The imaging diagnosis of cholecystitis in the adult ED: a comparative multi-reader, multivariable analysis of CT and US image features.

PURPOSE: The purposes of this study are (1) to utilize multivariable logistic regression in order to evaluate which image feature combination is most predictive in the diagnosis of cholecystitis for computed tomography (CT) and ultrasound (US) in adult ED patients and (2) to use these results to compare the accuracy of CT and US. METHODS: For RUQ pain patients undergoing US and CT at the same visit, multiple image features were evaluated independently by 2 radiologists blinded to additional data. Inter-reader variability was measured with the Kappa statistic. Sonographic Murphy's sign (SMS) information was obtained from original reports. Multivariable logistic regression was utilized to develop optimal predictive models for each modality. For US, models with/without SMS were compared to establish its relative value. RESULTS: 446 patients met inclusion criteria. For CT, the combination of cholelithiasis, short-axis gallbladder diameter > 3 cm, pericholecystic fluid or inflammation, and mural thickening > 3 mm provided the optimal model for both readers. For US, the optimal model included cholelithiasis, short-axis diameter > 3 cm, mural heterogeneity/striation, and sludge/debris for both readers. Kappa = 0.79-0.96 for included image features. For both readers, CT and US models had equivalent diagnostic performances; the SMS did not contribute significantly to US models. CONCLUSION: For a diagnosis of cholecystitis in the ED, (1) the optimal image feature combination for CT is cholelithiasis, short-axis diameter > 3 cm, pericholecystic fluid or inflammation, mural thickening > 3 mm; and cholelithiasis, short-axis diameter > 3 cm, mural heterogeneity/striation, sludge/debris for US; (2) CT and US have equivalent diagnostic performance; (3) inter-reader reliability is substantial to excellent for utilized image features; (4) the SMS does not affect US model accuracy.

55 Mn-based fiducial markers for rapid and automated RF coil localization for hyperpolarized 13 C MRI.

PURPOSE: To use fiducial markers containing manganese 55 to rapidly localize carbon 13 (13 C) RF coils for correcting images for B1 variation. METHODS: Hollow high-density polyethylene spheres were filled with 3M sodium permanganate and affixed to a rectangular 13 C-tuned RF coil. The relative positions of the markers and coil conductors were mapped using CT. Marker positions were measured by MRI using a series of 1D projections and automated peak detection. Once the coil location was determined, coil sensitivity was estimated using a quasi-static calculation. Simulations were performed to determine the minimum number of projections required for robust localization. Phantom experiments were used to confirm the accuracy of marker localization as well as the calculated coil sensitivity. Finally, in vivo validation was performed using hyperpolarized 13 C pyruvate in a rat model. RESULTS: In simulations, our algorithm was accurate in determining marker positions when at least 6 projections were used (RMSE 1.4 ± 0.9 mm). These estimates were verified in phantom experiments, where markers locations were determined with an RMS accuracy of 1.3 mm. A minimum SNR of 4 was required for automated detection to perform accurately. Computed coil sensitivity had a median error of 17% when taken over the entire measured area and 5.7% over a central region. In a rat, correction for nonuniform reception and flip angle was able to normalize the signals arising from asymmetrically positioned kidneys. CONCLUSION: Manganese 55 fiducial markers are an inexpensive and reliable method for rapidly localizing 13 C RF coils and correcting 13 C images for B1 variation without user intervention.

Role of Ultrasound and CT in the Workup of Right Upper Quadrant Pain in Adults in the Emergency Department: A Retrospective Review of More Than 2800 Cases.

OBJECTIVE. The purpose of this study was to compare ultrasound and CT in the diagnostic evaluation of right upper quadrant pain in adults in the emergency department. MATERIALS AND METHODS. A retrospective review was conducted of adult emergency department encounters for right upper quadrant pain over a 5-year period, excluding those for prior cholecystectomy, current pregnancy, and trauma. Imaging study reports were reviewed for gallbladder and nongallbladder explanations of right upper quadrant pain and were considered positive for cholecystitis when two or more supportive features were present. Encounter outcomes were evaluated on the basis of pathology data and clinical management. RESULTS. Among the encounters, 2859 met the study inclusion criteria, and 18% met the study definition of cholecystitis by pathologic or clinical criteria. The following metrics showed no statistically significant difference between ultrasound and CT, respectively: sensitivity, 61% and 55%; specificity, 91% and 92%; positive predictive value, 63% and 63%; and negative predictive value, 91% and 90%. Both modalities were performed in 20% of encounters. In the subgroup analysis, ultrasound showed an acute nongallbladder abnormality that was missed at a preceding CT examination in only 1 of 238 cases (0.4%). CT showed an acute nongallbladder abnormality missed at a preceding ultrasound examination in 103 of 322 cases (32%). CONCLUSION. CT is noninferior to ultrasound in both ruling in and ruling out the diagnosis of cholecystitis in adult patients undergoing emergency evaluation of right upper quadrant pain and offers the advantage of depicting acute nongallbladder abnormalities.

ACR TI-RADS: Pitfalls, Solutions, and Future Directions.

The high prevalence of thyroid nodules combined with the generally indolent growth of thyroid cancer present a challenge for optimal patient care. Risk classification models based on US features have been created by multiple professional societies, including the American College of Radiology (ACR), which published the Thyroid Imaging Reporting and Data System (TI-RADS) in 2017. ACR TI-RADS uses a standardized lexicon for assessment of thyroid nodules to generate a numeric scoring of features, designate categories of relative probability of benignity or malignancy, and provide management recommendations, with the aim of reducing unnecessary biopsies and excessive surveillance. Adopting ACR TI-RADS may require practice-level changes involving image acquisition and workflow, interpretation, and reporting. Significant resources should be devoted to educating sonographers and radiologists to accurately recognize features that contribute to the scoring of a nodule. Following a system that uses approved terminology generates reproducible and relevant reports while providing clarity of language and preventing misinterpretation. Comprehensive documentation facilitates quality improvement efforts. It also creates opportunities for outcome data and other performance metrics to be integrated with research. The authors review ACR TI-RADS, describe challenges and potential solutions related to its implementation based on their experiences, and highlight possible future directions in its evolution. ©RSNA, 2019 See discussion on this article by Hoang.

From Inguinal Hernias to Spermatic Cord Lipomas: Pearls, Pitfalls, and Mimics of Abdominal and Pelvic Hernias.

Abdominal and pelvic hernias may be indolent and detected incidentally, manifest acutely with pain and distress, or cause chronic discomfort. Physical examination findings are often ambiguous and insufficient for optimal triage. Therefore, accurate anatomic delineation and identification of complications are critical for effective treatment planning. Imaging, particularly computed tomography, provides a vital understanding of the hernia's location and size, involved viscera, and severity of associated complications. Reader familiarity with the imaging appearances and anatomic landmarks of hernias is important for correct diagnosis, which may impact preoperative planning and reduce morbidity. This article reviews the appearance of anatomic structures in the abdominal wall and pelvis that are important for diagnosing common and uncommon abdominal and pelvic hernias, and it highlights key imaging features that are helpful for differentiating hernias, mimics, and their complications. Online DICOM image stacks are available for this article . ©RSNA, 2017.

Automatic generation of boundary conditions using demons nonrigid image registration for use in 3-D modality-independent elastography.

Modality-independent elastography (MIE) is a method of elastography that reconstructs the elastic properties of tissue using images acquired under different loading conditions and a biomechanical model. Boundary conditions are a critical input to the algorithm and are often determined by time-consuming point correspondence methods requiring manual user input. This study presents a novel method of automatically generating boundary conditions by nonrigidly registering two image sets with a demons diffusion-based registration algorithm. The use of this method was successfully performed in silico using magnetic resonance and X-ray-computed tomography image data with known boundary conditions. These preliminary results produced boundary conditions with an accuracy of up to 80% compared to the known conditions. Demons-based boundary conditions were utilized within a 3-D MIE reconstruction to determine an elasticity contrast ratio between tumor and normal tissue. Two phantom experiments were then conducted to further test the accuracy of the demons boundary conditions and the MIE reconstruction arising from the use of these conditions. Preliminary results show a reasonable characterization of the material properties on this first attempt and a significant improvement in the automation level and viability of the method.

Non-rigid registration of breast surfaces using the laplace and diffusion equations.

A semi-automated, non-rigid breast surface registration method is presented that involves solving the Laplace or diffusion equations over undeformed and deformed breast surfaces. The resulting potential energy fields and isocontours are used to establish surface correspondence. This novel surface-based method, which does not require intensity images, anatomical landmarks, or fiducials, is compared to a gold standard of thin-plate spline (TPS) interpolation. Realistic finite element simulations of breast compression and further testing against a tissue-mimicking phantom demonstrate that this method is capable of registering surfaces experiencing 6 - 36 mm compression to within a mean error of 0.5 - 5.7 mm.

Modality independent elastography (MIE): potential applications in dermoscopy.

The use of palpation information for skin disease characterization is not as commonly used as in other soft tissues, although mechanical differences within lesions have been noted. For example, regions of hyperkeratosis have the potential to transform into cancerous lesions and likely feature different material properties from those of surrounding normal tissue due to varying cytoarchitecture. As a result, the spatial distribution of lesion mechanical properties may serve to assist a diagnosis or enhance visualization of the complete extent of a cancerous region, i.e., accurate information regarding the margins of disease for surgical therapy. In this work, a multiresolution extension to a novel elastographic imaging method called Modality Independent Elastography (MIE) is used to characterize the mechanical properties of a skin-like phantom embedded with a mock stiff lesion. Simulation studies were also performed to investigate the potential for characterizing realistic melanoma lesions. Elasticity image reconstructions from the phantom experiments localized the stiff inclusion and had good correlation between the Young's modulus contrast ratio and experimental measurements from material testing. In addition, multiresolution MIE was shown to be a more robust framework than its single-resolution version. Results from the melanoma simulation demonstrate the potential for using multiresolution MIE with dermoscopic images.