Aad Van der Lugt (Netherlands)
ErasmusMC RadiologyAuthor Of 2 Presentations
DEEP LEARNING-BASED INTRACRANIAL PERFORATION DETECTION IN DSA IMAGES OBTAINED DURING ENDOVASCULAR THROMBECTOMY
Abstract
Background and Aims
Intracranial vessel perforation is a procedural complication during endovascular thrombectomy (EVT). Its occurrence is strongly associated with unfavorable treatment outcomes. Early identification of perforation would allow therapeutic actions to prevent situation deterioration. Due to its low occurrence and the large heterogeneity in image appearance, perforations may initially be missed by the interventionalists. In this work, we study the feasibility of automated intracranial vessel perforation detection during EVT using deep learning techniques.
Methods
From the MR CLEAN registry, fifty-three patients (149 acquisitions) with vessel perforations were identified and annotated by an experienced neuroradiologist. Another 150 acquisitions (from 150 patients) without perforations were randomly selected as negative samples. The proposed solution builds on top of state-of-the-art object detection algorithms. It incorporates temporal information of DSA with bidirectional convolutional gated recurrent units (Bi-ConvGRU), further followed by a problem-tailored acquisition level optimization to reduce false positives based on temporal consistency.
Results
In ten-fold cross-validation on 203 patients (299 acquisitions, 3607 images), the proposed method achieves an area under the receiver-operating characteristic curve of 90% for acquisition-level classification of acquisitions with a perforation. The sensitivity and specificity were 81% and 86%, respectively.
Conclusions
The proposed deep learning-based algorithm achieves promising performance in vessel perforation detection in DSA for stroke patients, and can potentially be deployed in clinical practice to detect perforation early and allow direct clinical decision making.
ASSOCIATION BETWEEN THROMBUS CHARACTERISTICS, SUCCESSFUL REPERFUSION AND FUNCTIONAL OUTCOME AFTER ENDOVASCULAR TREATMENT IN PATIENTS WITH POSTERIOR CIRCULATION STROKE.
Abstract
Background and Aims
Thrombus perviousness and length are important thrombus imaging characteristics related to outcome in patients with anterior circulation ischemic stroke. We investigated the association of thrombus characteristics with reperfusion and functional outcome in patients with posterior circulation stroke (PCS) who underwent endovascular treatment (EVT).
Methods
Thrombus imaging characteristics (perviousness, density, and length) were measured in patients with PCS registered in the MR CLEAN Registry between 2014-2018. All characteristics were assessed on thin-slice (<2.5mm) non-contrast computed tomography and computed tomography angiography imaging acquired within 30 minutes from each other. We compared thrombus characteristics in patients with or without excellent reperfusion (expanded Treatment In Cerebral Ischemia (eTICI) score ≥2C) and with or without favorable functional outcome (modified Rankin Scale score (mRS) 0-3) using the Mann-Whitney U-test.
Results
Sixty-six patients were analyzed. Median thrombus perviousness was 7 (IQR -1-21) Hounsfield Units (HU). Median density was 54 (IQR 44-66) HU and median length was 24 (IQR 15-42) mm. No significant between-group differences were found in thrombus characteristics (Table 1).
| eTICI<2C (n=32) | eTICI≥2C (n=29) | p value | mRS 0-3 (n=27) | mRS 4-6 (n=37) | p value | |
| Thrombus imaging characteristics | ||||||
| Length (mm), median (IQR) | 24 (14-49) | 21 (15-29) | 0.45 | 21 (15-29) | 25 (15-48) | 0.24 |
| Perviousness (HU), median (IQR) | 11 (1-21) | 7 (-4-18) | 0.42 | 2 (-2-20) | 10 (1-21) | 0.29 |
| Density (HU), median (IQR) | 55 (47-66) | 51 (43-62) | 0.63 | 55 (43-66) | 50 (44-65) | 0.64 |
Conclusions
In this study of patients with PCS treated with EVT, thrombus imaging characteristics were not associated with reperfusion or functional outcome.