Computer-assisted Reporting and Decision Support Increases Compliance with Follow-up Imaging and Hormonal Screening of Adrenal Incidentalomas.

OBJECTIVE: To assess the impact of using a computer-assisted reporting and decision support (CAR/DS) tool at the radiologist point-of-care on ordering provider compliance with recommendations for adrenal incidentaloma workup. METHOD: Abdominal CT reports describing adrenal incidentalomas (2014 - 2016) were retrospectively extracted from the radiology database. Exclusion criteria were history of cancer, suspected functioning adrenal tumor, dominant nodule size < 1 cm or ≥ 4 cm, myelolipomas, cysts, and hematomas. Multivariable logistic regression models were employed to predict follow-up imaging (FUI) and hormonal screening orders as a function of patient age and sex, nodule size, and CAR/DS use. CAR/DS reports were compared to conventional reports regarding ordering provider compliance with, frequency, and completeness of, guideline-warranted recommendations for FUI and hormonal screening of adrenal incidentalomas using Chi-square test. RESULT: Of 174 patients (mean age 62.4; 51.1% women) with adrenal incidentalomas, 62% (108/174) received CAR/DS-based recommendations versus 38% (66/174) unassisted recommendations. CAR/DS use was an independent predictor of provider compliance both with FUI (Odds Ratio [OR]=2.47, p = 0.02) and hormonal screening (OR=2.38, p = 0.04). CAR/DS reports recommended FUI (97.2%,105/108) and hormonal screening (87.0%,94/108) more often than conventional reports (respectively, 69.7% [46/66], 3.0% [2/66], both p <0.0001). CAR/DS recommendations more frequently included instructions for FUI time, protocol, and modality than conventional reports (all p <0.001). CONCLUSION: Ordering providers were at least twice as likely to comply with report recommendations for FUI and hormonal evaluation of adrenal incidentalomas generated using CAR/DS versus unassisted reporting. CAR/DS-directed recommendations were more adherent to guidelines than those generated without.

Artificial Intelligence Enabling Radiology Reporting.

The radiology reporting process is beginning to incorporate structured, semantically labeled data. Tools based on artificial intelligence technologies using a structured reporting context can assist with internal report consistency and longitudinal tracking. To-do lists of relevant issues could be assembled by artificial intelligence tools, incorporating components of the patient's history. Radiologists will review and select artificial intelligence-generated and other data to be transmitted to the electronic health record and generate feedback for ongoing improvement of artificial intelligence tools. These technologies should make reports more valuable by making reports more accessible and better able to integrate into care pathways.

Computer-Assisted Reporting and Decision Support in Standardized Radiology Reporting for Cancer Imaging.

PURPOSE: Recent advances in structured reporting are providing an opportunity to enhance cancer imaging assessment to drive value-based care and improve patient safety. METHODS: The computer-assisted reporting and decision support (CAR/DS) framework has been developed to enable systematic ingestion of guidelines as clinical decision structured reporting tools embedded within the radiologist's workflow. RESULTS: CAR/DS tools can reduce the radiology reporting variability and increase compliance with clinical guidelines. The lung cancer use-case is used to describe various scenarios of a cancer imaging structured reporting pathway, including incidental findings, screening, staging, and restaging or continued care. Various aspects of these tools are also described using cancer-related examples for different imaging modalities and applications such as calculators. Such systems can leverage artificial intelligence (AI) algorithms to assist with the generation of structured reports and there are opportunities for new AI applications to be created using the structured data associated with CAR/DS tools. CONCLUSION: These AI-enabled systems are starting to allow information from multiple sources to be integrated and inserted into structured reports to drive improvements in clinical decision support and patient care.

Rectal MRI after neoadjuvant chemoradiation therapy: a pictorial guide to interpretation.

Magnetic resonance imaging (MRI) is the current reference standard imaging modality for restaging rectal cancer after neoadjuvant chemoradiation and is used to guide clinical management decisions. This pictorial essay provides an illustrative atlas of the key MRI features used to assess rectal cancer after treatment. MRI findings of residual tumor including non-mucinous, mucinous, and signet-ring cell adenocarcinoma subtypes are correlated with histopathology. Imaging appearances of treatment changes that mimic residual tumor in the setting of confirmed pathological complete response at resection are illustrated. Treatment complications are also shown. Knowledge of these imaging findings and their importance may help radiologists comply with all elements of the structured reporting templates proposed by the Rectal Cancer Disease Focused Panel of the Society of Abdominal Radiology and by the European Society of Gastrointestinal and Abdominal Radiology.

A Radiology-focused Analysis of Transparency and Usability of Top U.S. Hospitals' Chargemasters.

RATIONALE AND OBJECTIVES: In 2019, Centers for Medicare and Medicaid Services enforced regulation from the Affordable Care Act, requiring all U.S. hospitals to publish standard hospital charges annually. This study assesses top U.S academic hospitals' chargemasters for selected advanced diagnostic imaging services and the usability of publicly available information to allow consumers to determine out-of-pocket costs. MATERIALS AND METHODS: Publicly available chargemasters and associated websites for the top 20 ranked hospitals in U.S. News and World Report were assessed for several features including: file format, inclusion of CPT codes, disclaimers on charges versus costs and professional fees, and tools allowing determination of actual out-of-pocket costs for selected advanced diagnostic imaging examinations. RESULTS: All hospitals had publicly available chargemasters, 90% of which were in Microsoft Excel format. Ten percent of chargemasters included CPT codes. All chargemaster websites had disclaimers regarding differences between charges versus patient costs; 20% had disclaimers regarding professional fees. 20% of hospitals provided out-of-pocket costs for uninsured patients or tools allowing out-of-pocket cost determination. Median (range) MR exam charges were: brain with and without contrast: $5375 ($834-$13,857), noncontrast knee: $3402 (4530-$6924); noncontrast lumbar spine: $ 3449 ($473-$7367). Median (range) CT exam charges were: noncontrast head: $1923 ($165-$4974), noncontrast chest: $1947 ($282-$2991); contrast abdomen/pelvis: $4307 ($486-$11,726). CONCLUSION: While all top-ranked hospitals had publicly available chargemasters, they rarely provided transparent information to allow patients to determine out-of-pocket costs for advanced diagnostic imaging services.

Frequency and Reliability of the Reversed Halo Sign in Patients With Septic Pulmonary Embolism Due to IV Substance Use Disorder.

OBJECTIVE. The purpose of this article is to assess the "reversed halo" sign in patients with septic pulmonary embolism (PE) due to IV substance use disorder. MATERIALS AND METHODS. A retrospective analysis was performed of chest CT scans obtained between 2007 and 2017 that had findings of septic PE associated with IV substance use disorder. Inclusion criteria were history of IV substance use disorder, findings of septic PE on chest CT scans, and confirmation of infection. Image analysis was performed by three radiologists to assess the frequency, appearance, and evolution of the reversed halo sign. Interreader agreement to characterize the reversed halo sign was assessed using kappa statistical analysis. The chi-square test was used to correlate reversed halo sign shape with evolution on follow-up scans. RESULTS. Of 62 patients who met the inclusion criteria (54.8% women; mean age, 32.8 ± 8.3 [SD] years), 59.7% (37/62) had reversed halo signs (κ = 0.837-0.958, p < 0.0001). The mean number of unique reversed halo signs per patient was 2.1 ± 1.7 (46.7% of patients had more than one reversed halo sign). Of 78 unique reversed halo signs, 93.6% (73/78) were peripherally located and 51.3% (40/78) were located at the lower lobe, 52.6% (41/78) were pyramidal and 47.4% (37/78) were round shaped, 89.7% (70/78) had central low-attenuation areas, and 34.6% (27/78) had internal reticulations. Cavitation developed in 37.2% (29/78) of reversed halo signs and more often in pyramid-shaped ones (70.8%, 17/24), whereas consolidation occurred in 30.8% (24/78) and more often in round-shaped ones (58.6%; 17/29, p = 0.03). CONCLUSION. Septic PE should be considered in the differential diagnosis of patients with IV substance use disorder presenting with reversed halo sign. The reversed halo sign was reliably and frequently observed on the chest CT scans of patients with IV substance use disorder-related septic PE. Characteristics of reversed halo sign presentation were identified as potential features to differentiate septic PE from other causes of pulmonary infarct manifesting with reversed halo sign.

Artificial Intelligence and Clinical Decision Support for Radiologists and Referring Providers.

Recent advances in artificial intelligence (AI) are providing an opportunity to enhance existing clinical decision support (CDS) tools to improve patient safety and drive value-based imaging. We discuss the advantages and potential applications that may be realized with the synergy between AI and CDS systems. From the perspective of both radiologist and ordering provider, CDS could be significantly empowered using AI. CDS enhanced by AI could reduce friction in radiology workflows and can aid AI developers to identify relevant imaging features their tools should be seeking to extract from images. Furthermore, these systems can generate structured data to be used as input to develop machine learning algorithms, which can drive downstream care pathways. For referring providers, an AI-enabled CDS solution could enable an evolution from existing imaging-centric CDS toward decision support that takes into account a holistic patient perspective. More intelligent CDS could suggest imaging examinations in highly complex clinical scenarios, assist on the identification of appropriate imaging opportunities at the health system level, suggest appropriate individualized screening, or aid health care providers to ensure continuity of care. AI has the potential to enable the next generation of CDS, improving patient care and enhancing providers' and radiologists' experience.

Temporal Trends in Imaging Utilization for Suspected Substance Use Disorder in an Academic Emergency Radiology Department.

PURPOSE: To assess temporal trends and utilization patterns of diagnostic imaging performed for substance use disorder (SUD)-related indications in an academic radiology emergency department (ED). METHODS: Retrospective analyses of ED imaging examinations acquired from 2005 to 2015 were performed. Imaging examinations performed for suspected SUD-related indications, based on the order history, were compared with those without a SUD-related indication. Unadjusted analyses comparing demographic and imaging characteristics between SUD-related versus non-SUD-related indications used Wilcoxon and Pearson's χ2 tests. Multivariable logistic regression models, within each imaging modality subgroup and combined, were employed to examine the odds of imaging examinations having an SUD-related indication as a function of demographic and imaging characteristics. RESULTS: Among 938,245 examinations, 0.17% had an SUD-related indication. Patients with SUD-related indications were younger (mean 37.2 ± 11.1 versus 53.5 ± 22.4, P < .001) and more commonly male (65% versus 52%, P < .001). The proportions of MR (17%), spine (17%), and extremities (33%) studies performed for SUD-related indications were larger among SUD than non-SUD indications (6%, 8%, 26%, respectively, all P < .001). Regression analysis demonstrated the odds of acquiring an ED imaging examination with an SUD-related indication significantly increased over time (P < .001, adjusted odds ratio [aOR] = 1.06), which was most pronounced among MR (P < .001, aOR = 1.23). For all regression models, younger age, male gender, and body part being imaged were identified as independent predictors of an SUD-related indication for ED imaging. CONCLUSION: Imaging performed for an SUD-related indication represented a small but increasing subset of overall ED imaging. Utilization of MR for SUD-related indications significantly outpaced growth of MR without SUD-related indications.

Accuracy and timeliness of an abbreviated emergency department MRCP protocol for choledocholithiasis.

PURPOSE: To determine the diagnostic accuracy and time savings of an abbreviated magnetic resonance cholangiopancreatography (A-MRCP) protocol for detecting choledocholithiasis in patients visiting the emergency department (ED) for suspected biliary obstruction. METHODS AND MATERIALS: This retrospective study evaluated adult patients (ages 18+ years) visiting an academic Level 1 trauma center between January 1, 2016, and December 31, 2017, who were imaged with MRCP for suspected biliary obstruction. Patients were scanned with either a four-sequence A-MRCP protocol or a conventional eight-sequence MRCP (C-MRCP) protocol. Image acquisition and MRI room time were compared. The radiology report was used to determine whether a study was limited by motion or prematurely aborted, as well as for the presence of pertinent biliary findings. Diagnostic accuracy of A-MRCP studies were compared with any available endoscopic retrograde cholangiopancreatography (ERCP) report within 30 days. RESULTS: One hundred sixteen patients met inclusion criteria; 85 were scanned with the A-MRCP protocol (45.9% male, mean 57.4 years) and 31 with the C-MRCP protocol (38.7% male, mean 58.3 years). Mean image acquisition time and MRI room time for the A-MRCP protocol were significantly lower compared to those for the C-MRCP protocol (16 and 34 min vs. 42 and 61 min, both p < 0.0001). Choledocholithiasis was seen in 23.5% of A-MRCP cases and 19.4% of C-MRCP cases. Non-biliary findings were common in both cohorts, comprising 56.5% of A-MRCP cases and 41.9% of C-MRCP cases. 44.7% of A-MRCP patients received subsequent (diagnostic or therapeutic) ERCP (mean follow-up time 3 days), in which A-MRCP accurately identified choledocholithiasis in 86.8% of cases, with sensitivity of 85%, specificity of 88.9%, positive predictive value (PPV) of 89.5%, and negative predictive value (NPV) of 84.2%. In comparison, 38.7% of C-MRCP patients underwent ERCP (mean follow-up of 2.3 days) with an accuracy of 91.7%, sensitivity of 80%, specificity of 100%, PPV of 100%, and NPV of 87.5%. Only 4.7% of A-MRCP exams demonstrated motion artifact vs. 12.9% of C-MRCP exams. One study was prematurely aborted due to patient discomfort in the A-MRCP cohort while no studies were terminated in the C-MRCP cohort. CONCLUSION: An abbreviated MRCP protocol to evaluate for choledocholithiasis provides significant time savings and reduced motion artifact over the conventional MRCP protocol while providing similar diagnostic accuracy.

Impact of Radiology Report Wording on Care of Patients With Acute Epiploic Appendagitis.

OBJECTIVE. The purpose of this study was to evaluate the association between the diagnostic certainty expressed by the wording of CT report impressions and subsequent use of standard treatment with analgesics versus nonstandard antibiotic administration in patients with acute epiploic appendagitis (EA). MATERIALS AND METHODS. Demographic, clinical, and radiologic data from a 10-year cohort of patients with acute EA were retrospectively analyzed and correlated with standard treatment with analgesics versus nonstandard treatment with antibiotics. A level of certainty was assigned to the CT report language based on the wording of the impression statements by two radiologists; their interreader agreement was assessed with kappa statistics. Bivariate analyses were performed to correlate all variables with antibiotic administration and to assess for collinearity. Multivariate logistic regression was performed to identify independent predictors of antibiotic use in patients with acute EA. RESULTS. Of 124 patients with CT-diagnosed acute EA, 22% (27/124) received antibiotic treatment. After the CT report impressions were evaluated, 27% (34/124) were categorized as low certainty and 73% (90/124) as high certainty (κ = 0.958, p < 0.001). Multivariate regression was significant (p < 0.001, Nagelkerke R2 = 0.249) and found CT report impressions' level of certainty (odds ratio [OR] = 6.1, p < 0.001) and evaluation in an outpatient clinic rather than an emergency department (ED) (OR = 4.4, p = 0.003) to be independent predictors of antibiotic administration for patients with acute EA. Outpatient presentation was also correlated with age, abdominal pain duration, and left-colonic involvement in the bivariate analysis (all p ≤ 0.01). CONCLUSION. The diagnostic certainty conveyed by the wording of CT report impressions correlated with antibiotic treatment decisions for patients with acute EA. Patients whose report impressions expressed low rather than high certainty were six times more likely to receive antibiotic therapy; patients evaluated at outpatient clinics rather than EDs were four times more likely.

The added value of cross-sectional imaging in the detection of additional radiographically occult fractures in the setting of a Chopart fracture.

PURPOSE: Radiography has a low sensitivity for the detection of fractures related to the talonavicular and calcaneocuboid articulations, also known as Chopart fractures. The purpose of this study is to determine the sensitivity of radiographs for detecting additional foot and ankle fractures related to Chopart fracture using CT or MRI as the reference standard. METHOD: We performed an IRB-approved, retrospective review of radiology reports between 2010 and 2014. Inclusion criteria were (1) diagnosis of a Chopart fracture and (2) at least one radiograph and subsequent cross-sectional imaging (CT or MR). CT or MRI was considered the diagnostic reference standard. Results were stratified by the energy of trauma and by type of radiograph performed (weight-bearing (WB) versus non-WB). RESULTS: One hundred eight patients met the inclusion criteria. The calcaneocuboid articulation was the most commonly involved type of Chopart fracture, seen in 75% of cases (81/108). Chopart fractures were detected on the initial radiographs in 67.6% of cases (73/108). Additional fractures of the ankle and midfoot were diagnosed in 34.2% of cases (37/108), with 56.7% (21/37) of these cases having at least one additional fracture seen on CT or MRI that was not seen on the initial radiographs, with fractures of the midfoot most often missed. In 56.7% (17/30) patients whose radiographs detected Chopart fractures, at least one additional fracture was missed; 30% of them demonstrated intra-articular extension and 56.7% were considered displaced. High-energy trauma was related to higher incidence of additional fractures. There was no significant difference in the sensitivity of radiographs to detect additional fractures between high versus low-energy trauma (p = 0.3) and WB versus non-WB radiographs (p = 0.5). Most patients were treated nonoperatively (56.5%, 61/108), with surgical intervention more frequent in patients with a high energy of trauma (51.7% versus 33.3%, p = 0.05). CONCLUSION: In the setting of a Chopart fracture, CT or MRI can add significant value in the detection of additional ankle or midfoot fractures, irrespective of the energy of trauma. Since additional fractures can have important management implications, CT or MRI should be considered as part of the standard workup for all midfoot fractures.

Advances in Pancreatic CT Imaging.

OBJECTIVE: The purpose of this article is to discuss the advances in CT acquisition and image postprocessing as they apply to imaging the pancreas and to conceptualize the role of radiogenomics and machine learning in pancreatic imaging. CONCLUSION: CT is the preferred imaging modality for assessment of pancreatic diseases. Recent advances in CT (dual-energy CT, CT perfusion, CT volumetry, and radiogenomics) and emerging computational algorithms (machine learning) have the potential to further increase the value of CT in pancreatic imaging.

Complications of Intravenous Substance Use Disorders (SUDs): A Multimodality Pictorial Essay for Cardiothoracic Radiologists.

With the growing substance use disorders epidemic in the United States, a wide range of complications may be seen in both the acute and chronic clinical setting. This article will use a multimodality approach to review complications of intravenous substance use that cardiothoracic radiologists are likely to encounter. Radiologists should be aware of these multisystem complications that can affect the cardiovascular, respiratory, and musculoskeletal systems in order to make an accurate and timely diagnosis, which can drastically alter the management of these patients.

Quantifying the Impact of Noninterpretive Tasks on Radiology Report Turn-Around Times.

PURPOSE: Traditional radiology productivity metrics do not account for noninterpretive tasks (NITs). This study aimed to systematically quantify NITs and their impact on report turn-around time (RTAT) during solo academic neuroradiology overnight coverage in the emergency department. METHODS: Retrospective analysis of 1 week of data, including phone call quantity and duration, clinician identification badge access to the reading room ("badge swipes"), suspected acute strokes, imaging examination volume, and emergency department patient volume, was performed. Univariate analyses were employed to quantify NITs. Multivariate linear regression was used to determine if NITs within an hour are predictive of RTAT of studies completed within that hour. RESULTS: Sixty-three hours of overnight neuroradiology coverage were analyzed. The mean number of phone calls per hour was 8.7 (SD: 5.7), and mean duration of phone calls per hour was 12 min (SD: 9.6 min, range 1-46). The mean number of badge swipes per hour was 2.1 (SD 1.6). The mean number of examinations (CT and MRI) performed per hour was 2.2 (SD: 1.7). Regression analyses found total duration of phone calls in an hour as the strongest independent predictor of RTAT (unstandardized ß = 4.25, P < .001). The overall multivariate model was also significant (P < .001, R2 = 0.596; adjusted R2 = 0.578). CONCLUSIONS: For every 1-min increase in total duration of calls in an hour, mean RTAT increased by 4.25 min. Standardizing capture of NITs may aid development of strategies that address productivity, communication, and value in radiology.