Deep learning-based classification of DSA image sequences of patients with acute ischemic stroke.

PURPOSE: Recently, a large number of patients with acute ischemic stroke benefited from the use of thrombectomy, a minimally invasive intervention technique for mechanically removing thrombi from the cerebrovasculature. During thrombectomy, 2D digital subtraction angiography (DSA) image sequences are acquired simultaneously from the posterior-anterior and the lateral view to control whether thrombus removal was successful, and to possibly detect newly occluded areas caused by thrombus fragments split from the main thrombus. However, such new occlusions, which would be treatable by thrombectomy, may be overlooked during the intervention. To prevent this, we developed a deep learning-based approach to automatic classification of DSA sequences into thrombus-free and non-thrombus-free sequences. METHODS: We performed a retrospective study based on the single-center DSA data of thrombectomy patients. For classifying the DSA sequences, we applied Long Short-Term Memory or Gated Recurrent Unit networks and combined them with different Convolutional Neural Networks used as feature extractor. These network variants were trained on the DSA data by using five-fold cross-validation. The classification performance was determined on a test data set with respect to the Matthews correlation coefficient (MCC) and the area under the curve (AUC). Finally, we evaluated our models on patient cases, in which overlooking thrombi during thrombectomy had happened. RESULTS: Depending on the specific model configuration used, we obtained a performance of up to 0.77[Formula: see text]0.94 for the MCC[Formula: see text]AUC, respectively. Additionally, overlooking thrombi could have been prevented in the reported patient cases, as our models would have classified the corresponding DSA sequences correctly. CONCLUSION: Our deep learning-based approach to thrombus identification in DSA sequences yielded high accuracy on our single-center test data set. External validation is now required to investigate the generalizability of our method. As demonstrated, using this new approach may help reduce the incident risk of overlooking thrombi during thrombectomy in the future.

Acute thromboses and occlusions of dual layer carotid stents in endovascular treatment of tandem occlusions.

PURPOSE: To evaluate the occurrence and risk factors of acute in-stent thrombosis or stent occlusion in patients with tandem occlusions receiving intracranial mechanical thrombectomy and emergent extracranial internal carotid artery stenting with a dual layer carotid stent. METHODS: Multicenter retrospective data collection and analysis of stroke databases of seven comprehensive stroke centers from three European countries. RESULTS: Overall, 160 patients (mean (SD) age 66 (12) years; 104 men (65%); median (IQR) baseline NIHSS 14 (9-18); IV lysis, n=97 (60.6%)) were treated for a cervical carotid artery occlusion or stenosis using a CASPER stent (MicroVention), and received mechanical thrombectomy for an intracranial occlusion between April 2014 and November 2018. During the procedure or within 72 hours, formation of thrombus and complete occlusion of the CASPER stent was observed in 33/160 (20.8%) and in 12/160 patients (7.5%), respectively. In 25/33 (75.8%) and in 9/12 patients (75%), respectively, this occurred during the procedure. No statistically significant difference was observed between patients with and without thrombus formation with regard to pre-existing long term medication with anticoagulants or intraprocedural administration of heparin, acetylsalicylic acid (ASA), or heparin and ASA. Favorable early neurological outcome was similar in patients with (n=15; 45.5%) and without (n=63; 49.6%) thrombus formation at the CASPER stent. CONCLUSION: Acute thrombosis or occlusion of CASPER stents in thrombectomy patients receiving emergent extracranial internal carotid artery stenting for tandem occlusions were observed more often during the procedure than within 72 hours of follow-up, were less frequent then previously reported, and showed no impact on early neurological outcome.

Occlusion of a mural type vein of Galen malformation in a 10-month-old boy with three Woven EndoBridge (WEB 17) and two coils.

We report the case of a 10-month-old boy with an enlarged head circumference and severe motor developmental delay. MRI showed a vein of Galen malformation (VGAM) with a heavily dilated median prosencephalic vein. Digital subtraction angiography confirmed a mural type VGAM with three feeding arteries arising from the posterior cerebral arteries. Due to the short length of the feeding arteries and the high flow, occlusion of the feeding vessels with detachable coils was not possible because of repeated coil dislocation into the dilated vein. Embolization of the three feeding vessels was then performed with a Woven EndoBridge single layer device (WEB SL17). In two arteries complete occlusion was accomplished with the WEB alone and in one artery additional deployment of two coils was necessary. Follow-up imaging at day 1 after treatment as well as 3 and 9 months after embolization showed persistent occlusion.