Index by author

Convolutional neural networks were trained on combinations of DWI, ADC, and low b-value-weighted images from 116 subjects. The performances of the networks (measured by the Dice score, sensitivity, and precision) were compared with one another and with ensembles of 5 networks. An ensemble of convolutional neural networks trained on DWI, ADC, and low b-value-weighted images produced the most accurate acute infarct segmentation over individual networks. Automated volumes correlated with manually measured volumes for the independent cohort.

The authors performed a retrospective analysis of the prospectively maintained Endovascular Treatment of Ischemic Stroke multicentric registry. Data from consecutive patients who benefited from thrombectomy with a first-line Sofia approach between January 2013 and April 2018 were studied. We excluded other first-line approaches (stent retriever or combined aspiration and stent retriever) and extracranial occlusions. During the study period, 296 patients were treated. Mean age and initial NIHSS score were, respectively, 69.5 years and 16. Successful reperfusion, defined by the modified TICI 2b/3, was obtained in 86.1%. Complete reperfusion (modified TICI 3) was obtained in 41.2%. A first-pass effect was achieved in 24.2%. A rescue stent retriever approach was required in 29.7%. The first-line contact aspiration approach appeared safe and efficient with Sofia catheters. These devices achieved very high reperfusion rates with a low requirement for stent retriever rescue therapy, especially for M1 occlusions.