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

2021 ◽  
Vol 2021 ◽  
pp. 1-12
Author(s):  
Changhua Liu ◽  
Tao Qin ◽  
Liangjin Liu

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.

2020 ◽  
Author(s):  
Zhengqi Zhu ◽  
Ru Zhang ◽  
Kaixuan Ren ◽  
Ruochen Cong ◽  
Xiangyang Zhu ◽  
...  

Abstract Background: Intravenous thrombolysis (IVT) is a rapid and effective treatment in the early stage of ischemic stroke patients and the purpose of this work is to explore the significance of Hounsfield unit(HU) value in Alberta Stroke Program Early CT Score (ASPECTS) for predicting the clinical prognosis of stroke patients with middle cerebral artery occlusion (MCAO) treated by IVT. Methods: The 84 stroke patients with MCAO treated by IVT were divided into good prognosis group (48 cases) and poor prognosis group (36 cases). HU ratio and HU difference calculated from non-contrast computed tomography (NCCT) between groups were analyzed. Results: The HU ratio of good prognosis group was higher than that in poor prognosis group and the HU difference of good prognosis group was lower than that in poor prognosis group (P<0.05). The HU ratio was negatively correlated with the infarct volume, and the HU difference was positively correlated with the infarct volume (P<0.05). HU difference was an independent risk factor for prognosis of patients with MCAO treated by IVT. The area under the curve (AUC) of HU ratio and HU difference for prognosis was 0.743 and 0.833 respectively (P<0.05). Conclusion: The HU value changes are related to the clinical prognosis of stroke patients with MCAO treated by IVT, HU value may be a prognostic indicator for stroke patients with MCAO treated by IVT.


Stroke ◽  
2014 ◽  
Vol 45 (suppl_1) ◽  
Author(s):  
Ana I. Calleja ◽  
Elisa Cortijo ◽  
Pablo García-Bermejo ◽  
Mario Martinez-Gáldamez ◽  
Santiago Pérez-Fernández ◽  
...  

Background: Selection of best responders to endovascular reperfusion could be aided by predicting the expected duration of tissue-at-risk viability, which may be dependant on collateral circulation status. We aimed to identify the best predictor of good collateral circulation among perfusion computed tomography (PCT) parameters in middle cerebral artery (MCA) ischemic stroke, and to analyze how early MCA response to intravenous thrombolysis and PCT-derived markers of good collaterals interact to determine stroke outcome. Methods: We prospectively studied acute MCA ischemic stroke patients treated with intravenous thrombolysis who underwent PCT prior to treatment showing a target-mismatch profile. Collateral status was assessed using a PCT source-images based score. PCT maps were quantitatively analyzed. Cerebral blood volume (CBV), cerebral blood flow (CBF) and Tmax were calculated within the hypoperfused volume and in the equivalent region of unaffected hemisphere. Occluded MCAs were monitored by transcranial Duplex to assess early recanalization. Main outcome variables were final infarct volume and modified Rankin score at day 90. Results: One hundred MCA ischemic stroke patients imaged by PCT received intravenous thrombolysis, and 68 met all inclusion criteria. A relative CBV > 0.93 emerged as the most robust predictor of good collaterals [OR 12.6 (95% CI 2.9-55.9), p=0.001]. Early MCA recanalization was strongly associated with better long-term outcome and lower infarct volume in patients with rCBV <0.93, but not in patients with high rCBV. None of the patients with rCBV < 0.93 achieved good outcome in absence of early recanalization. Conclusion: rCBV was the strongest marker of collateral status and may help predict the duration of tissue-at-risk viability in hyperacute MCA ischemic stroke


2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Zhengqi Zhu ◽  
Ru Zhang ◽  
Kaixuan Ren ◽  
Ruochen Cong ◽  
Xiangyang Zhu ◽  
...  

Abstract Background Intravenous thrombolysis (IVT) is a rapid and effective treatment in the early stage of ischemic stroke patients and the purpose of this work is to explore the significance of Hounsfield unit (HU) value in Alberta Stroke Program Early CT Score (ASPECTS) for predicting the clinical prognosis of stroke patients with middle cerebral artery occlusion (MCAO) treated by IVT. Methods The 84 stroke patients with MCAO treated by IVT were divided into good prognosis group (48 cases) and poor prognosis group (36 cases). HU ratio and HU difference calculated from non-contrast computed tomography between groups were analyzed. Results The HU ratio of good prognosis group was higher than that in poor prognosis group and the HU difference of good prognosis group was lower than that in poor prognosis group (P < 0.05). The HU ratio and ASPECTS were negatively correlated with the infarct volume, and the HU difference was positively correlated with the infarct volume (P < 0.05). HU difference was an independent risk factor for prognosis of patients with MCAO treated by IVT. The area under the receiver operating characteristic curve of HU ratio and HU difference for prognosis was 0.743 and 0.833 respectively. Conclusion The HU value changes are related to the clinical prognosis of stroke patients with MCAO treated by IVT, HU value may be a prognostic indicator for stroke patients with MCAO treated by IVT.


Stroke ◽  
2021 ◽  
Vol 52 (Suppl_1) ◽  
Author(s):  
Shadi Yaghi ◽  
Eva Mistry ◽  
Adam H De Havenon ◽  
Christopher Leon Guerrero ◽  
Amre Nouh ◽  
...  

Background and Purpose: Multiple studies have established that intravenous thrombolysis with alteplase improves outcome after acute ischemic stroke. However, assessment of thrombolysis’ efficacy in stroke patients with atrial fibrillation (AF) has yielded mixed results. We sought to determine the association of alteplase with mortality, hemorrhagic transformation (HT), infarct volume, and mortality in patients with AF and acute ischemic stroke. Methods: We retrospectively analyzed consecutive acute ischemic stroke patients with AF included in the Initiation of Anticoagulation after Cardioembolic stroke (IAC) study, which pooled data from 8 comprehensive stroke centers in the United States. 1889 (90.6%) had available 90-day follow up data and were included. For our primary analysis we used a cohort of 1367/1889 (72.4%) patients who did not undergo mechanical thrombectomy (MT). Secondary analyses were repeated in the patients that underwent MT (n=522). Binary logistic regression was used to determine whether alteplase use was independently associated with risk of HT, final infarct volume, and 90-day mortality, respectively, adjusting for potential confounders. Results: In our primary analyses we found that alteplase use was independently associated with an increased risk for HT (adjusted OR 2.14, 95% CI 1.49 - 3.07, p <0.001) but overall reduced risk of 90-day mortality (adjusted OR 0.58, 95% CI 0.39 - 0.87, p = 0.009). Among patients undergoing MT, alteplase use was associated with a trend towards a reduction in 90-day mortality (adjusted OR 0.68 95% CI 0.45 - 1.04, p = 0.077). In the subgroup of patients prescribed DOAC treatment (n = 327; 24 received alteplase), alteplase treatment was associated with a trend towards smaller infarct size (< 10 mL), (adjusted OR 0.40, 95% CI 0.15 - 1.12, p = 0.082) without a significant difference in the odds of 90-day mortality (adjusted OR 0.51, 95% CI 0.12 - 2.13, p = 0.357) or hemorrhagic transformation (adjusted OR 0.27, 95% CI 0.03 - 2.07, p = 0.206). Conclusion: Thrombolysis with intravenous alteplase was associated with reduced 90-day mortality in AF patients with acute ischemic stroke not undergoing MT. Further study is required to assess the safety and efficacy of alteplase in AF patients undergoing MT and those on DOACs.


Stroke ◽  
2020 ◽  
Vol 51 (Suppl_1) ◽  
Author(s):  
David Baker ◽  
Dinesh Jillella ◽  
Takashi Shimoyama ◽  
Ken Uchino

Introduction: In patients with large vessel occlusion presenting with acute ischemic stroke, cerebral perfusion is a major determinant of stroke severity. However, limited data exists to guide hemodynamic management of these patients early after presentation. In this study, we aim to evaluate the effect of blood pressure reductions during the hyper-acute period on infarct size. Methods: From a clinical stroke registry at a single comprehensive stroke center, we reviewed patients with middle cerebral artery (M1) or internal carotid artery occlusion who underwent hyperacute magnetic resonance imaging (MRI) for endovascular treatment decision in 2018. Infarct volume was determined by area of reduced apparent diffusion coefficient using RAPID software. Collateral circulation was scored based on baseline CT angiogram (good collaterals constituted >50% filling, poor collaterals ≤50% filling). Average mean arterial pressure (MAP) readings from the first hour of presentation were compared to average MAP readings from the hour prior to magnetic resonance imaging. For the purposes of our study, a drop of > 20% in the average MAP was regarded as a significant decrease. We hypothesized that both significant drop in MAP and the presence of good collateral circulation were independent predictors of infarct volume expressed as a logarithmic value in multivariable regression model. Results: Of the 35 patients (mean age 67, mean NIHSS 16) meeting inclusion criteria, 11% of patients experienced an early significant drop in MAP prior to time of MRI. Among patients with a significant drop in MAP, the average decrease was 35 mm Hg ±3.3 among those with significant drop from a baseline mean MAP of 125 mm Hg. In the multivariable analysis adjusting for collateral status, a significant drop in average MAP was independently associated with an increase in infarct volume (β = -0.727, p=0.0306). Collateral status also independently predicted infarct size (β=0.775, p=0.0007). Conclusion: Among ischemic stroke patients with large vessel occlusion, a >20% drop in MAP during the hyper-acute period is associated with larger infarct volumes. Further studies are needed to optimize early blood pressure management in these patients.


Stroke ◽  
2017 ◽  
Vol 48 (suppl_1) ◽  
Author(s):  
Marta Rubiera ◽  
Alvaro Garcia-Tornell ◽  
Sandra Boned ◽  
Nicolas Romero ◽  
Pilar Coscojuela ◽  
...  

Good collateral circulation (CC) is a strong outcome predictor in acute stroke patients. CT angiography (CTA) is wide-world available but does not provide accurate information about parenchymal status. CT perfusion (CTP) is frequently used to determine ischemic core and tissue at risk. Our aim was to identify an easy and quick method to evaluate CC status by CTP. Methods: Consecutive ischemic stroke patients <8h from symptoms onset evaluated for reperfusion therapies were studied. Non-contrast CT, CTP and multiphase CTA were performed. Patients with confirmed M1-MCA or TICA occlusion on CTA were included. CC evaluation was determined by multiphase CTA (mCTA) according to the Calgary CC Scale and classified as poor (grades 0-2) or good (grades 3-5). In CTP maps, one single ipsi- and contralateral regions of interest (ROI) were defined in the MCA cortical territory (M4, M5, M6). We studied the association of absolute and relative to contralateral ROI-CTP values with CC degree determined by mCTA. Results: 33 patients were included, median NIHSS 17.5 (2-22). Twenty-five patients (75.8%) presented a M1 and 8 (24.2%) a TICA occlusion. On mCTA, 27 (81.8%) patients presented with a favourable CC status and 6 (18.2%) with poor CC. Mean ROI values in the ischemic MCA territory were: CBV 3.5±1.5 ml/100mg, CBF 46.9±29.3 ml/100mg/min, MTT 8.1±3.1 s, Tmax 23.2±4.4 s. In the contralateral non-ischemic MCA, the mean ROI values were: CBV 3.48±1.4, CBF 66.5±32.7, MTT 5.6±2.3, Tmax 20.4±4.8. Absolute and relative CBV-ROI data (relCBV= ischemic CBV value / contralateral CBV value) were the only values significantly associated with CC status on mCTA (good CC mean CBV: 3.8 ml/100g VS poor CC mean CBV: 1.9, p=0.006; good CC mean relCBV 1.1 vs poor CC mean relCBV 0.6, p=0.019). A ROC curve defined 2.5 ml/100mg as the better cut-off point of ROI-CBV that identified patients with good CC status (sensitivity 96%, specificity 84%, VPP 0.96, VPN 0.83). Patients with a ROI-CBV >2.5 presented lower median NIHSS after 24 hours (4 vs 18, p= 0.012) and smaller mean infarct volume on control CT (27.9 vs 88.3, p=0.021). Conclusion: A single cortical ROI-CBV allows an easy and quick accurate evaluation of collateral circulation in CTP. ROI-CBV>2.5 ml/100mg is related to good clinical and radiological outcomes.


2019 ◽  
Author(s):  
Yake Zheng ◽  
Peng Zhao ◽  
yajun lian ◽  
Lihao Li ◽  
Yuan Chen ◽  
...  

Abstract Background We set out to investigate the characteristics and factors related to non-inflammation cerebrospinal fluid (CSF) and normal brain magnetic resonance images (MRI) of autoimmune encephalitis (AE) in patients. Methods The distribution and characteristics of brain MRI and CSF in 124 patients who were living with anti-NMDAR(71), LGI1(26),CASPR2(4),GABAR(23) encephalitis and who had been admitted between October 2016 and May 2018 were analyzed prospectively. Results 12 of the 124 patients(1%) had a normal MRI and non-inflammation CSF.Ten of them were LGI1(83%),while the remaining 1 patient was NMDAR(8.3%),1 patient was CASPR2(8.3%).The clinical symptoms including epilepsy, psychosis, cognitive disorders, conscious disorders, headache, faciobrachial dystonic seizure (FBDS), speech disorders and hypoventilation. AE with non-inflammation CSF and normal MRI with good clinical prognosis. The median modified Rankin Scale (mRS) was low, and recurrence rate was also low. Conclusion The clinical manifestations of on-inflammation CSF and brain MRI-negative patients with AE are not specific, but suggest a better prognosis and a lower recurrence rates.


2015 ◽  
Vol 40 (3-4) ◽  
pp. 182-190 ◽  
Author(s):  
Harri Rusanen ◽  
Jukka T. Saarinen ◽  
Niko Sillanpää

Background: We studied the impact of collateral circulation on CT perfusion (CTP) parametric maps and the amount of salvaged brain tissue, the imaging and clinical outcome at 24 h and at 3 months in a retrospective acute (<3 h) stroke cohort (105 patients) with anterior circulation thrombus treated with intravenous thrombolysis. Methods: Baseline clinical and imaging information were collected and groups with different collateral scores (CS) were compared. Binary logistic regression analyses using good CS (CS ≥2) as the dependent variable were calculated. Results: CTP Alberta Stroke Program Early CT Score (ASPECTS) was successfully assessed in 58 cases. Thirty patients displayed good CS. Poor CS were associated with more severe strokes according to National Institutes of Health Stroke Scale (NIHSS) at arrival (15 vs. 7, p = 0.005) and at 24 h (10 vs. 3, p = 0.003) after intravenous thrombolysis. Good CS were associated with a longer mean onset-to-treatment time (141 vs. 121 min, p = 0.009) and time to CTP (102 vs. 87 min, p = 0.047), better cerebral blood volume (CBV) ASPECTS (9 vs. 6, p < 0.001), better mean transit time (MTT) ASPECTS (6 vs. 3, p < 0.001), better noncontrast CT (NCCT) ASPECTS (10 vs. 8, p < 0.001) at arrival and with favorable clinical outcome at 3 months (modified Rankin Scale ≤2, p = 0.002). The fraction of penumbra that was salvageable at arrival and salvaged at 24 h was higher with better CS (p < 0.001 and p = 0.035, respectively). In multivariate analysis, time from the onset of symptoms to imaging (p = 0.037, OR 1.04 per minute, 95% CI 1.00-1.08) and CBV ASPECTS (p = 0.001, OR 2.11 per ASPECTS point, 95% CI 1.33-3.34) predicted good CS. In similar multivariable models, MTT ASPECTS (p = 0.04, OR 1.46 per ASPECTS point, 95% CI 1.02-2.10) and NCCT ASPECTS predicted good CS (p = 0.003, OR 4.38 per CT ASPECTS point, 95% CI 1.66-11.55) along with longer time from the onset of symptoms to imaging (p = 0.045, OR 1.03 per minute, 95% CI 1.00-1.06 and p = 0.02, OR 1.05 per minute, 95% CI 1.00-1.09, respectively). CBV ASPECTS had a larger area under the receiver operating characteristic curve for good CS (0.837) than NCCT ASPECTS (0.802) or MTT ASPECTS (0.752) at arrival. Conclusions: Favorable CBV ASPECTS, NCCT ASPECTS and MTT ASPECTS are associated with good CS along with more salvageable tissue and longer time from the onset of symptoms to imaging in ischemic stroke patients treated with intravenous thrombolysis.


2014 ◽  
Vol 57 (2) ◽  
pp. 171-178 ◽  
Author(s):  
Roland Faigle ◽  
Amy W. Wozniak ◽  
Elisabeth B. Marsh ◽  
Rafael H. Llinas ◽  
Victor C. Urrutia

2015 ◽  
Vol 40 (5-6) ◽  
pp. 279-285 ◽  
Author(s):  
Dezhi Liu ◽  
Fabien Scalzo ◽  
Sidney Starkman ◽  
Neal M. Rao ◽  
Jason D. Hinman ◽  
...  

Background: Lesion patterns may predict prognosis after acute ischemic stroke within the middle cerebral artery (MCA) territory; yet it remains unclear whether such imaging prognostic factors are related to patient outcome after intravenous thrombolysis. Aims: The aim of this study is to investigate the clinical outcome after intravenous thrombolysis in acute MCA ischemic strokes with respect to diffusion-weighted imaging (DWI) lesion patterns. Methods: Consecutive acute ischemic stroke cases of the MCA territory treated over a 7-year period were retrospectively analyzed. All acute MCA stroke patients underwent a MRI scan before intravenous thrombolytic therapy was included. DWI lesions were divided into 6 patterns (territorial, other cortical, small superficial, internal border zone, small deep, and other deep infarcts). Lesion volumes were measured by dedicated imaging processing software. Favorable outcome was defined as modified Rankin scale (mRS) of 0-2 at 90 days. Results: Among the 172 patients included in our study, 75 (43.6%) were observed to have territorial infarct patterns or other deep infarct patterns. These patients also had higher baseline NIHSS score (p < 0.001), a higher proportion of large cerebral artery occlusions (p < 0.001) and larger infarct volume (p < 0.001). Favorable outcome (mRS 0-2) was achieved in 89 patients (51.7%). After multivariable analysis, groups with specific lesion patterns, including territorial infarct and other deep infarct pattern, were independently associated with favorable outcome (OR 0.40; 95% CI 0.16-0.99; p = 0.047). Conclusions: Specific lesion patterns predict differential outcome after intravenous thrombolysis therapy in acute MCA stroke patients.


Sign in / Sign up

Export Citation Format

Share Document