Dual Energy CT for Deep Learning-Based Segmentation and Volumetric Estimation of Early Ischemic Infarcts.

Ischemic changes are not visible on non-contrast head CT until several hours after infarction, though deep convolutional neural networks have shown promise in the detection of subtle imaging findings. This study aims to assess if dual-energy CT (DECT) acquisition can improve early infarct visibility for machine learning. The retrospective dataset consisted of 330 DECTs acquired up to 48 h prior to confirmation of a DWI positive infarct on MRI between 2016 and 2022. Infarct segmentation maps were generated from the MRI and co-registered to the CT to serve as ground truth for segmentation. A self-configuring 3D nnU-Net was trained for segmentation on (1) standard 120 kV mixed-images (2) 190 keV virtual monochromatic images and (3) 120 kV + 190 keV images as dual channel inputs. Algorithm performance was assessed with Dice scores with paired t-tests on a test set. Global aggregate Dice scores were 0.616, 0.645, and 0.665 for standard 120 kV images, 190 keV, and combined channel inputs respectively. Differences in overall Dice scores were statistically significant with highest performance for combined channel inputs (p < 0.01). Small but statistically significant differences were observed for infarcts between 6 and 12 h from last-known-well with higher performance for larger infarcts. Volumetric accuracy trended higher with combined inputs but differences were not statistically significant (p = 0.07). Supplementation of standard head CT images with dual-energy data provides earlier and more accurate segmentation of infarcts for machine learning particularly between 6 and 12 h after last-known-well.

Optimizing Acute Stroke Segmentation on MRI Using Deep Learning: Self-Configuring Neural Networks Provide High Performance Using Only DWI Sequences.

Segmentation of infarcts is clinically important in ischemic stroke management and prognostication. It is unclear what role the combination of DWI, ADC, and FLAIR MRI sequences provide for deep learning in infarct segmentation. Recent technologies in model self-configuration have promised greater performance and generalizability through automated optimization. We assessed the utility of DWI, ADC, and FLAIR sequences on ischemic stroke segmentation, compared self-configuring nnU-Net models to conventional U-Net models without manual optimization, and evaluated the generalizability of results on an external clinical dataset. 3D self-configuring nnU-Net models and standard 3D U-Net models with MONAI were trained on 200 infarcts using DWI, ADC, and FLAIR sequences separately and in all combinations. Segmentation results were compared between models using paired t-test comparison on a hold-out test set of 50 cases. The highest performing model was externally validated on a clinical dataset of 50 MRIs. nnU-Net with DWI sequences attained a Dice score of 0.810 ± 0.155. There was no statistically significant difference when DWI sequences were supplemented with ADC and FLAIR images (Dice score of 0.813 ± 0.150; p = 0.15). nnU-Net models significantly outperformed standard U-Net models for all sequence combinations (p < 0.001). On the external dataset, Dice scores measured 0.704 ± 0.199 for positive cases with false positives with intracranial hemorrhage. Highly optimized neural networks such as nnU-Net provide excellent stroke segmentation even when only provided DWI images, without significant improvement from other sequences. This differs from-and significantly outperforms-standard U-Net architectures. Results translated well to the external clinical environment and provide the groundwork for optimized acute stroke segmentation on MRI.

Sociodemographic biases in a commercial AI model for intracranial hemorrhage detection.

PURPOSE: To evaluate whether a commercial AI tool for intracranial hemorrhage (ICH) detection on head CT exhibited sociodemographic biases. METHODS: Our retrospective study reviewed 9736 consecutive, adult non-contrast head CT scans performed between November 2021 and February 2022 in a single healthcare system. Each CT scan was evaluated by a commercial ICH AI tool and a board-certified neuroradiologist; ground truth was defined as final radiologist determination of ICH presence/absence. After evaluating the AI tool's aggregate diagnostic performance, sub-analyses based on sociodemographic groups (age, sex, race, ethnicity, insurance status, and Area of Deprivation Index [ADI] scores) assessed for biases. χ2 test or Fisher's exact tests evaluated for statistical significance with p ≤ 0.05. RESULTS: Our patient population was 50% female (mean age 60 ± 19 years). The AI tool had an aggregate accuracy of 93% [9060/9736], sensitivity of 85% [1140/1338], specificity of 94% [7920/ 8398], positive predictive value (PPV) of 71% [1140/1618] and negative predictive value (NPV) of 98% [7920/8118]. Sociodemographic biases were identified, including lower PPV for patients who were females (67.3% [62,441/656] vs. 72.7% [699/962], p = 0.02), Black (66.7% [454/681] vs. 73.2% [686/937], p = 0.005), non-Hispanic/non-Latino (69.7% [1038/1490] vs. 95.4% [417/437]), p = 0.009), and who had Medicaid/Medicare (69.9% [754/1078]) or Private (66.5% [228/343]) primary insurance (p = 0.003). Lower sensitivity was seen for patients in the third quartile of national (78.8% [241/306], p = 0.001) and state ADI scores (79.0% [22/287], p = 0.001). CONCLUSIONS: In our healthcare system, a commercial AI tool had lower performance for ICH detection than previously reported and demonstrated several sociodemographic biases.

Conference Attendance Tracking and Evaluation in the Era of Virtual Conferences.

RATIONALE AND OBJECTIVES: The coronavirus pandemic upended in-person radiology education and led to a transition to virtual platforms. We needed a new method to monitor lecture attendance, previously relying on a physical badge system. Our goal was to develop and implement a virtual conference attendance system that is user-friendly, automated, useable in any virtual conference environment, and accurate. MATERIALS AND METHODS: We developed a web-based platform to serve as a virtual conference attendance tracking and evaluation platform. Daily, the application synchronizes with our lecture calendar to identify scheduled conferences and generates a unique attendance link for each event. The link is automatically posted in the conference chat and attendees must be logged in by the time it is posted to click the link, prompting single sign-on authentication. We integrated the system with resident schedules to excuse residents when appropriate. Real-time attendance reports are accessible in a user-friendly dashboard with a 5-star lecture review and comment system. We surveyed residents on satisfaction with the application after 1-year of use. RESULTS: Over the 2020-2021 academic year, we registered 376 conferences with 5,040 virtual swipes from 65 users. Once set up, virtual swipes take seconds to perform with minimal disruption to the conference. Average satisfaction for the platform was rated as 4.69 on a scale of 1 to 5. All respondents agreed or strongly agreed that use of the platform should be continued for future years, with 85% strongly agreeing. CONCLUSION: We developed an online platform for radiology conference attendance logging and evaluation, designed for virtual conferences.

Prediction of Coronary Artery Calcium and Cardiovascular Risk on Chest Radiographs Using Deep Learning.

PURPOSE: To assess the ability of deep convolutional neural networks (DCNNs) to predict coronary artery calcium (CAC) and cardiovascular risk on chest radiographs. MATERIALS AND METHODS: In this retrospective study, 1689 radiographs in patients who underwent cardiac CT and chest radiography within the same year, between 2013 and 2018, were included (mean age, 56 years ± 11 [standard deviation]; 969 radiographs in women). Agatston scores were used as ground truth labels for DCNN training on radiographs. DCNNs were trained for binary classification of (a) nonzero or zero total calcium scores, (b) presence or absence of calcium in each coronary artery, and (c) total calcium scores above or below varying thresholds. Results from classification of test images were compared with established 10-year atherosclerotic cardiovascular disease (ASCVD) risk scores in each cohort. Classifier performance was measured using area under the receiver operating characteristic curve (AUC) with attention maps to highlight areas of decision-making. RESULTS: Binary classification between zero and nonzero total calcium scores reached an AUC of 0.73 on frontal radiographs, with similar performance on laterals (AUC, 0.70; P = .56). Performance was similar for binary classification of absolute total calcium score above or below 100 (AUC, 0.74). Frontal radiographs that tested positive for a predicted nonzero CAC score correlated with a higher 10-year ASCVD risk of 17.2% ± 10.9 compared with 11.9% ± 10.2 for a negative test, indicating predicted CAC score of zero (P < .001). Multivariate logistic regression demonstrated the algorithm could predict a nonzero calcium score independent of traditional cardiovascular risk factors. Performance was reduced for individual coronary arteries. Heat maps primarily localized to the cardiac silhouette and occasionally other cardiovascular findings. CONCLUSION: DCNNs trained on chest radiographs had modest accuracy for predicting the presence of CAC correlating with cardiovascular risk.Keywords: Coronary Arteries, Cardiac, Calcifications/Calculi, Neural NetworksSee also the commentary by Gupta and Blankstein in this issue.©RSNA, 2021.

Preparing junior radiology residents for overnight call via peer-led, hands-on simulation.

PURPOSE: The aim of this study was to assess the perceived value and impact of a hands-on mock call simulation program on resident confidence with interpretation of emergency department overnight call cases. METHODS: A five-session course was implemented in June of 2018 for rising PGY-3/R2 residents to mimic the experience of overnight call with indirect supervision. Sessions were led by senior residents in the program and consisted of timed, independent interpretation of 15-20 high-yield cases per day which highlighted "do-not miss" critical findings and simulated workflow interruptions including phone calls, consultations, and questions from technologists. IRB-approved, and anonymous pre- and post-course surveys were administered to participants which assessed residents' degree of confidence in interpretation of on-call cases and comparison of the mock call experience with existing preparatory strategies. Survey responses were analyzed using McNemar's test and Mann-Whitney U test. RESULTS: Our survey response rate was 91% (29/32). After completing the mock call simulation, there was a significant increase in the mean Likert score of resident confidence levels and feelings of preparedness from 4.59 to 7.38 (p < 0.01). The majority of respondents (72.4% [21/29]) felt that the mock call simulation was "extremely useful." One hundred percent of respondents indicated that the mock call simulation should be implemented for the following year. CONCLUSION: Implementation of a hands-on mock call simulation significantly improves the confidence levels of radiology residents before assuming on-call responsibilities and may serve as an adjunct to existing preparatory strategies.

Patient-Centered Radiology with FHIR: an Introduction to the Use of FHIR to Offer Radiology a Clinically Integrated Platform.

Fast Healthcare Interoperability Resources (FHIR) is an open interoperability standard that allows external software to quickly search for and access clinical information from the electronic medical record (EMR) in a method that is developer-friendly, using current internet technology standards. In this article, we highlight the new FHIR standard and illustrate how FHIR can be used to offer the field of radiology a more clinically integrated and patient-centered system, opening the EMR to external radiology software in ways unfeasible with traditional standards. We explain how to construct FHIR queries relevant to medical imaging using the Society for Imaging Informatics in Medicine (SIIM) Hackathon application programming interface (API), provide sample queries for use, and suggest solutions to offer a patient-centered, rather than an image-centered, workflow that remains clinically relevant.

Metabolic regulation of collagen gel contraction by porcine aortic valvular interstitial cells.

Despite a high incidence of calcific aortic valve disease in metabolic syndrome, there is little information about the fundamental metabolism of heart valves. Cell metabolism is a first responder to chemical and mechanical stimuli, but it is unknown how such signals employed in valve tissue engineering impact valvular interstitial cell (VIC) biology and valvular disease pathogenesis. In this study porcine aortic VICs were seeded into three-dimensional collagen gels and analysed for gel contraction, lactate production and glucose consumption in response to manipulation of metabolic substrates, including glucose, galactose, pyruvate and glutamine. Cell viability was also assessed in two-dimensional culture. We found that gel contraction was sensitive to metabolic manipulation, particularly in nutrient-depleted medium. Contraction was optimal at an intermediate glucose concentration (2 g l(-1)) with less contraction with excess (4.5 g l(-1)) or reduced glucose (1 g l(-1)). Substitution with galactose delayed contraction and decreased lactate production. In low sugar concentrations, pyruvate depletion reduced contraction. Glutamine depletion reduced cell metabolism and viability. Our results suggest that nutrient depletion and manipulation of metabolic substrates impacts the viability, metabolism and contractile behaviour of VICs. Particularly, hyperglycaemic conditions can reduce VIC interaction with and remodelling of the extracellular matrix. These results begin to link VIC metabolism and macroscopic behaviour such as cell-matrix interaction.