter was estimated and subtracted from the bCT scans of three patients to obtain the scatter-corrected image. The scatter-corrected projections were reconstructed and compared with the uncorrected reconstructions by evaluating the correction of the cupping artifact, increase in image contrast, and contrast-to-noise ratio (CNR). RESULTS: The mean MRD and MAE across all cases (min, max) for the internal validation set were 0.04% (-1.1%, 1.3%) and 2.94% (2.7%, 3.2%), while for the external test set they were -0.64% (-1.6%, 0.2%) and 2.84% (2.3%, 3.5%), respectively. For MC-simulated reconstruction slices, the computed SSIM was 0.99 and the MAE was 0.11% (range: 0%, 0.35%) with a single outlier slice of 2.06%. For the three patient bCT reconstructed images, the correction increased the contrast by a mean of 25% (range: 20%, 30%), and reduced the cupping artifact. The mean CNR increased by 0.32 after scatter correction, which was not found to be significant (95% confidence interval: [-0.01, 0.65], p = 0.059). The time required to correct the scatter in a single bCT projection was 0.2 s on an NVIDIA GeForce GTX 1080 GPU. CONCLUSION: The developed DL model could accurately estimate scatter in bCT projection images and could enhance contrast and correct for cupping artifact in reconstructed patient images without significantly affecting the CNR. The time required for correction would allow its use in daily clinical practice, and the reported accuracy will potentially allow quantitative reconstructions.