Incidental Coronary Artery Calcium: Opportunistic Screening of Previous Nongated Chest Computed Tomography Scans to Improve Statin Rates (NOTIFY-1 Project).

BACKGROUND: Coronary artery calcium (CAC) can be identified on nongated chest computed tomography (CT) scans, but this finding is not consistently incorporated into care. A deep learning algorithm enables opportunistic CAC screening of nongated chest CT scans. Our objective was to evaluate the effect of notifying clinicians and patients of incidental CAC on statin initiation. METHODS: NOTIFY-1 (Incidental Coronary Calcification Quality Improvement Project) was a randomized quality improvement project in the Stanford Health Care System. Patients without known atherosclerotic cardiovascular disease or a previous statin prescription were screened for CAC on a previous nongated chest CT scan from 2014 to 2019 using a validated deep learning algorithm with radiologist confirmation. Patients with incidental CAC were randomly assigned to notification of the primary care clinician and patient versus usual care. Notification included a patient-specific image of CAC and guideline recommendations regarding statin use. The primary outcome was statin prescription within 6 months. RESULTS: Among 2113 patients who met initial clinical inclusion criteria, CAC was identified by the algorithm in 424 patients. After chart review and additional exclusions were made, a radiologist confirmed CAC among 173 of 194 patients (89.2%) who were randomly assigned to notification or usual care. At 6 months, the statin prescription rate was 51.2% (44/86) in the notification arm versus 6.9% (6/87) with usual care (P<0.001). There was also more coronary artery disease testing in the notification arm (15.1% [13/86] versus 2.3% [2/87]; P=0.008). CONCLUSIONS: Opportunistic CAC screening of previous nongated chest CT scans followed by clinician and patient notification led to a significant increase in statin prescriptions. Further research is needed to determine whether this approach can reduce atherosclerotic cardiovascular disease events. REGISTRATION: URL: https://www. CLINICALTRIALS: gov; Unique identifier: NCT04789278.

Diagnostic performance of single-phase dual-energy CT to differentiate vascular and nonvascular incidental renal lesions on portal venous phase: comparison with CT.

OBJECTIVES: To determine whether single-phase dual-energy CT (DECT) differentiates vascular and nonvascular renal lesions in the portal venous phase (PVP). Optimal iodine threshold was determined and compared to Hounsfield unit (HU) measurements. METHODS: We retrospectively included 250 patients (266 renal lesions) who underwent a clinically indicated PVP abdominopelvic CT on a rapid-kilovoltage-switching single-source DECT (rsDECT) or a dual-source DECT (dsDECT) scanner. Iodine concentration and HU measurements were calculated by four experienced readers. Diagnostic accuracy was determined using biopsy results and follow-up imaging as reference standard. Area under the curve (AUC) was calculated for each DECT scanner to differentiate vascular from nonvascular lesions and vascular lesions from hemorrhagic/proteinaceous cysts. Univariable and multivariable logistic regression analyses evaluated the association between variables and the presence of vascular lesions. RESULTS: A normalized iodine concentration threshold of 0.25 mg/mL yielded high accuracy in differentiating vascular and nonvascular lesions (AUC 0.93, p < 0.001), with comparable performance to HU measurements (AUC 0.93). Both iodine concentration and HU measurements were independently associated with vascular lesions when adjusted for age, gender, body mass index, and lesion size (AUC 0.95 and 0.95, respectively). When combined, diagnostic performance was higher (AUC 0.96). Both absolute and normalized iodine concentrations performed better than HU measurements (AUC 0.92 vs. AUC 0.87) in differentiating vascular lesions from hemorrhagic/proteinaceous cysts. CONCLUSION: A single-phase (PVP) DECT scan yields high accuracy to differentiate vascular from nonvascular renal lesions. Iodine concentration showed a slightly higher performance than HU measurements in differentiating vascular lesions from hemorrhagic/proteinaceous cysts. KEY POINTS: • A single-phase dual-energy CT scan in the portal venous phase differentiates vascular from nonvascular renal lesions with high accuracy (AUC 0.93). • When combined, iodine concentration and HU measurements showed the highest diagnostic performance (AUC 0.96) to differentiate vascular from nonvascular renal lesions. • Compared to HU measurements, iodine concentration showed a slightly higher performance in differentiating vascular lesions from hemorrhagic/proteinaceous cysts.

Split-Bolus, Single-Acquisition, Dual-Phase Abdominopelvic CT Angiography for the Evaluation of Lung Transplant Candidates: Image Quality and Resource Utilization.

OBJECTIVE. The purpose of this study was to assess the image quality and resource utilization of single-injection, split-bolus, dual-enhancement abdominopelvic CT angiography (hereafter referred to as dual-enhancement CTA) performed for combined vascular and solid organ assessment compared with those of single-injection, single-enhancement abdominopelvic CT angiography (hereafter referred to as single-enhancement CTA) for vascular assessment in combination with additional examinations (CT, MRI, and US) performed to assess for malignancy in lung transplant candidates. MATERIALS AND METHODS. We retrospectively reviewed 100 patients who underwent abdominopelvic CTA examinations before lung transplant. Cohort A (n = 50) underwent dual-enhancement CTA and cohort B (n = 50) underwent single-enhancement CTA. Contrast opacification of the vasculature was assessed along the abdominal aorta through the right femoral artery. Solid organ enhancement was assessed in the right lobe of the liver and the right renal cortex. Measurements of mean radiation dose, contrast exposure, and cost of the studies (in U.S. dollars) were compared. RESULTS. Mean (± SD) vascular enhancement on dual-enhancement CTA and single-enhancement CTA was 334.2 ± 26.5 HU (coefficient of variation, 8.3%) and 340.0 ± 21.6 HU (coefficient of variation, 6.5%) (p = 0.23), respectively. For dual-enhancement CTA and single-enhancement CTA, mean liver enhancement was 125.8 ± 30.5 HU and 60.4 ± 6.9 HU (p < 0.01), respectively, whereas mean renal cortical enhancement was 260.3 ± 62.2 HU and 133.4 ± 38.6 HU (p < 0.01), respectively. The mean IV contrast volume was 150 mL for dual-enhancement CTA and 75 mL for single-enhancement CTA. Cohort A underwent six additional imaging studies (one of which was a CT colonography study with an effective dose of 19.0 mSv) at a total cost of $9840 per patient. Cohort B underwent 44 additional imaging studies (mean effective dose, 12.7 ± 6.5 mSv) at a total cost of $12,846 per patient (resulting in a 30.6% reduction in cost for dual-enhancement CTA studies; p < 0.0001). CONCLUSION. Dual-enhancement abdominopelvic CTA allows combined vascular and abdominopelvic solid organ assessment with improved image quality and a lower cost compared with traditional imaging pathways.

Virtual Unenhanced Images at Dual-Energy CT: Influence on Renal Lesion Characterization.

Background Dual-energy (DE) CT allows reconstruction of virtual noncontrast (VNC) images from a single-phase contrast agent-enhanced examination, potentially reducing the need for multiphasic CT to characterize renal lesions. However, data regarding diagnostic performance of VNC images for the characterization of renal lesions are limited. Purpose To determine whether renal mass CT performed by using VNC images allows for reliable identification of renal lesions and differentiation of contrast-enhanced from unenhanced lesions, compared with unenhanced images. Materials and Methods This is a retrospective study of 293 patients (105 women [mean age, 65 years; age range, 18-91 years] and 188 men [mean age, 66 years; age range, 23-90 years] with 379 renal lesions [craniocaudal diameter, 1.0-4.0 cm]) who underwent a single-energy unenhanced CT examination followed by a nephrographic-phase DE CT between June 2013 and October 2017 by using one of four different DE CT platforms from two vendors. VNC images were calculated by using vendor-specific algorithms. Each lesion was classified in a blinded and independent fashion by using the VNC or unenhanced image in combination with the nephrographic images. Attenuation measurements were obtained on the VNC, unenhanced, and nephrographic images. Unenhanced images and pathologic or imaging follow-up for more than 24 months served as reference standard. Results There was strong overall agreement between VNC and unenhanced images for renal lesion characterization (Cramer V = 0.85). VNC images yielded a high diagnostic performance (area under the receiver operating characteristic curve, 0.91; 95% confidence interval: 0.86, 0.95) for facilitation of differentiation of contrast-enhanced from unenhanced renal lesions. However, there was a reduction in diagnostic performance for depicting contrast-enhanced renal lesions by using VNC compared with unenhanced images (area under the receiver operating characteristic curve, 0.91 [95% confidence interval: 0.86, 0.95] vs 0.96 [95% confidence interval: 0.93, 0.99]; P < .001). Mean absolute difference between the VNC and unenhanced attenuation was 9.2 HU ± 8.7. Conclusion Virtual noncontrast images enabled accurate renal lesion characterization, albeit with a reduction in diagnostic performance for contrast-enhanced lesion characterization. © RSNA, 2019 Online supplemental material is available for this article.

Reproducibility of CT Radiomic Features within the Same Patient: Influence of Radiation Dose and CT Reconstruction Settings.

Background Results of recent phantom studies show that variation in CT acquisition parameters and reconstruction techniques may make radiomic features largely nonreproduceable and of limited use for prognostic clinical studies. Purpose To investigate the effect of CT radiation dose and reconstruction settings on the reproducibility of radiomic features, as well as to identify correction factors for mitigating these sources of variability. Materials and Methods This was a secondary analysis of a prospective study of metastatic liver lesions in patients who underwent staging with single-energy dual-source contrast material-enhanced staging CT between September 2011 and April 2012. Technique parameters were altered, resulting in 28 CT data sets per patient that included different dose levels, section thicknesses, kernels, and reconstruction algorithm settings. By using a training data set (n = 76), reproducible intensity, shape, and texture radiomic features (reproducibility threshold, R2 ≥ 0.95) were selected and correction factors were calculated by using a linear model to convert each radiomic feature to its estimated value in a reference technique. By using a test data set (n = 75), the reproducibility of hierarchical clustering based on 106 radiomic features measured with different CT techniques was assessed. Results Data in 78 patients (mean age, 60 years ± 10; 33 women) with 151 liver lesions were included. The percentage of radiomic features deemed reproducible for any variation of the different technical parameters was 11% (12 of 106). Of all technical parameters, reconstructed section thickness had the largest impact on the reproducibility of radiomic features (12.3% [13 of 106]) if only one technical parameter was changed while all other technical parameters were kept constant. The results of the hierarchical cluster analysis showed improved clustering reproducibility when reproducible radiomic features with dedicated correction factors were used (ρ = 0.39-0.71 vs ρ = 0.14-0.47). Conclusion Most radiomic features are highly affected by CT acquisition and reconstruction settings, to the point of being nonreproducible. Selecting reproducible radiomic features along with study-specific correction factors offers improved clustering reproducibility. © RSNA, 2019 Online supplemental material is available for this article. See also the editorial by Sosna in this issue.

Comparison of Iodine Quantification and Conventional Attenuation Measurements for Differentiating Small, Truly Enhancing Renal Masses From High-Attenuation Nonenhancing Renal Lesions With Dual-Energy CT.

OBJECTIVE. The purpose of this study is to determine whether iodine quantification techniques from contrast-enhanced dual-energy CT (DECT) data allow equal differentiation of small enhancing renal masses from high-attenuation (> 20 HU of unenhanced attenuation) nonenhancing lesions, compared with conventional attenuation measurements. MATERIALS AND METHODS. A total of 220 nonconsecutive patients (mean [± SD] age, 66 ± 13 years; 130 men and 90 women) with 265 high-attenuation renal lesions (mean attenuation, 54 ± 33 HU; 91 enhancing lesions) were included. Each patient underwent single-energy unenhanced CT followed by DECT during the nephrographic phase using one of four different high-end DECT platforms (first- and second-generation rapid-kilovoltage-switching DECT platforms and second- and third-generation dual-source DECT platforms). Iodine quantification measurements and conventional attenuation change measurements were calculated for each lesion. Diagnostic accuracy was determined by pathologic analysis, confirmation with another imaging modality, or greater than 24 months of imaging follow-up as the reference standard. RESULTS. The diagnostic accuracy for differentiating enhancing from nonenhancing renal lesions was significantly higher for conventional attenuation change measurements, compared with iodine quantification measurements (AUC values, 0.973 vs 0.875; p < 0.0001). The diagnostic performance of iodine quantification measurements improved only marginally with the utilization of DECT platform-specific optimized iodine quantification thresholds, yielding AUC values of 0.907 and 0.893 for the rapid-kilovoltage-switching DECT and dual-source DECT platforms, respectively. Unenhanced lesion attenuation (p = 0.0010) and intraparenchymal location (p = 0.0249) significantly influenced the diagnostic accuracy of the iodine quantification techniques. CONCLUSION. Iodine quantification from DECT data yields inferior diagnostic accuracy when compared with conventional attenuation change measurements for differentiating small, truly enhancing renal masses and high-attenuation renal lesions.

Systematic Review and Meta-Analysis Investigating the Diagnostic Yield of Dual-Energy CT for Renal Mass Assessment.

OBJECTIVE. The objective of our study was to perform a systematic review and meta-analysis to evaluate the diagnostic accuracy of dual-energy CT (DECT) for renal mass evaluation. MATERIALS AND METHODS. In March 2018, we searched MEDLINE, Cochrane Database of Systematic Reviews, Embase, and Web of Science databases. Analytic methods were based on Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). Pooled estimates for sensitivity, specificity, and diagnostic odds ratios were calculated for DECT-based virtual monochromatic imaging (VMI) and iodine quantification techniques as well as for conventional attenuation measurements from renal mass CT protocols. I2 was used to evaluate heterogeneity. The methodologic quality of the included studies and potential bias were assessed using items from the Quality Assessment Tool for Diagnostic Accuracy Studies 2 (QUADAS-2). RESULTS. Of the 1043 articles initially identified, 13 were selected for inclusion (969 patients, 1193 renal masses). Cumulative data of sensitivity, specificity, and summary diagnostic odds ratio for VMI were 87% (95% CI, 80-92%; I2, 92.0%), 93% (95% CI, 90-96%; I2, 18.0%), and 183.4 (95% CI, 30.7-1093.4; I2, 61.6%), respectively. Cumulative data of sensitivity, specificity, and summary diagnostic odds ratio for iodine quantification were 99% (95% CI, 97-100%; I2, 17.6%), 91% (95% CI, 89-94%; I2, 84.2%), and 511.5 (95% CI, 217-1201; I2, 0%). No significant differences in AUCs were found when comparing iodine quantification to conventional attenuation measurements (p = 0.79). CONCLUSION. DECT yields high accuracy for renal mass evaluation. Determination of iodine content with the iodine quantification technique shows diagnostic accuracy similar to conventional attenuation measurements from renal mass CT protocols. The iodine quantification technique may be used to characterize incidental renal masses when a dedicated renal mass protocol is not available.

Dual-Energy CT Material Density Iodine Quantification for Distinguishing Vascular From Nonvascular Renal Lesions: Normalization Reduces Intermanufacturer Threshold Variability.

OBJECTIVE: The purpose of this study was to determine whether a single, uniform normalized iodine threshold reduces variability and enables reliable differentiation between vascular and nonvascular renal lesions independent of the dual-energy CT (DECT) platform used. MATERIALS AND METHODS: In this retrospective, HIPAA-compliant, institutional review board-approved study, 247 patients (156 men, 91 women; mean age ± SD, 67 ± 12 years old) with 263 renal lesions (193 nonvascular, 70 vascular) underwent unenhanced single-energy and contrast-enhanced DECT scans. One hundred and six nonvascular and 38 vascular lesions were scanned on two dual-source DECT (dsDECT) scanners, and 87 nonvascular and 32 vascular lesions were scanned on two rapid-kilovoltage-switching single-source DECT (rsDECT) scanners. Optimal absolute and normalized (to aorta) lesion iodine thresholds were determined for each platform type and for the entire cohort combined. RESULTS: Mean optimal absolute discriminant thresholds were 1.3 mg I/mL (95% CI, 1.2-1.9 mg I/mL), 1.6 mg I/mL (95% CI, 0.9-1.5 mg I/mL), and 1.5 mg I/mL (95% CI, 1.4-1.7 mg I/mL) for dsDECT, rsDECT, and combined cohorts, respectively. Optimal normalized discriminant thresholds were 0.3 mg I/mL (95% CI, 0.2-0.4 mg I/mL) for both the dsDECT and rsDECT cohorts, and 0.3 mg I/mL (0.3-0.4 mg I/mL) for the combined cohort. The AUC, sensitivity, and specificity for the combined optimal normalized discriminant threshold of 0.3 mg I/mL was 0.96 (95% CI, 0.92-1.00), 0.93 (0.84-0.97), and 0.95 (0.91-0.98), respectively. Normalization resulted in decreased variability and better lesion separation (effect size, 1.77 vs 1.69, p < 0.0001). CONCLUSION: The optimal absolute discriminant threshold for evaluating renal lesions varies depending on the type of DECT platform, though this difference is not statistically significant. Variation can be reduced with a better separation of vascular and nonvascular lesions by normalizing iodine quantification to the aorta.

Deep-learning-assisted diagnosis for knee magnetic resonance imaging: Development and retrospective validation of MRNet.

BACKGROUND: Magnetic resonance imaging (MRI) of the knee is the preferred method for diagnosing knee injuries. However, interpretation of knee MRI is time-intensive and subject to diagnostic error and variability. An automated system for interpreting knee MRI could prioritize high-risk patients and assist clinicians in making diagnoses. Deep learning methods, in being able to automatically learn layers of features, are well suited for modeling the complex relationships between medical images and their interpretations. In this study we developed a deep learning model for detecting general abnormalities and specific diagnoses (anterior cruciate ligament [ACL] tears and meniscal tears) on knee MRI exams. We then measured the effect of providing the model's predictions to clinical experts during interpretation. METHODS AND FINDINGS: Our dataset consisted of 1,370 knee MRI exams performed at Stanford University Medical Center between January 1, 2001, and December 31, 2012 (mean age 38.0 years; 569 [41.5%] female patients). The majority vote of 3 musculoskeletal radiologists established reference standard labels on an internal validation set of 120 exams. We developed MRNet, a convolutional neural network for classifying MRI series and combined predictions from 3 series per exam using logistic regression. In detecting abnormalities, ACL tears, and meniscal tears, this model achieved area under the receiver operating characteristic curve (AUC) values of 0.937 (95% CI 0.895, 0.980), 0.965 (95% CI 0.938, 0.993), and 0.847 (95% CI 0.780, 0.914), respectively, on the internal validation set. We also obtained a public dataset of 917 exams with sagittal T1-weighted series and labels for ACL injury from Clinical Hospital Centre Rijeka, Croatia. On the external validation set of 183 exams, the MRNet trained on Stanford sagittal T2-weighted series achieved an AUC of 0.824 (95% CI 0.757, 0.892) in the detection of ACL injuries with no additional training, while an MRNet trained on the rest of the external data achieved an AUC of 0.911 (95% CI 0.864, 0.958). We additionally measured the specificity, sensitivity, and accuracy of 9 clinical experts (7 board-certified general radiologists and 2 orthopedic surgeons) on the internal validation set both with and without model assistance. Using a 2-sided Pearson's chi-squared test with adjustment for multiple comparisons, we found no significant differences between the performance of the model and that of unassisted general radiologists in detecting abnormalities. General radiologists achieved significantly higher sensitivity in detecting ACL tears (p-value = 0.002; q-value = 0.019) and significantly higher specificity in detecting meniscal tears (p-value = 0.003; q-value = 0.019). Using a 1-tailed t test on the change in performance metrics, we found that providing model predictions significantly increased clinical experts' specificity in identifying ACL tears (p-value < 0.001; q-value = 0.006). The primary limitations of our study include lack of surgical ground truth and the small size of the panel of clinical experts. CONCLUSIONS: Our deep learning model can rapidly generate accurate clinical pathology classifications of knee MRI exams from both internal and external datasets. Moreover, our results support the assertion that deep learning models can improve the performance of clinical experts during medical imaging interpretation. Further research is needed to validate the model prospectively and to determine its utility in the clinical setting.

Deep learning for chest radiograph diagnosis: A retrospective comparison of the CheXNeXt algorithm to practicing radiologists.

BACKGROUND: Chest radiograph interpretation is critical for the detection of thoracic diseases, including tuberculosis and lung cancer, which affect millions of people worldwide each year. This time-consuming task typically requires expert radiologists to read the images, leading to fatigue-based diagnostic error and lack of diagnostic expertise in areas of the world where radiologists are not available. Recently, deep learning approaches have been able to achieve expert-level performance in medical image interpretation tasks, powered by large network architectures and fueled by the emergence of large labeled datasets. The purpose of this study is to investigate the performance of a deep learning algorithm on the detection of pathologies in chest radiographs compared with practicing radiologists. METHODS AND FINDINGS: We developed CheXNeXt, a convolutional neural network to concurrently detect the presence of 14 different pathologies, including pneumonia, pleural effusion, pulmonary masses, and nodules in frontal-view chest radiographs. CheXNeXt was trained and internally validated on the ChestX-ray8 dataset, with a held-out validation set consisting of 420 images, sampled to contain at least 50 cases of each of the original pathology labels. On this validation set, the majority vote of a panel of 3 board-certified cardiothoracic specialist radiologists served as reference standard. We compared CheXNeXt's discriminative performance on the validation set to the performance of 9 radiologists using the area under the receiver operating characteristic curve (AUC). The radiologists included 6 board-certified radiologists (average experience 12 years, range 4-28 years) and 3 senior radiology residents, from 3 academic institutions. We found that CheXNeXt achieved radiologist-level performance on 11 pathologies and did not achieve radiologist-level performance on 3 pathologies. The radiologists achieved statistically significantly higher AUC performance on cardiomegaly, emphysema, and hiatal hernia, with AUCs of 0.888 (95% confidence interval [CI] 0.863-0.910), 0.911 (95% CI 0.866-0.947), and 0.985 (95% CI 0.974-0.991), respectively, whereas CheXNeXt's AUCs were 0.831 (95% CI 0.790-0.870), 0.704 (95% CI 0.567-0.833), and 0.851 (95% CI 0.785-0.909), respectively. CheXNeXt performed better than radiologists in detecting atelectasis, with an AUC of 0.862 (95% CI 0.825-0.895), statistically significantly higher than radiologists' AUC of 0.808 (95% CI 0.777-0.838); there were no statistically significant differences in AUCs for the other 10 pathologies. The average time to interpret the 420 images in the validation set was substantially longer for the radiologists (240 minutes) than for CheXNeXt (1.5 minutes). The main limitations of our study are that neither CheXNeXt nor the radiologists were permitted to use patient history or review prior examinations and that evaluation was limited to a dataset from a single institution. CONCLUSIONS: In this study, we developed and validated a deep learning algorithm that classified clinically important abnormalities in chest radiographs at a performance level comparable to practicing radiologists. Once tested prospectively in clinical settings, the algorithm could have the potential to expand patient access to chest radiograph diagnostics.

Energy-Specific Optimization of Attenuation Thresholds for Low-Energy Virtual Monoenergetic Images in Renal Lesion Evaluation.

OBJECTIVE: The purpose of this study was to determine in vitro and in vivo the optimal threshold for renal lesion vascularity at low-energy (40-60 keV) virtual monoenergetic imaging. MATERIALS AND METHODS: A rod simulating unenhanced renal parenchymal attenuation (35 HU) was fitted with a syringe containing water. Three iodinated solutions (0.38, 0.57, and 0.76 mg I/mL) were inserted into another rod that simulated enhanced renal parenchyma (180 HU). Rods were inserted into cylindric phantoms of three different body sizes and scanned with single- and dual-energy MDCT. In addition, 102 patients (32 men, 70 women; mean age, 66.8 ± 12.9 [SD] years) with 112 renal lesions (67 nonvascular, 45 vascular) measuring 1.1-8.9 cm underwent single-energy unenhanced and contrast-enhanced dual-energy CT. Optimal threshold attenuation values that differentiated vascular from nonvascular lesions at 40-60 keV were determined. RESULTS: Mean optimal threshold values were 30.2 ± 3.6 (standard error), 20.9 ± 1.3, and 16.1 ± 1.0 HU in the phantom, and 35.9 ± 3.6, 25.4 ± 1.8, and 17.8 ± 1.8 HU in the patients at 40, 50, and 60 keV. Sensitivity and specificity for the thresholds did not change significantly between low-energy and 70-keV virtual monoenergetic imaging (sensitivity, 87-98%; specificity, 90-91%). The AUC from 40 to 70 keV was 0.96 (95% CI, 0.93-0.99) to 0.98 (95% CI, 0.95-1.00). CONCLUSION: Low-energy virtual monoenergetic imaging at energy-specific optimized attenuation thresholds can be used for reliable characterization of renal lesions.

Characterization of Small Incidental Indeterminate Hypoattenuating Hepatic Lesions: Added Value of Single-Phase Contrast-Enhanced Dual-Energy CT Material Attenuation Analysis.

OBJECTIVE: The objective of this study is to determine whether single-phase contrast-enhanced dual-energy CT (DECT) material attenuation analysis improves the characterization of small (< 2.0 cm) incidental indeterminate hypoattenuating hepatic lesions, compared with conventional single-energy CT evaluation. MATERIALS AND METHODS: This retrospective study involved 55 patients (24 men and 31 women; mean [± SD] age, 63.9 ± 15.3 years) with 77 incidental hypoattenuating hepatic lesions (59 benign and 18 malignant lesions) measuring 0.5-2.0 cm who underwent single-phase contrast-enhanced DECT of the abdomen for pain. For each lesion, attenuation measurements were obtained using blended 120-kVp-equivalent images and contrast map images. DECT material attenuation images were used for iodine quantification. Optimal lesion attenuation and iodine concentration threshold values that best distinguished benign lesions from malignant lesions were generated using smooth bootstrapping. The diagnostic accuracy of the optimized thresholds was compared using the Wilcox rank sum test. RESULTS: The optimal mean (± standard error) attenuation threshold values that best differentiated benign and malignant lesions were 50.2 ± 5.2 HU and 11.5 ± 2.0 HU when blended 120-kVp and contrast map images, respectively, were used. The iodine concentration (expressed as milligrams of iodine per milliliter) differed significantly (p < 0.0001) between benign lesions (0.6 ± 0.4 mg I/mL) and malignant lesions (1.7 ± 0.4 mg I/mL). The optimal iodine concentration that best distinguished between benign and malignant lesions was 1.2 ± 0.1 mg I/mL. The sensitivity, specificity, and AUC value were highest for iodine concentration (0.94, 0.93, and 0.97, respectively), compared with blended images (0.89, 0.70, and 0.81, respectively) and contrast map images (0.94, 0.64, 0.77, respectively). CONCLUSION: Iodine quantification performed using single-phase contrast-enhanced DECT material attenuation images improves the characterization of small (< 2 cm) incidental indeterminate hypoattenuating hepatic lesions, compared with conventional attenuation measurements.

Clinically Acceptable Optimized Dose Reduction in Computed Tomographic Imaging of Necrotizing Pancreatitis Using a Noise Addition Software Tool.

OBJECTIVE: This study aimed to determine potential radiation dose reduction of contrast-enhanced computed tomography (CECT) for imaging necrotizing pancreatitis (NP) using a noise addition tool. METHODS: Eighty-four patients were identified with at least 1 abdominopelvic CECT for NP within a 2-year period. Sixty consecutive scans were selected as reference radiation dose data sets. A noise addition software was used to simulate 4 data sets of increased noise. Readers rated confidence for identifying (i) anatomic structures, (ii) complications of NP, and (iii) diagnostic acceptability. Noise and dose levels were identified at acceptability threshold where observer scores were statistically indistinguishable from full-dose computed tomographies. RESULTS: Observers' perception of image tasks decreased progressively with increasing noise (P < 0.05). Acceptability and statistical analysis indicated that noise can be increased from 10 to 25 HU corresponding to an 84% reduction in dose without change in observer perception (P < 0.05). CONCLUSIONS: Higher image noise levels may be tolerated in CECT in patients with NP.

Financial Forecasting and Stochastic Modeling: Predicting the Impact of Business Decisions.

In health care organizations, effective investment of precious resources is critical to assure that the organization delivers high-quality and sustainable patient care within a supportive environment for patients, their families, and the health care providers. This holds true for organizations independent of size, from small practices to large health systems. For radiologists whose role is to oversee the delivery of imaging services and the interpretation, communication, and curation of imaging-informed information, business decisions influence where and how they practice, the tools available for image acquisition and interpretation, and ultimately their professional satisfaction. With so much at stake, physicians must understand and embrace the methods necessary to develop and interpret robust financial analyses so they effectively participate in and better understand decision making. This review discusses the financial drivers upon which health care organizations base investment decisions and the central role that stochastic financial modeling should play in support of strategically aligned capital investments. Given a health care industry that has been slow to embrace advanced financial analytics, a fundamental message of this review is that the skills and analytical tools are readily attainable and well worth the effort to implement in the interest of informed decision making. © RSNA, 2017 Online supplemental material is available for this article.

Characterization of Small (< 4 cm) Focal Renal Lesions: Diagnostic Accuracy of Spectral Analysis Using Single-Phase Contrast-Enhanced Dual-Energy CT.

OBJECTIVE: The purpose of this study is to determine whether single-phase contrast-enhanced dual-energy quantitative spectral analysis improves the accuracy of diagnosis of small (< 4.0 cm) renal lesions, compared with conventional single-energy attenuation measurements. MATERIALS AND METHODS: In this retrospective study, 136 consecutive patients (95 men and 41 women; mean age, 54 years) with 144 renal lesions (111 benign and 33 malignant) underwent single-energy unenhanced and dual-energy contrast-enhanced CT of the abdomen. For each renal lesion, attenuation measurements were obtained, and an attenuation change of 15 HU or greater was considered evidence of enhancement. Dual-energy spectral attenuation curves were generated for each lesion. The slope of each curve was measured between 40 and 50 keV (λHU40-50), 40 and 70 keV (λHU40-70), and 40 and 140 keV (λHU40-140). Mean lesion attenuation values and spectral attenuation curve parameters were compared between benign and malignant renal lesions by use of the two-sample t test. Diagnostic accuracy was assessed and validated using cross-validation analysis. RESULTS: With the use of cross-validated optimal thresholds at 100% sensitivity, specificity for differentiating between benign and malignant renal lesions improved significantly when both λHU40-70 and λHU40-140 were used, compared with conventional enhancement measurements (93% [103/111; 95% CI, 86-97%] vs 81% [90/111; 95% CI, 73-88%]) (p = 0.02). The sensitivity of λHU40-70 and λHU40-140 was also higher than that of conventional enhancement measurements, although it was not statistically significant. CONCLUSION: Single-phase contrast-enhanced dual-energy quantitative spectral analysis significantly improves the specificity for characterization of small (< 4.0 cm) renal lesions, compared with conventional single-energy attenuation measurements.

Preoperative Multidetector CT Diagnosis of Extrapancreatic Perineural or Duodenal Invasion Is Associated with Reduced Postoperative Survival after Pancreaticoduodenectomy for Pancreatic Adenocarcinoma: Preliminary Experience and Implications for Patient Care.

Purpose To test the hypothesis that patients with pancreatic adenocarcinoma who otherwise are viewed to have resectable disease but have preoperative findings of extrapancreatic perineural invasion (EPNI) and/or duodenal invasion at multidetector computed tomography (CT) have reduced postoperative survival after pancreaticoduodenectomy for pancreatic ductal adenocarcinoma (PDAC). Materials and Methods This study was approved by the institutional review board and complied with HIPAA. The authors retrospectively evaluated 76 consecutive patients with PDAC who underwent preoperative multidetector CT and subsequent pancreaticoduodenectomy. Two radiologists blinded to surgical pathology results and clinical outcome evaluated multidetector CT images for evidence of EPNI and duodenal invasion; discrepancies were resolved by consensus. Also determined for each patient were resected lymph node status, tumor size, surgical margin status, time to progression, and time to death. Data were assessed with the Goodman-Kruskal gamma for correlations among indicators and the log-rank test, Kaplan-Meier estimates, and multivariate Cox proportional hazards regression for survival analysis. Results In univariate analysis, duodenal invasion and/or EPNI on preoperativemultidetector CT images was associated with significantly decreased progression-free survival (P < .0001) and overall survival (P = .0013), and the clinical indicators (lymph node status, tumor size, and surgical margin status) as well as duodenal invasion and/or EPNI showed correlation with each other. In multivariate regression that included multidetector CT findings as well as the three traditional clinical indicators, only duodenal invasion and/or EPNI showed significant independent association with reduction in both modes of survival (P < .0001 and P = .014, respectively). Interobserver agreement was substantial with respect to EPNI and duodenal invasion (κ = 0.691 and 0.682, respectively). Conclusion Patients with evidence of EPNI and/or duodenal invasion on preoperative multidetector CT images have significantly reduced survival after pancreaticoduodenectomy for PDAC. © RSNA, 2016.

MDCT Diagnosis of Perineural Invasion Involving the Celiac Plexus in Intrahepatic Cholangiocarcinoma: Preliminary Observations and Clinical Implications.

OBJECTIVE: The purpose of this study was to test the hypothesis that soft-tissue infiltration along the celiac plexus and delayed enhancement exceeding two-thirds of the tumor area on preoperative MDCT correlate with histologic evidence of perineural invasion in resected intrahepatic cholangiocarcinomas. MATERIALS AND METHODS: Two experienced abdominal radiologists retrospectively reviewed preoperative multiphasic MDCT scans of 20 patients who underwent resection of intrahepatic cholangiocarcinoma, identifying soft-tissue infiltration along the celiac plexus, delayed enhancement exceeding two-thirds of the tumor area, and maximum tumor diameter. Consensus findings were compared with intratumoral perineural invasion in resected intrahepatic cholangiocarcinomas using the Fisher exact test. RESULTS: Six patients had histologic intratumoral perineural invasion, five of whom had soft-tissue infiltration along the celiac plexus on preoperative MDCT, with corresponding 83.3% sensitivity and 92.9% specificity for perineural invasion and significant association between these MDCT and histologic findings (p = 0.002). No patients with histologic perineural invasion had enhancement exceeding two-thirds of the tumor area on MDCT; sensitivity was 0.0% for this finding. Tumor diameter on MDCT was not significantly associated with perineural invasion at histopathology (p = 0.530). CONCLUSION: Soft-tissue infiltration along the celiac plexus on MDCT is an indicator of perineural invasion in patients with intrahepatic cholangiocarcinoma. The data did not confirm an association between delayed enhancement exceeding two-thirds of the tumor area and perineural invasion. Because perineural invasion from intrahepatic cholangiocarcinoma is associated with a very poor prognosis and is generally a contraindication to surgery, the MDCT diagnosis of celiac plexus perineural invasion in patients with intrahepatic cholangiocarcinoma may have important implications for prognosis and treatment planning.

Dual-Energy Multidetector-Row Computed Tomography of the Hepatic Arterial System: Optimization of Energy and Material-Specific Reconstruction Techniques.

PURPOSE: To investigate the optimal dual-energy reconstruction technique for the visualization of the hepatic arterial system during dual-energy multidetector computed tomographic (MDCT) angiography of the liver. MATERIALS AND METHODS: Twenty-nine nonconsecutive patients underwent dual-energy MDCT angiography of the liver. Synthesized monochromatic (40, 50, 60, and 80 keV) and iodine density data sets were reconstructed. Aortic attenuation, noise, and contrast-to-noise ratio (CNR) were measured. In addition, volume-rendered images were generated and qualitatively assessed by 2 independent readers, blinded to technique. The impact of body size on the readers' scores was also assessed. RESULTS: Aortic attenuation, noise, and CNR increased progressively with decreasing keV and were significantly higher between 40 and 60 keV (P < 0.001). There was a significant improvement of readers' visualization of arterial anatomy at lower monochromatic energies (P < 0.001). Iodine density images yielded significantly higher CNR compared with all monochromatic data sets (P < 0.001). However, iodine density images were scored nondiagnostic by the 2 readers. CONCLUSIONS: Synthesized monochromatic images between 40 and 60 keV maximize the magnitude of arterial enhancement and improve visualization of hepatic arterial anatomy at dual-energy MDCT angiography of the liver. Larger body sizes may counteract the benefits of using lower monochromatic energies.

The role of CT-guided percutaneous drainage of loculated air collections: an institutional experience.

OBJECTIVE: The purpose of this study is to describe our experience with the role of CT-guided percutaneous drainage of loculated intra-abdominal collections consisting entirely of gas. MATERIALS AND METHODS: An IRB-approved retrospective study analyzing patients with air-only intra-abdominal collections over an 8-year period was undertaken. Seven patients referred for percutaneous drainage were included. Size of collections, subsequent development of fluid, and microbiological yield were determined. Clinical outcome was also analyzed. RESULTS: Out of 2835 patients referred for percutaneous drainage between 2004 and 2012, seven patients (5M, 2F; average age 63, range 54-85) met criteria for inclusion with CT showing air-only collections. Percutaneous drain placement (five 8 Fr, one 10 Fr, and one 12 Fr) using Seldinger technique was performed. Four patients (57%) had recently undergone surgery (2 Whipple, 1 colectomy, 1 hepatic resection) while two (29%) had a remote surgery (1 abdominoperineal resection, 1 sigmoidectomy). Despite the lack of detectable fluid on the original CT, 6 patients (86%) had air and fluid aspirated at drainage, 5 (83%) of the aspirates developed positive microbacterial cultures. Four patients (57%) presented with fever at the time of the initial scan, all of whom had positive cultures from aspirated fluid. Four patients (57%) had leukocytosis, all of whom had positive cultures from aspirated fluid. CONCLUSIONS: Although relatively rare in occurrence, patients with air-only intra-abdominal collections with signs of infection should be considered for percutaneous management similar to that of conventional infected fluid collections. Although fluid is not visible on CT, these collections can produce fluid that contains organisms.