Recent advances and future perspectives of machine learning techniques offer promising applications in medical imaging. Machine learning has the potential to improve different steps of the radiology workflow including order scheduling and triage, clinical decision support systems, detection and interpretation of findings, postprocessing and dose estimation, examination quality control, and radiology reporting. In this article, the authors review examples of current applications of machine learning and artificial intelligence techniques in diagnostic radiology. In addition, the future impact and natural extension of these techniques in radiology practice are discussed.

This copy is for personal use only. To order printed copies, contact reprints@rsna.org I n P r e s s Abbreviations: ICU = intensive care unit; ACE2 = angiotensin converting enzyme 2; COVID-19 = Coronavirus disease 2019; RUQ = right upper quadrant; SARS-CoV-2 = Severe acute respiratory syndrome coronavirus 2.Key Results: -33% of inpatients with COVID-19 had abdominal imaging and 17% had cross-sectional imaging. Imaging was associated with age (OR 1.03 per year increase) and intensive care unit (ICU) admission (OR 17.3). -54% of right upper quadrant ultrasounds demonstrated findings of cholestasis. -31% of CTs showed bowel wall abnormalities. Signs of late ischemia were seen on 20% of CTs in ICU patients (2.7% of ICU patients), with pathologic correlation suggesting small vessel thrombosis. Summary Statement: Bowel abnormalities, including ischemia, and cholestasis were common findings on abdominal imaging of inpatients with COVID-19. I n P r e s s Abstract:Background: Angiotensin converting enzyme 2 (ACE2), a target of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), demonstrates its highest surface expression in the lung, small bowel, and vasculature, suggesting abdominal viscera may be susceptible to injury.Purpose: To report abdominal imaging findings in patients with coronavirus disease 2019 . Materials and Methods:In this retrospective cross-sectional study, patients consecutively admitted to a single quaternary care center from 3/27/2020 to 4/10/2020 who tested positive for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) were included. Abdominal imaging studies performed in these patients were reviewed and salient findings recorded.Medical records were reviewed for clinical data. Univariable analysis and logistic regression were performed. Results: 412 patients (average age 57 years; range 18->90 years; 241 men, 171 women) were evaluated. 224 abdominal imaging studies were performed (radiographs, n=137; ultrasound, n=44; CT, n=42; MRI, n=1) in 134 patients (33%). Abdominal imaging was associated with age (odds ratio [OR] 1.03 per year increase, p=0.001) and ICU admission (OR 17.3, p<0.001). Bowel wall abnormalities were seen on 31% of CT scans (13 of 42) and were associated with ICU admission (OR 15.5, p=0.01). Bowel findings included pneumatosis or portal venous gas, seen on 20% of CT scans in ICU patients (4 of 20). Surgical correlation (n=4) revealed unusual yellow discoloration of bowel (n=3) and bowel infarction (n=2). Pathology demonstrated ischemic enteritis with patchy necrosis and fibrin thrombi in arterioles (n=2). Of right upper quadrant ultrasounds, 87% (32 of 37) were performed for liver laboratory findings, and 54% (20 of 37) demonstrated a dilated sludge-filled gallbladder suggestive of cholestasis. Patients with a cholecystostomy tube placed (n=4) had negative bacterial cultures. Conclusion: Bowel abnormalities and cholestasis were common findings on abdominal imaging of inpatients with COVID-19. Patients who went to laparotomy often had ischemia, possibly due to sma...

Tissue stiffness has long been known to be a biomarker of tissue pathology. Ultrasound elastography measures tissue mechanical properties by monitoring the response of tissue to acoustic energy. Different elastographic techniques have been applied to many different tissues and diseases. Depending on the pathology, patient-based factors, and ultrasound operator-based factors, these techniques vary in accuracy and reliability. In this review, we discuss the physical principles of ultrasound elastography, discuss differences between different ultrasound elastographic techniques, and review the advantages and disadvantages of these techniques in clinical practice.

).q RSNA, 2014 Purpose:To evaluate the accuracy of shear-wave elastography (SWE) for staging liver fibrosis in patients with diffuse liver disease (including patients with hepatitis C virus [HCV]) and to determine the relative accuracy of SWE measurements obtained from different hepatic acquisition sites for staging liver fibrosis. Materials and Methods:The institutional review board approved this single-institution prospective study, which was performed between January 2010 and March 2013 in 136 consecutive patients who underwent SWE before their scheduled liver biopsy (age range, 18-76 years; mean age, 49 years; 70 men, 66 women). Informed consent was obtained from all patients. SWE measurements were obtained at four sites in the liver. Biopsy specimens were reviewed in a blinded manner by a pathologist using METAVIR criteria. SWE measurements and biopsy results were compared by using the Spearman correlation and receiver operating characteristic (ROC) curve analysis. Results:SWE values obtained at the upper right lobe showed the highest correlation with estimation of fibrosis (r = 0.41, P , .001). Inflammation and steatosis did not show any correlation with SWE values except for values from the left lobe, which showed correlation with steatosis (r = 0.24, P = .004). Abbreviations: AUC = area under the ROC curve CI = confidence interval CLD = chronic liver disease DANA = difference between the mean fibrosis stage of advanced fibrosis and the mean fibrosis stage of nonadvanced fibrosis HCV = hepatitis C virus ROC = receiver operating characteristic SWE = shear-wave elastography Author contributions:Guarantors of integrity of entire study, A.E.S., M.D.; study concepts/study design or data acquisition or data analysis/ interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; agrees to ensure any questions related to the work are appropriately resolved, all authors; literature research, A.E.S., M.D., A.V., J.M.N., K.E.C.; clinical studies, A.E.S., A.V., A.K.B., J.M.N., K.E.C., R.T.C.; statistical analysis, M.D., E.F.H.; and manuscript editing, A.E.S., M.D., A.V., R.T.C. Funding:R.T.C. supported by the National Institutes of Health (grant DK078772).Conflicts of interest are listed at the end of this article. (1), with as many as 150 000 new cases diagnosed each year (2)-20% of which had cirrhosis at presentation. The multiple causes of CLD follow a common pathway of progressive liver fibrosis, ultimately culminating in cirrhosis. These include hepatitis C virus (HCV), hepatitis B virus, nonalcoholic fatty liver disease, and alcoholic liver disease (3). Although the prevalence of major causes of CLD remains stable, data from the National Health and Nutrition Examination Surveys show that nonalcoholic fatty liver disease will be a substantial burden on the prevalence of CLD in the United States (1). Advanced fibrosis, cirrhosis, and hepatocellular carcinoma develop in about 17%-55% of patients with HCV ...

Non-alcoholic fatty liver disease (NAFLD) is the most common diffuse liver disease, with a worldwide prevalence of 20% to 46%. NAFLD can be subdivided into simple steatosis and nonalcoholic steatohepatitis. Most cases of simple steatosis are non-progressive, whereas nonalcoholic steatohepatitis may result in chronic liver injury and progressive fibrosis in a significant minority. Effective risk stratification and management of NAFLD requires evaluation of hepatic parenchymal fat, fibrosis, and inflammation. Liver biopsy remains the current gold standard; however, non-invasive imaging methods are rapidly evolving and may replace biopsy in some circumstances. These methods include well-established techniques, such as conventional ultrasonography, computed tomography, and magnetic resonance imaging and newer imaging technologies, such as ultrasound elastography, quantitative ultrasound techniques, magnetic resonance elastography, and magnetic resonance-based fat quantitation techniques. The aim of this article is to review the current status of imaging methods for NAFLD risk stratification and management, including their diagnostic accuracy, limitations, and practical applicability.

Ultrasound (US) imaging is the most commonly performed cross-sectional diagnostic imaging modality in the practice of medicine. It is low-cost, non-ionizing, portable, and capable of real-time image acquisition and display. US is a rapidly evolving technology with significant challenges and opportunities. Challenges include high inter- and intra-operator variability and limited image quality control. Tremendous opportunities have arisen in the last decade as a result of exponential growth in available computational power coupled with progressive miniaturization of US devices. As US devices become smaller, enhanced computational capability can contribute significantly to decreasing variability through advanced image processing. In this paper, we review leading machine learning (ML) approaches and research directions in US, with an emphasis on recent ML advances. We also present our outlook on future opportunities for ML techniques to further improve clinical workflow and US-based disease diagnosis and characterization.

No abstract

BackgroundThere currently is a need for a non-invasive measure of renal fibrosis. We aim to explore whether shear wave elastography (SWE)-derived estimates of tissue stiffness may serve as a non-invasive biomarker that can distinguish normal and abnormal renal parenchymal tissue.MethodsParticipants with CKD (by estimated GFR) and healthy volunteers underwent SWE. Renal elasticity was estimated as Young’s modulus (YM) in kilopascals (kPa). Univariate Wilcoxon rank-sum tests were used.ResultsTwenty-five participants with CKD (median GFR 38 mL/min; quartile 1, quartile 3 28, 42) and 20 healthy controls without CKD underwent SWE performed by a single radiologist. CKD was associated with increased median YM (9.40 [5.55, 22.35] vs. 4.40 [3.68, 5.70] kPa; p = 0.002) and higher median intra-subject inter-measurement estimated YM’s variability (4.27 [2.89, 9.90] vs. 1.51 [1.21, 2.05] kPa; p < 0.001).ConclusionsSWE-derived estimates of renal stiffness and intra-subject estimated stiffness variability are higher in patients with CKD than in healthy controls. Renal fibrosis is a plausible explanation for the observed difference in YM. Further studies are required to determine the relationship between YM, estimated renal stiffness, and renal fibrosis severity.Electronic supplementary materialThe online version of this article (doi:10.1186/s12882-015-0120-7) contains supplementary material, which is available to authorized users.