Search

Rationale and Objectives In diagnostic accuracy studies, cases in which a reader does not see the condition of interest are often given the same score for ROC analysis (e.g. confidence score of 0%). However, many of these cases can be further distinguished and doing so may result in more robust ROC results. Materials and Methods We examined two recent, real-world studies which used different procedures to encourage readers to further distinguish subjects who appear to be without the condition of interest. For each study, we calculated the results under two conditions. In the “zeros distinguished” (ZD) condition, we incorporated the confidence scores collected to further distinguish the normal-looking subjects. In the “zeros not distinguished” (ZND) condition, we disregarded these scores and simply gave the unit of analysis a score of zero whenever the reader did not identify the condition of interest in that unit. We compared the two conditions on (1) coverage of the ROC space and (2) discrepancy between parametric and nonparametric results. Results Compared to the ZND condition, coverage of the ROC space was improved in the ZD condition for all ROC curves in both studies. In the first study, there was a significant reduction in the discrepancy between parametric and nonparametric results (median discrepancy in ZND vs ZD condition: 0.033 vs 0.011, p = 0.012). A similar reduction was not seen in the second study, though the discrepancies were very low in both conditions (0.003 vs 0.006, p = 0.313). Conclusion Prompting readers to further distinguish cases in which they do not see the condition of interest may result in more robust ROC results, though further exploration of this topic is warranted.

Previous studies have reported mixed results regarding whether the use of structured reporting (SR) leads to a change in interpretation times. The objective of this study was to quantify any change in interpretation times after the implementation of SR for multiple sclerosis (MS) follow-up magnetic resonance imaging (MRI) of the brain.Interpretation times before and after the transition to MS MRI SR were compared over a 5-year period. To control for changing practice patterns, a control group of non-MS (intracranial masses) reports not using SR was also assessed. In a secondary analysis, interpretation times for 2D and 3D MRI MS protocols after the initiation of SR were compared to determine whether increased image number with the 3D protocol affected interpretation times.Mean and median interpretation times before the initiation of SR for MS MRI were 11.0 and 8.0 minutes versus 8.5 and 6.0 minutes after the implementation of SR (p0.001). Although non-MS MRI interpretation times also decreased, an interaction analysis demonstrated that the decrease in MS interpretation times was significantly higher (p0.001). Mean and median interpretation times using 3D protocols were slighter increased compared to interpretation times with 2D protocols (p = 0.036).After the implementation of SR for MS follow-up MRI at our institution, interpretation times significantly decreased despite the increased number of images with some of the examinations due to the adoption of 3D protocols. The adoption of SR for MS MRI follow-up scans may improve radiologist efficiency.

To describe the MRI features of deltoid tears and to evaluate tear characteristics in patient groups based on history of trauma and rotator cuff tear (RCT). The records of patients who underwent shoulder MRI at our institution between July 2007 and June 2018 were retrospectively reviewed to identify deltoid tears, and patients were divided into groups based on history of recent trauma and presence of RCT. Images were reviewed to identify the location and size of the deltoid tear; the presence or absence of RCT, muscle atrophy, tendon retraction, humeral head subluxation, soft tissue edema, and additional pathologies were also noted. Medical records were reviewed for information about history of steroid injection, previous rotator cuff surgery, and treatments used. Among 69 patients with deltoid tears (45 men; mean age, 65.2 years; range, 19–89 years), patients with RCTs and no trauma had the highest frequency of deltoid tears in the middle portion (p = 0.005). Only patients with RCTs had undergone steroid injection or rotator cuff surgery. Two patients had deltoid tear without RCT and without recent trauma; these patients demonstrated evidence of calcific tendinopathy and chronic subacromial-subdeltoid bursitis. The middle (acromial) portion of the deltoid is more frequently affected in patients with RCTs than in those with trauma. Although deltoid tears are commonly associated with RCT, calcific tendinopathy and chronic bursitis may also be seen in patients with deltoid tears.

To identify clinical variables, including use of newer Viatorr TIPS endoprosthesis with controlled expansion (VCX) that may affect the occurrence and risk of hepatic encephalopathy after transjugular intrahepatic portosystemic shunt (TIPS) creation.A total of 376 patients who underwent TIPS creation at our institution between 2003 and 2018 were retrospectively identified. Of these patients, 71 received a Viatorr controlled expansion endoprosthesis and 305 received a Viatorr TIPS endoprosthesis (older version without controlled expansion). Multivariate regression analysis was used to identify factors predicting the occurrence of hepatic encephalopathy after TIPS creation; a Cox proportional hazard model was used to assess risk of HE through time to HE onset RESULTS: A total of 194 patients (52%) developed hepatic encephalopathy after TIPS creation, including 28 of 71 patients (39%) who received a VCX endoprosthesis. Older patient age and the use of Viatorr endoprosthesis without controlled expansion were significantly associated with the development of hepatic encephalopathy overall. Pre-TIPS pressure variables, patient age, plasma international normalized ratio, and model for end-stage liver disease score were risk factors for time to hepatic encephalopathy.Several variables are mild predictors of early hepatic encephalopathy development after TIPS creation, and the use of VCX endoprosthesis in TIPS creation is associated with a modest lower risk of hepatic encephalopathy. These preliminary findings should be considered in regard to patient selection, endoprosthesis selection, and post-transjugular intrahepatic portosystemic shunt creation monitoring for the development of hepatic encephalopathy.

The design and analysis of multireader multicase (MRMC) studies are quite challenging. These studies differ from most medical studies because they need a reference standard and sampling from two populations (ie, reader and patient populations). They are quite expensive to conduct, requiring a good deal of readers' time for image interpretation. One common problem is the use of imperfect reference standards, often correlated with the test or tests being evaluated. Another common issue is oversimplification of the multidimensional MRMC data. In this study, the fundamentals of MRMC study design and analysis are reviewed. The goal is to provide investigators with a guide to the fundamentals of MRMC design and analysis, with references to more detailed discussions. In addition, readers are updated on newer areas of research, including correction for studies with multiple diagnostic accuracy end points and adjustment for location bias.

PURPOSE The purpose of this study was to evaluate patterns of vascular and lung parenchymal enhancement in patients with suspected chronic thromboembolic pulmonary hypertension (CTEPH) and in those with acute pulmonary embolism (PE) and compare those two groups. MATERIALS AND METHODS We retrospectively studied 186 thoracic DECT studies referred for evaluation of CTEPH or pulmonary hypertension. A total of 80 of these patients had a negative scan (control group), 13 had acute PE, and 53 had chronic thromboembolic disease (CTED)/CTEPH. Five different DECT-based parameters were evaluated that highlight patterns of vascular kinetics. Specifically, total DECT-based parenchymal attenuation in Hounsfield Unit (HU) (LungHU), percentage of perfused blood volume (PBV), peak enhancement of main pulmonary artery (PApeak in HU), maximum enhancement corresponding to 100 (PAmax), and the ratio of PApeak to LungHU were calculated. RESULTS Compared with patients with negative CT, patients with CTED/CTEPH tended to have lower LungHU (median: 27 vs. 38, P

We sought to evaluate the correlation between MRI phenotypes of prostate cancer as defined by PI-RADS v2 and the Decipher Genomic Classifier (used to estimate the risk of early metastases). This single-center, retrospective study included 72 nonconsecutive men with prostate cancer who underwent MRI before radical prostatectomy performed between April 2014 and August 2017 and whose MRI registered lesions were microdissected from radical prostatectomy specimens and then profiled using Decipher (89 lesions; 23 MRI invisible [PI-RADS v2 scores ≤ 2] and 66 MRI visible [PI-RADS v2 scores ≥ 3]). Linear regression analysis was used to assess clinicopathologic and MRI predictors of Decipher results; correlation coefficients (r) were used to quantify these associations. AUC was used to determine whether PI-RADS v2 could accurately distinguish between low-risk (Decipher score

OBJECTIVE. The purpose of this study is to assess CT-based markers predictive of the development of chronic thromboembolic pulmonary hypertension (CTEPH) after acute pulmonary embolism. MATERIALS AND METHODS. Identified from a search of local registries, 48 patients who had CTEPH develop were included in the study group, and 113 patients who had complete resolution of acute pulmonary embolism were included in the control group. Baseline CT scans obtained at the time of the initial pulmonary embolism event were evaluated for the degree of clot-induced vessel obstruction, the quantitative Walsh score, the ratio of the right ventricle diameter to the left ventricle diameter, the right atrium diameter, the pulmonary artery diameter, right heart thrombus, pericardial effusion, lung infarction, and mosaic attenuation. Classification and regression tree analysis was used to create a decision tree. The decision tree was externally validated on an anonymized cohort of 50 control subjects and 50 patients with CTEPH. RESULTS. During univariable analysis, an increase in the degree occlusive clot on initial imaging, a decrease in the Walsh score, absence of pericardial effusion, presence of lung infarction, and the presence of mosaic attenuation were associated with an increased probability of CTEPH development. In the final decision tree, the occlusive nature of the clot remained. Two patients in the cohort used for external validation had nondiagnostic findings and were excluded. The decision process correctly classified 33% (16/48) of patients who had CTEPH develop and 86% (43/50) of patients who did not have CTEPH develop, for an odds ratio of 3.1 (95% CI, 1.1-8.3). CONCLUSION. The presence of an occlusive clot on initial imaging is associated with an increased probability of CTEPH development. Presence of mosaic attenuation and lung infarction may also predict CTEPH development, although additional studies are needed.

OBJECTIVE. The objective of our study was to compare morphologic and functional dual-energy CT (DECT) parameters in patients with chronic thromboembolic disease (CTED) and chronic thromboembolic pulmonary hypertension (CTEPH). MATERIALS AND METHODS. Using the local CTEPH registry, we identified 28 patients with CTED and 72 patients with CTEPH. On each DECT examination, a clot burden score was calculated by assigning the following scores for chronic changes by location: pulmonary trunk, 5; each main pulmonary artery (MPA), 4; each lobar branch, 3; each segmental branch, 2; and each subsegmental branch, 1. The total clot burden score was calculated by adding the individual scores from both lungs. Functional parameters were assessed using perfused blood volume (PBV) maps and included lung enhancement (in Hounsfield units), percentage of PBV, MPA peak enhancement (in Hounsfield units), maximum enhancement corresponding to 100, and the ratio of MPA peak enhancement to lung enhancement. A perfusion defect (PD) score was calculated by assigning 1 point to each segmental PD. Patients with CTED and patients with CTEPH were matched using propensity score matching to account for potential confounders. RESULTS. After matching, the CTEPH group showed a higher PD score than the CTED group and unilateral disease was more common in the CTED group than in the CTEPH group. In the unmatched sample, patients with CTED showed increased percentages of PBV for both lungs (PBV total) and for the right lung as compared with the CTEPH group (adjusted p = 0.040 and 0.028, respectively); after adjustment for clot burden, the difference between groups was still noted but was not statistically significant. No statistically significant differences were noted in the various functional DECT parameters after propensity score matching. CONCLUSION. Patients with CTED show anatomic and functional changes in the pulmonary vasculature and lung parenchyma similar to those seen in patients with CTEPH. Functional DECT parameters support the observation that CTED is an intermediate clinical phenotype in the population with chronic pulmonary embolism.

To evaluate the effect of enema and dietary restrictions on prostate MR image quality metrics and to assess inter-reader agreement for these metrics.This retrospective study included 195 men divided into groups based on their compliance with preparation instructions before prostate MRI (Enema + Diet, n = 98; Enema, n = 42; Diet, n = 35; Control [no compliance], n = 20). Four readers independently assessed six image quality metrics on a 5-point scale. Between-group comparisons were made using Wilcoxon rank sum test. Inter-reader agreement was calculated using Fleiss' kappa.Compared with the Control group, image quality with respect to rectal stool/gas, distortion of diffusion-weighted images, overall image quality, and confidence in assessment was higher in the Enema + Diet, Enema, and Diet groups (p 0.05 for all comparisons). The Enema + Diet and Enema groups had significantly higher scores than the Diet group for rectal stool/gas (p0.001 and 0.005, respectively). The Enema + Diet and Diet groups had higher scores than the Control group for rectal peristalsis (p = 0.027 and 0.009, respectively), but there were no significant differences in motion artifacts on T2-weighted images. Agreement among readers was fair, with kappa values ranging from 0.25 to 0.37.Enema and dietary restriction can improve the quality of prostate MRI by decreasing rectal distension and distortion of diffusion-weighted images and by increasing reader confidence in image assessment. Inter-reader agreement using subjective criteria for analysis of MRI quality is fair.

Purpose The purpose of this study was to assess the effect of obesity and iterative reconstruction on the ability to reduce exposure by studying the accuracy for detection of low‐contrast low‐attenuation (LCLA) liver lesions on computed tomography (CT) using a phantom model. Methods A phantom with four unique LCLA liver lesions (5‐ to 15‐mm spheres, –24 to –6 HU relative to 90‐HU background) was scanned without (“thin” phantom) and with (“obese” phantom) a 5‐cm thick fat‐attenuation ring at 150 mAs (thin phantom) and 450 mAs (obese phantom) standard exposures and at 33% and 67% exposure reductions. Images were reconstructed using standard filtered back projection (FBP) and with iterative reconstruction (Adaptive Model‐Based Iterative Reconstruction strength 3, ADMIRE). A noninferiority analysis of lesion detection was performed. Results Mean area under the curve (AUC) values for lesion detection were significantly higher for the thin phantom than for the obese phantom regardless of exposure level (P

To characterize the appearance, location, severity, and prevalence of focal glenoid retroversion on shoulder MRI in professional baseball pitchers versus age-matched controls. In this retrospective review, two musculoskeletal radiologists evaluated shoulder MRI examinations from 40 professional baseball pitchers and 40 age-matched controls. Images were scored for the presence of a focal posterior glenoid convexity and the clock face positions involved. A consensus interpretation was also performed. A third reader identified the presence of additional shoulder abnormalities. After consensus, 60% (24/40) of pitchers were found to have focal glenoid retroversion versus 20% (8/40) of controls (p = 0.001). The most apparent location was posterior or posterior-superior, from 7.75 o’clock to 10.5 o’clock (median, 9.75 o’clock; interquartile range, 9.25–10 o’clock). The median clock angle of involvement was greater in pitchers than in controls (30° vs. 0°; p

The aim of this brief report is to determine the safety and reliability of minimally invasive video-assisted thoracic surgery (VATS) resection without the aid of intraoperative fluoroscopy after computed tomography (CT)–guided microcoil localization of small peripheral pulmonary nodules. Twenty patients with peripheral lung nodules underwent percutaneous needle localization with a microcoil that was tagged back to the visceral pleural surface. Same-day VATS resection was performed without the use of intraoperative fluoroscopy. All 20 nodules were successfully localized in the CT procedure room, and all 20 nodules were resected with negative margins and no major complications.

To determine the fat content of myxoid liposarcomas (MLS) on MRI and to identify any association between lipid content and survival. The fat percentage of MLS diagnosed between January 2006 and December 2016 at a single institution was assessed by two radiologists on preoperative MR images. A Cox proportional hazard model was used to determine any association between tumor fat percentage and survival time. Tumor fat percentage was the single predictor in the model. A significance level of 0.05 was used. The Kaplan–Meier estimator was also used to provide a nonparametric estimate of the survivor function within the entire sample and within two patient subgroups consists of lipid-rich and lipid-poor tumors. Lipid-rich tumors were defined as any tumors showing more than 20% of fat on MRI. A 20% cutoff was determined arbitrarily. Of the 43 cases identified through retrospective review, 8 tumors demonstrated ≥10% fat on MRI, and 4 tumors demonstrated ≥20% fat (highest fat percentage, 38%). There was no significant survival difference between patients with high tumor fat, which was defined as ≥20% fat, compared with those with little to no tumor fat. Myxoid liposarcomas may demonstrate a higher fat content on MRI than has previously been reported in the literature. Increased tumor fat percentage in lipid-rich tumors was not found to be associated with increased risk of death. Radiologists must be aware of the existence of MLS lesions with higher fat content.

Rational and Objectives This study aims to evaluate the diagnostic utility of breast imaging using transmission ultrasound. We present readers' accuracy in determining whether a breast lesion is a cyst versus a solid using transmission ultrasound as an adjunct to mammography. Materials and Methods This retrospective multi-reader, multi-case receiver operating characteristic study included 37 lesions seen on mammography and transmission ultrasound. Cyst cases were confirmed as cysts using their appearance on handheld ultrasound. Solid cases were confirmed as solids with pathology results. Fourteen readers performed blinded, randomized reads with mammography + quantitative transmission scan images, assigning both a confidence score (0–100) and a binary classification of cyst or solid. A 95% percentile bootstrap confidence interval (CI) was computed for the readers' mean receiver operating characteristic area, sensitivity, and specificity. Results Using the readers' binary classification of cyst or solid lesions, the mean sensitivity and specificity were 0.933 [95% CI: 0.837, 0.995] and 0.858 [95% CI: 0.701, 0.985], respectively. When the readers' confidence scores were used to distinguish a cyst versus solid, the mean receiver operating characteristic area was 0.920 [95% CI: 0.827, 0.985]. Conclusions Transmission ultrasound can provide an accurate assessment of a cyst versus a solid lesion in the breast. Prospective clinical trials will further delineate the role of transmission ultrasound as an adjunct to mammography to increase specificity in breast evaluation.

Purpose We did a comparative analysis of matched and mismatched defects in pre- and post-operative V/Q scans in CTEPH patients. We correlated the number of these defects with pre-operative clinical and hemodynamic parameters. Methods This was a retrospective study on 27 patients with CTEPH who underwent surgery. Pre- and post-operative V/Q scans were graded for each lung segment as normal, matched or mismatched defect. Additional pre- and post-operative clinical and hemodynamic parameters that were collected include New York Heart Association functional class, six-minute walk distance in feet, N-terminal pro b-type natriuretic peptide, forced expiratory volume in one second/forced vital capacity, diffusing capacity of the lung for carbon monoxide, pulmonary arterial pressure (systolic, diastolic and mean), right atrial pressure, cardiac output and cardiac index. Pulmonary vascular resistance was then calculated. Results On a segmental basis, 176 mismatched defects were noted in 27 patients, of which 111 improved post-surgery (63%). 22 of the 34 matched defects improved following surgery (64%). 31 new mismatched defects were observed. The number of pre-operative matched defects per patient ranged from 0 to 6. No statistically significant associations were observed between the number of pre-operative matched defects and pre-operative clinical parameters. No statistically significant associations were observed between the number of improved matched defects and the change in clinical parameters (pre- to post-surgery). Conclusion Both matched and mismatched defects on preoperative V/Q scans can show normalization post-surgery. The extent of matched defects on a preoperative V/Q scan does not correlate significantly with other clinical and hemodynamic parameters.

The objective of this study was to measure the accuracy and interobserver agreement of the Prostate Imaging Reporting and Data System, version 2 (PI-RADSv2), for the characterization of prostate lesions on multiparametric MRI.This retrospective study included 170 men examined at a single institution between August 2014 and February 2015 on a 3-T MRI scanner. Study patients were found to have lesions concerning for prostate cancer that were targeted for MRI/transrectal ultrasound fusion biopsy. Two experienced readers independently assigned a PI-RADSv2 assessment category to the dominant lesion in each patient. The AUC was calculated to determine reader accuracy for the detection of clinically significant prostate cancer (Gleason score ≥ 3 + 4). The Cohen kappa statistic was used to quantify interobserver agreement.The prevalence of clinically significant prostate cancer was 0.36 (61/170 patients). The AUCs for readers 1 and 2 were 0.871 and 0.882, respectively. The AUCs were greater for peripheral zone lesions than for transition zone lesions. When a PI-RADSv2 assessment category ≥ 3 was considered positive, the agreement between readers was good overall (κ = 0.63) and was fair for transition zone lesions (κ = 0.53). When a PI-RADSv2 assessment category ≥ 4 was considered positive, the agreement was excellent overall (κ = 0.91) and was excellent for both peripheral zone lesions (κ = 0.91) and transition zone lesions (κ = 0.87).Two experienced readers were able to accurately identify patients with clinically significant prostate cancer using PI-RADSv2 with good interobserver agreement overall.

Transjugular intrahepatic portosystemic shunts (TIPS) were historically placed using uncovered bare-metal stents. Current practice has now shifted toward the use of polytetrafluoroethylene (PTFE)-covered stents, given the improved primary patency seen with these stents. The aim of this study was to determine whether there is any added value, such as overall survival or stent patency, when using covered stents versus uncovered stents in TIPS placement in a large cohort. From April 1995 to June 2012, a total of 744 consecutive adult patients underwent de novo TIPS placement (378 receiving uncovered stents, 366 receiving covered stents). Information was obtained on demographics, baseline clinical variables, and outcomes after TIPS placement. Data were collected, compared, and analyzed to assess outcomes including mortality, primary patency (determined via repeat intervention), and secondary patency (determined via ultrasound parameters). Covered stents were associated with significantly improved primary patency (P

The purpose of this study was to investigate and validate adaptation of a cardiovascular CT angiography contrast injection protocol for lower tube potential.Eighty-three patients evaluated for thoracic aortic disease with a 256-MDCT scanner were imaged at 120 kV (group 1) or 100 kV (group 2) with the same contrast protocol (90 mL iopromide 370 mg I/mL at 3.5 mL/s). A pharmacokinetic model was validated and used to simulate aortic attenuation in group 2 patients with 20%, 33%, and 44% reduction in contrast volume. A 44% volume reduction was applied to 50 additional patients who underwent imaging at 100 kV (group 3). Patient characteristics, scanning and radiation parameters, and objective and subjective image indexes were compared among groups.Group 2 patients had higher mean aortic blood attenuation (399±61 HU) than group 1 patients (281±48 HU) (p0.001) but similar image noise. Group 3 and group 1 patients had similar mean aortic attenuation and noise. Subjective assessment of image quality indicated that group 3 and group 1 had comparable percentages of images with good or excellent diagnostic confidence scores (reader 1, 98% vs 96%; reader 2, 96% vs 96%).Lower tube potential (100 kV) for cardiothoracic CT could be accompanied by a 44% reduction in contrast volume with satisfactory aortic blood-pool attenuation in most patients. More personalized adaptation of the contrast protocol that takes into account patient characteristics and tube potential is necessary to ensure sufficient contrast enhancement for all patients.

The purpose of this study was to evaluate the impact of attenuation-based kilovoltage (kV) pair selection in dual source dual energy (DSDE)-pulmonary embolism (PE) protocol examinations on radiation dose savings and image quality. A prospective study was carried out on 118 patients with suspected PE. In patients in whom attenuation-based kV pair selection selected the 80/140Sn kV pair, the pre-scan 100/140Sn CTDIvol (computed tomography dose index volume) values were compared with the pre-scan 80/140Sn CTDIvol values. Subjective and objective image quality parameters were assessed. Attenuation-based kV pair selection switched to the 80/140Sn kV pair (“switched” cohort) in 63 out of 118 patients (53%). The mean 100/140Sn pre-scan CTDIvol was 8.8 mGy, while the mean 80/140Sn pre-scan CTDIvol was 7.5 mGy. The average estimated dose reduction for the “switched” cohort was 1.3 mGy (95% CI 1.2, 1.4; p

Purpose To compare the diagnostic accuracy and image quality of computed tomographic (CT) enterographic images obtained at half dose and reconstructed with filtered back projection (FBP) and sinogram-affirmed iterative reconstruction (SAFIRE) with those of full-dose CT enterographic images reconstructed with FBP for active inflammatory terminal or neoterminal ileal Crohn disease. Materials and Methods This retrospective study was compliant with HIPAA and approved by the institutional review board. The requirement to obtain informed consent was waived. Ninety subjects (45 with active terminal ileal Crohn disease and 45 without Crohn disease) underwent CT enterography with a dual-source CT unit. The reference standard for confirmation of active Crohn disease was active terminal ileal Crohn disease based on ileocolonoscopy or established Crohn disease and imaging features of active terminal ileal Crohn disease. Data from both tubes were reconstructed with FBP (100% exposure); data from the primary tube (50% exposure) were reconstructed with FBP and SAFIRE strengths 3 and 4, yielding four datasets per CT enterographic examination. The mean volume CT dose index (CTDIvol) and size-specific dose estimate (SSDE) at full dose were 13.1 mGy (median, 7.36 mGy) and 15.9 mGy (median, 13.06 mGy), respectively, and those at half dose were 6.55 mGy (median, 3.68 mGy) and 7.95 mGy (median, 6.5 mGy). Images were subjectively evaluated by eight radiologists for quality and diagnostic confidence for Crohn disease. Areas under the receiver operating characteristic curves (AUCs) were estimated, and the multireader, multicase analysis of variance method was used to compare reconstruction methods on the basis of a noninferiority margin of 0.05. Results The mean AUCs with half-dose scans (FBP, 0.908; SAFIRE 3, 0.935; SAFIRE 4, 0.924) were noninferior to the mean AUC with full-dose FBP scans (0.908; P < .003). The proportion of images with inferior quality was significantly higher with all half-dose reconstructions than with full-dose FBP (mean proportion: 0.117 for half-dose FBP, 0.054 for half-dose SAFIRE 3, 0.054 for half-dose SAFIRE 4, and 0.017 for full-dose FBP; P < .001). Conclusion The diagnostic accuracy of half-dose CT enterography with FBP and SAFIRE is statistically noninferior to that of full-dose CT enterography for active inflammatory terminal ileal Crohn disease, despite an inferior subjective image quality. (©) RSNA, 2016 Online supplemental material is available for this article.

A major initiative of the Quantitative Imaging Biomarker Alliance (QIBA) is to develop standards-based documents called “Profiles”, which describe one or more technical performance claims for a given imaging modality. The term “actor” denotes any entity (device, software, person) whose performance must meet certain specifications in order for the claim to be met. The objective of this paper is to present the statistical issues in testing actors’ conformance with the specifications. In particular, we present the general rationale and interpretation of the claims, the minimum requirements for testing whether an actor achieves the performance requirements, the study designs used for testing conformity, and the statistical analysis plan. We use three examples to illustrate the process: apparent diffusion coefficient (ADC) in solid tumors measured by MRI, change in Perc 15 as a biomarker for the progression of emphysema, and percent change in solid tumor volume by CT as a biomarker for lung cancer progression.

Receiver operating characteristic (ROC) analysis is a tool used to describe the discrimination accuracy of a diagnostic test or prediction model. While sensitivity and specificity are the basic metrics of accuracy, they have many limitations when characterizing test accuracy, particularly when comparing the accuracies of competing tests. In this article we review the basic study design features of ROC studies, illustrate sample size calculations, present statistical methods for measuring and comparing accuracy, and highlight commonly used ROC software. We include descriptions of multi-reader ROC study design and analysis, address frequently seen problems of verification and location bias, discuss clustered data, and provide strategies for testing endpoints in ROC studies. The methods are illustrated with a study of transmission ultrasound for diagnosing breast lesions.

Objectives: Injuries to the shoulder are a common cause of missed playing time in professional baseball athletes. These injuries predominately involve the supraspinatus tendon or the glenoid labrum. The purpose of this project is to describe a series of throwing-related injuries to the subscapularis muscle-tendon complex that have been previously unreported amongst baseball players. Additionally, it is the goal to describe a mechanism of injury as well as a treatment program for these athletes. Methods: A retrospective review of shoulder magnetic resonance imaging (MRI) scans of players from 1 Major League Baseball organization over the course of 5 years was performed to identify cases with findings suggestive of a subscapularis injury. The MRIs were graded and the medical records were reviewed to assess clinical findings, treatment plans, and follow-up. In the injured cohort, preinjury baseline measurements of shoulder motion including external rotation at 90° of abduction were compared to measurements from a non-injured baseball athlete cohort to evaluate whether decrease external rotation is a risk factor for this injury. Results: A total of 133 MRI scans of the shoulder were evaluated. Eleven of the 133 MRIs demonstrated signal changes suggesting subscapularis injury. Ten of these 11 patients had clinical findings supporting a diagnosis of throwing-related subscapularis strain. There were 4 grade I, 4 grade II, and 2 grade III injuries. All injuries occurred in the inferior half of the subscapularis at the myotendinous junction. Risk of subscapularis injury increased with lower levels of dominant arm external rotation (odds ratio, 0.89; 95% CI, 0.82-0.94; P < .001). A threshold of dominant arm external rotation of less than 106° demonstrated sensitivity of 0.700 (95% CI, 0.392-0.897) and specificity of 0.951 (95% CI, 0.888-0.982) for subscapularis injury. All players were treated conservatively without surgery; only 1 did not return to play. Of the remaining players, the mean number of days missed because of injury was 27 (range, 11-61 days). Conclusion: Throwing-related subscapularis injuries occur in the inferior half of the muscle at the myotendinous junction. The subscapularis muscle fibers are oriented along a fan shape because of the relatively narrow insertion on the lesser tuberosity of the proximal humerus and a broader origin on the scapula with muscle fiber alignment dependent on arm position. When the thrower's arm is at 90 degrees of abduction and external rotation, the inferior fibers of the subscapularis comes to be aligned in the transverse plane across the glenohumeral joint and are placed on greatest stretch during the late cocking and early acceleration phases of throwing. Our data suggest that there is an increased risk of these injuries with lower levels of dominant arm external rotation as this shorter arc requires greater force production by the subscapularis muscle to achieve the same peak throwing velocity.

The purpose of this study is to retrospectively evaluate the prevalence and extent of Celect inferior vena cava (IVC) filter penetration.All patients with Celect filters who underwent CT between 2007 and 2012 were assessed for penetration and related complications.Of the 690 patients with Celect filters placed at our institution, 265 underwent CT for various reasons. The mean (SD) interval between filter placement and last CT was 167.8±264.8 days (median, 56 days; range, 0-1592 days). Penetration of primary leg was observed in 39% (95% CI, 29-51%) of patients within 30 days and 80% (95% CI, 70-87%) of patients within 90 days after placement. The mean number of penetrated legs per patient was 1.8 at 30 days and 2.1 at 90 days after placement. Penetration into adjacent organs occurred in 35 of 265 (13.2%) filters (in 45 IVC filter limbs); penetration into two structures occurred with 10 filters. The sites involved were the duodenum (n=22), aorta (n=9), psoas muscle (n=4), vertebral body (n=3), pancreas (n=2), adrenal gland (n=1), liver (n=1), right kidney (n=1), lymph node (n=1), and diaphragm (n=1). One patient presented with abdominal pain related to penetration.This study confirms a high penetration rate for Celect IVC filters and shows that CT can characterize the extent of leg penetration. Most cases of penetration were asymptomatic, but the likelihood of penetration increased over time.

Purpose To investigate the prevalence and clinical sequelae of G2 filter (Bard Peripheral Vascular, Tempe, Arizona) fractures and fragment migration. Materials and Methods Patients who underwent G2 filter placement between October 2005 and February 2010 were assessed for filter fractures and complications. Fracture prevalence was estimated at various time points based on data from patients with known fracture status. Results Among 829 patients who underwent G2 filter placement, 684 had follow-up imaging and qualified for the study (381 men and 303 women; average age, 60.3 y; range, 15.8–95.2 y). For 541 (79%) patients, at least one image was available that contained the filter (imaging follow-up interval, 14.9 mo ± 20.0; range, 0–78.6 mo); images that did not include the filter but may have shown the migrated fracture fragment were available for 143 (21%) patients (follow-up interval, 11.2 mo ± 19.3; range, 0–83.4 mo). There were 16 fractured limbs identified in 13 patients (incidence, 1.9%; follow-up interval, 30.4 mo ± 18.7; range, 5.5–76.5 mo). Fracture fragments were identified in the pulmonary arteries (n = 4), right ventricle (n = 2), pericardium (n = 1), iliac vein (n = 1), and kidney (n = 1). Four fracture limbs remained near the filter; the remaining three could not be located. All patients with filter fracture were asymptomatic. The estimated 5-year fracture prevalence was 38% (95% confidence interval, 22.9%, 54.8%). Conclusions The early occurrence of G2 filter fractures was low, but the incidence increased over time. No life-threatening events occurred in patients with filter fracture during the study time frame.