Evaluation of Ischemic Penumbra in Stroke Patients Based on Deep Learning and Multimodal CT

In order to investigate the value of multimodal CT for quantitative assessment of collateral circulation, ischemic semidark zone, core infarct volume in patients with acute ischemic stroke (AIS), and prognosis assessment in intravenous thrombolytic therapy, segmentation model which is based on the self-attention mechanism is prone to generate attention coefficient maps with incorrect regions of interest. Moreover, the stroke lesion is not clearly characterized, and lesion boundary is poorly differentiated from normal brain tissue, thus affecting the segmentation performance. To address this problem, a primary and secondary path attention compensation network structure is proposed, which is based on the improved global attention upsampling U-Net model. The main path network is responsible for performing accurate lesion segmentation and outputting segmentation results. Likewise, the auxiliary path network generates loose auxiliary attention compensation coefficients, which compensate for possible attention coefficient errors in the main path network. Two hybrid loss functions are proposed to realize the respective functions of main and auxiliary path networks. It is experimentally demonstrated that both the improved global attention upsampling U-Net and the proposed primary and secondary path attention compensation networks show significant improvement in segmentation performance. Moreover, patients with good collateral circulation have a small final infarct area volume and a good clinical prognosis after intravenous thrombolysis. Quantitative assessment of collateral circulation and ischemic semidark zone by multimodal CT can better predict the clinical prognosis of intravenous thrombolysis.

Multimodal CT Imaging Characteristics in Predicting Prognosis of Wake-Up Stroke

Background: Multimodal CT imaging can evaluate cerebral hemodynamics and stroke etiology, playing an important role in predicting prognosis. This study aimed to summarize the comprehensive image characteristics of wake-up stroke (WUS), and to explore its value in prognostication.Methods: WUS patients with anterior circulation large vessel occlusion were recruited into this prospective study. According to the 90-day modified Rankin Scale (mRS), all patients were divided into good outcome (mRS 0–2) or bad (mRS 3–6). Baseline clinical information, multimodal CT imaging characteristics including NECT ASPECTS, clot burden score (CBS), collateral score, volume of penumbra and ischemic core on perfusion were compared. Multivariate logistic regression analysis was further used to analyze predictive factors for good prognosis. Area under curve (AUC) was calculated from the receiver operating characteristic (ROC) curve to assess prognostic value.Results: Forty WUS were analyzed in this study, with 20 (50%) achieving good outcome. Upon univariable analysis, the good outcome group demonstrated higher ASPECTS, higher CBS, higher rate of good collateral filling and lower penumbra volume when compared with the poor outcome group. Upon logistic regression analysis, poor outcome significantly correlated with penumbra volume (OR: 1.023, 95% CI = 1.003–1.043) and collateral score (OR: 0.140, 95% CI = 0.030–0.664). AUC was 0.715 for penumbra volume (95% CI, 0.550–0.846) and 0.825 for good collaterals (95% CI, 0.672–0.927) in predicting outcome.Conclusions:Penumbra volume and collateral score are the most relevant baseline imaging characters in predicting outcome of WUS patients. These imaging characteristics might be instructive to treatment selection. As the small sample size of current study, further studies with larger sample size are needed to confirm these observations.

Practicability and Diagnostic Yield of One-Stop Stroke CT with Delayed-Phase Cardiac CT in Detecting Major Cardioembolic Sources of Acute Ischemic Stroke

Purpose Recurrent stroke is considered to increase the incidence of severe disability and death. For correct risk assessment and patient management it is essential to identify the origin of stroke at an early stage. Transthoracic echocardiography (TTE) is the initial standard of care for evaluating patients in whom a cardioembolic source of stroke (CES) is suspected but its diagnostic capability is limited. Transesophageal echocardiography (TEE) is considered as gold standard; however, this approach is time consuming, semi-invasive and not always feasible. We hypothesized that adding a delayed-phase cardiac computed tomography (cCT) to initial multimodal CT might represent a valid alternative to routine clinical echocardiographic work-up. Material and Methods Patients with suspected acute cardioembolic stroke verified by initial multimodal CT and subsequently examined with cCT were included. The cCT was evaluated for presence of major CES and compared to routine clinical echocardiographic work-up. Results In all, 102 patients with suspected acute CES underwent cCT. Among them 60 patients underwent routine work-up with echocardiography (50 TTE and only 10 TEE). By cCT 10/60 (16.7%) major CES were detected but only 4 (6.7%) were identified by echocardiography. All CES observed by echocardiography were also detected by cCT. In 8 of 36 patients in whom echocardiography was not performed cCT also revealed a major CES. Conclusion These preliminary results show the potential diagnostic yield of delayed-phase cCT to detect major CES and therefore could accelerate decision-making to prevent recurrence stroke. To confirm these results larger studies with TEE as the reference standard and also compared to TTE would be necessary.