Faculty Bibliography

OBJECTIVE:To use Twitter to characterize public perspectives regarding artificial intelligence (AI) and radiology. METHODS AND MATERIALS/METHODS:Twitter was searched for all tweets containing the terms "artificial intelligence" and "radiology" from November 2016 to October 2017. Users posting the tweets, tweet content, and linked websites were categorized. RESULTS:Six hundred and five tweets were identified. These were from 407 unique users (most commonly industry-related individuals [22.6%]; radiologists only 9.3%) and linked to 216 unique websites. 42.5% of users were from the United States. The tweets mentioned machine/deep learning in 17.2%, industry in 14.0%, a medical society/conference in 13.4%, and a university in 9.8%. 6.3% mentioned a specific clinical application, most commonly oncology and lung/tuberculosis. 24.6% of tweets had a favorable stance regarding the impact of AI on radiology, 75.4% neutral, and none were unfavorable. 88.0% of linked websites leaned toward AI being positive for the field of radiology; none leaned toward AI being negative for the field. 51.9% of linked websites specifically mentioned improved efficiency for radiology with AI. 35.2% of websites described challenges for implementing AI in radiology. Of the 47.2% of websites that mentioned the issue of AI replacing radiologists, 77.5% leaned against AI replacing radiologists, 13.7% had a neutral view, and 8.8% leaned toward AI replacing radiologists. CONCLUSION/CONCLUSIONS:These observations provide an overview of the social media discussions regarding AI in radiology. While noting challenges, the discussions were overwhelmingly positive toward the transformative impact of AI on radiology and leaned against AI replacing radiologists. Greater radiologist engagement in this online social media dialog is encouraged.

RATIONALE AND OBJECTIVES/OBJECTIVE:To explore downstream costs associated with incidental pulmonary nodules detected on CT. MATERIALS AND METHODS/METHODS:The cohort comprised 200 patients with an incidental pulmonary nodule on chest CT. Downstream events (chest CT, PET/CT, office visits, percutaneous biopsy, and wedge resection) were identified from the electronic medical record. The 2017 Fleischner Society Guidelines were used to classify radiologists' recommendations and ordering physician management for the nodules. Downstream costs for nodule management were estimated from national Medicare rates, and average costs were determined. RESULTS:Average downstream cost per nodule was $393. Costs were greater when ordering physicians over-managed relative to radiologist recommendations ($940) vs. when adherent ($637) or under-managing ($166) relative to radiologists recommendations. Costs were also greater when ordering physicians over-managed relative to Fleischner Society guidelines ($860) vs. when under-managing ($208) or adherent ($292) to guidelines. Costs did not vary significantly based on whether or not radiologists recommended follow-up imaging ($167-$397), nor whether radiologists were adherent or under- or over-recommended relative to Fleischner Society guidelines ($313-$444). Costs were also higher in older patients, patients with a smoking history, and larger nodules. Five nodules underwent wedge resection and diagnosed as malignancies. No patient demonstrated recurrence or metastasis. Average cost per diagnosed malignancy was $3090. CONCLUSION/CONCLUSIONS:Downstream costs for incidental pulmonary nodules are highly variable and particularly high when ordering physicians over-manage relative to radiologist recommendations and Fleischner Society guidelines. To reduce unnecessary utilization and cost from over-management, radiologists may need to assume a greater role in partnering with ordering physicians to ensure appropriate, guideline-adherent, and follow-up testing.

Radiologists play a critical role in helping the health care system achieve greater value. Unfortunately, today radiology is often judged by simple "checkbox" metrics, which neither directly reflect the value radiologists provide nor the outcomes they help drive. To change this system, first, we must attempt to better define the elusive term value and, then, quantify the value of imaging through more relevant and meaningful metrics that can be more directly correlated with outcomes. This framework can further improve radiology's value by enhancing radiologists' integration into the care team and their engagement with patients. With these improvements, we can maximize the value of imaging in the overall care of patients.

The publisher regrets that this article has been temporarily removed. A replacement will appear as soon as possible in which the reason for the removal of the article will be specified, or the article will be reinstated. The full Elsevier Policy on Article Withdrawal can be found at https://www.elsevier.com/about/our-business/policies/article-withdrawal.

BACKGROUND:The value of dynamic contrast-enhanced (DCE) sequences in prostate MRI compared with noncontrast MRI is controversial. PURPOSE/OBJECTIVE:To evaluate the population net benefit of risk stratification using DCE-MRI for detection of high-grade prostate cancer (HGPCA), with or without high spatiotemporal resolution DCE imaging. STUDY TYPE/METHODS:Decision curve analysis. POPULATION/METHODS:Previously published patient studies on MRI for HGPCA detection, one using DCE with golden-angle radial sparse parallel (GRASP) images and the other using standard DCE-MRI. FIELD STRENGTH/SEQUENCE/UNASSIGNED:GRASP or standard DCE-MRI at 3 T. ASSESSMENT/RESULTS:Each study reported the proportion of lesions with HGPCA in each Prostate Imaging Reporting and Data System version 2 (PI-RADS v2) category (1-5), before and after reclassification of peripheral zone lesions from PI-RADS 3-4 based on contrast-enhanced images. This additional risk stratifying information was translated to population net benefit, when biopsy was hypothetically performed for: all lesions, no lesions, PI-RADS ≥3 (using NC-MRI), and PI-RADS ≥4 on DCE. STATISTICAL TESTS/UNASSIGNED:Decision curve analysis was performed for both GRASP and standard DCE-MRI data, translating the avoidance of unnecessary biopsies and detection of HGPCA to population net benefit. We standardized net benefit values for HGPCA prevalence and graphically summarized the comparative net benefit of biopsy strategies. RESULTS:For a clinically relevant range of risk thresholds for HGPCA (>11%), GRASP DCE-MRI with biopsy of PI-RADS ≥4 lesions provided the highest net benefit, while biopsy of PI-RADS ≥3 lesions provided highest net benefit at low personal risk thresholds (2-11%). In the same range of risk thresholds using standard DCE-MRI, the optimal strategy was biopsy for all lesions (0-15% risk threshold) or PI-RADS ≥3 on NC-MRI (16-33% risk threshold). DATA CONCLUSION/UNASSIGNED:GRASP DCE-MRI may potentially enable biopsy of PI-RADS ≥4 lesions, providing relatively preserved detection of HGPCA and avoidance of unnecessary biopsies compared with biopsy of all PI-RADS ≥3 lesions. J. Magn. Reson. Imaging 2019.

PURPOSE/OBJECTIVE:The aim of this study was to assess national and state-specific changes in emergency department (ED) chest imaging utilization from 1994 to 2015. METHODS:Using aggregate 100% Medicare Physician/Supplier Procedure Summary Master Files for 1994 to 2015, the annual frequency of chest imaging in Medicare Part B beneficiaries in the ED setting was identified, and utilization was normalized to annual Medicare enrollment as well as annual ED visits. Using individual Medicare beneficiary 5% research-identifiable files, similar determinations were performed for each state. RESULTS:Between 1994 and 2015, per 1,000 beneficiaries, ED utilization of chest radiography and CT increased by 173% (compound annual growth rate [CAGR] 4.9%) and 5,941.8% (CAGR 21.6%). Per 1,000 ED visits, utilization increased by 81% (CAGR 2.9%) and 3,915.4% (CAGR 19.2%), respectively. Across states, utilization was highly variable, with 2015 radiography utilization per 1,000 ED visits ranging from 82 (Wyoming) to 731 (Hawaii) and CT utilization ranging from 18 (Wyoming) to 76 (Hawaii). Between 2004 and 2015, most states demonstrated increases in the utilization of both radiography (maximal increase of CAGR 11.0% in Vermont) and CT (maximal increase of CAGR 21.0% in Maine). Nonetheless, utilization of radiography declined in four states and utilization of CT in a single state. CONCLUSIONS:Over the past two decades, ED utilization of chest imaging has increased. This was related not only to an increasing frequency of ED visits but also to increasing utilization per ED visit. Across states, utilization is highly variable, but with radiography and CT both increasing, the use of CT seems additive to, rather than replacing, radiography.

OBJECTIVE. The purpose of this article is to discuss the evolution of MRI in prostate cancer from the early 1980s to the current day, providing analysis of the key studies on this topic. CONCLUSION. The rapid diffusion of MRI technology has meant that residual variability remains between centers regarding the quality of acquisition and the quality and standardization of reporting.

OBJECTIVE:Recent research on inferior vena cava (IVC) filter utilization in the United States has largely focused on national aggregate Medicare datasets, showing recent declines. Whether these national Medicare trends are generalizable across regions and payer populations is unknown. We studied recent state-level variation in IVC filter utilization across both Medicare and private insurance populations. MATERIALS AND METHODS/METHODS:Using large individual beneficiary claims-level Medicare research identifiable files and a proprietary U.S. research database of the commercially insured population, we identified all billed IVC filter placement procedures performed between 2009 and 2015. We compared population-adjusted utilization rates by state and payer type. RESULTS:Between 2009 and 2015, IVC filter utilization across the United States declined by 36.3% (from 177.9 to 113.3 procedures per 100,000 beneficiaries) in the Medicare population and by 26.6% (from 32.7 to 24.0 procedures per 100,000 beneficiaries) in the privately insured population. For the Medicare population, state-level utilization rates varied 5.2-fold, from 48.4 to 251.3 procedures per 100,000 beneficiaries in Alaska and New Jersey, respectively. For the private insurance population, rates varied 5.5-fold, from 10.8 to 59.5 procedures per 100,000 beneficiaries in Oregon and Michigan, respectively. Nationally, utilization in the Medicare population was 5.0 times higher than that in the private insurance population (range by state, from 2.0 times higher in Hawaii to 11.1 times higher in Utah). Despite the national decline, utilization in Medicare and private insurance populations increased in five and seven states, respectively. State-level IVC filter utilization rates for the Medicare population correlated strongly with those for the privately insured population (r = 0.74; p < 0.001). In both the Medicare and privately insured populations, utilization rates correlated moderately with beneficiary age (r = 0.44 and r = 0.50, respectively; p < 0.001 for both). CONCLUSION/CONCLUSIONS:IVC filter utilization rates vary dramatically by state and payer population, and they likely depend in part on the age of the covered population. To better identify demographic and socioeconomic drivers of utilization, future research should prioritize nonaggregate multipayer claims-level approaches.

Technical advances in MRI have improved image quality and have led to expanding clinical indications for its use. However, long examination and interpretation times, as well as higher costs, still represent barriers to use of MRI. Abbreviated MRI protocols have emerged as an alternative to standard MRI protocols. These abbreviated MRI protocols seek to reduce longer MRI protocols by eliminating unnecessary or redundant sequences that negatively affect cost, MRI table time, patient comfort, image quality, and image interpretation time. However, the diagnostic information is generally not compromised. Abbreviated MRI protocols have already been used successfully for hepatocellular carcinoma screening, for prostate cancer detection, and for screening for nonalcoholic fatty liver disease as well as monitoring patients with this disease. It has been reported that image acquisition time and costs can be considerably reduced with abbreviated MRI prototcols, compared with standard MRI protocols, while maintaining a similar sensitivity and accuracy. Nevertheless, multiple applications still need to be explored in the abdomen and pelvis (eg, surveillance of metastases to the liver; follow-up of cystic pancreatic lesions, adrenal incidentalomas, and small renal masses; evaluation of ovarian cysts in postmenopausal women; staging of cervical and uterine corpus neoplasms; evaluation of müllerian duct anomalies). This article describes some successful applications of abbreviated MRI protocols, demonstrates how they can help in improving the MRI workflow, and explores potential future directions. ^©RSNA, 2019.