Search

Background: Accurate, rapid quantification of ventricular scar using cardiac magnetic resonance imaging (CMR) carries importance in arrhythmia management and patient prognosis. Artificial intelligence (AI) has been applied to other radiological challenges with success., Objective: We aimed to assess AI methodologies used for left ventricular scar identification in CMR, imaging sequences used for training, and its diagnostic evaluation., Methods: Following PRISMA recommendations, a systematic search of PubMed, Embase, Web of Science, CINAHL, OpenDissertations, arXiv, and IEEE Xplore was undertaken to June 2021 for full-text publications assessing left ventricular scar identification algorithms. No pre-registration was undertaken. Random-effect meta-analysis was performed to assess Dice Coefficient (DSC) overlap of learning vs predefined thresholding methods., Results: Thirty-five articles were included for final review. Supervised and unsupervised learning models had similar DSC compared to predefined threshold models (0.616 vs 0.633, P = .14) but had higher sensitivity, specificity, and accuracy. Meta-analysis of 4 studies revealed standardized mean difference of 1.11; 95% confidence interval -0.16 to 2.38, P = .09, I 2 = 98% favoring learning methods., Conclusion: Feasibility of applying AI to the task of scar detection in CMR has been demonstrated, but model evaluation remains heterogenous. Progression toward clinical application requires detailed, transparent, standardized model comparison and increased model generalizability., (© 2021 Published by Elsevier Inc. on behalf of Heart Rhythm Society.)

Background and Purpose- Little is known about the real-life factors that clinicians use in selection of patients that would receive endovascular treatment (EVT) in the real world. We sought to determine patient, practitioner, and health system factors associated with therapeutic decisions around endovascular treatment. Methods- We conducted a multinational cross-sectional web-based study comprising of 607 clinicians and interventionalists from 38 countries who are directly involved in acute stroke care. Participants were randomly allocated to 10 from a pool of 22 acute stroke case scenarios. Each case was classified as either Class I, Class II, or unknown evidence according to the current guidelines. We used logistic regression analysis applying weight of evidence approach. Main outcome measures were multilevel factors associated with EVT, adherence to current EVT guidelines, and practice gaps between current and ideal practice settings. Results- Of the 1330 invited participants, 607 (45.6%) participants completed the study (53.7% neurologists, 28.5% neurointerventional radiologists, 17.8% other clinicians). The weighed evidence approach revealed that National Institutes of Health Stroke Scale (34.9%), level of evidence (30.2%), ASPECTS (Alberta Stroke Program Early CT Score) or ischemic core volume (22.4%), patient's age (21.6%), and clinicians' experience in EVT use (19.3%) are the most important factors for EVT decision. Of 2208 responses that met Class I evidence for EVT, 1917 (86.8%) were in favor of EVT. In case scenarios with no available guidelines, 1070 of 1380 (77.5%) responses favored EVT. Comparison between current and ideal practice settings revealed a small practice gap (941 of 6070 responses, 15.5%). Conclusions- In this large multinational survey, stroke severity, guideline-based level of evidence, baseline brain imaging, patients' age and physicians' experience were the most relevant factors for EVT decision-making. The high agreement between responses and Class I guideline recommendations and high EVT use even when guidelines were not available reflect the real-world acceptance of EVT as standard of care in patients with disabling acute ischemic stroke.