scholarly journals Stroke Lesion Detection and Analysis in MRI Images Based on Deep Learning

2021 ◽  
Vol 2021 ◽  
pp. 1-9
Author(s):  
Shujun Zhang ◽  
Shuhao Xu ◽  
Liwei Tan ◽  
Hongyan Wang ◽  
Jianli Meng
Keyword(s):  
Ischemic Stroke ◽  
Deep Learning ◽  
Deep Neural Networks ◽  
Brain Mri ◽  
Lesion Detection ◽  
Learning Ability ◽  
Massive Data ◽  
Stroke Patients ◽  
Stroke Lesion ◽  
Assisted Diagnosis

Stroke is a kind of cerebrovascular disease that heavily damages people’s life and health. The quantitative analysis of brain MRI images plays an important role in the diagnosis and treatment of stroke. Deep neural networks with massive data learning ability supply a powerful tool for lesion detection. In order to study the property of the stroke lesions and complete intelligent automatic detection, we collaborated with two authoritative hospitals and collected 5,668 brain MRI images of 300 ischemic stroke patients. All the lesion regions in the images were accurately labeled by professional doctors to ensure the authority and effectiveness of the data. Three categories of deep learning object detection networks including Faster R-CNN, YOLOV3, and SSD are applied to implement automatic lesion detection with the best precision of 89.77%. Meanwhile, statistical analysis of the locations, shapes of the lesions, and possible related diseases is conducted with valid conclusions. The research contributes to the intelligent assisted diagnosis and prevention and treatment of ischemic stroke.

scholarly journals Deep Learning Framework to Detect Ischemic Stroke Lesion in Brain MRI Slices of Flair/DW/T1 Modalities

Symmetry ◽  
2021 ◽  
Vol 13 (11) ◽  
pp. 2080
Author(s):  
Venkatesan Rajinikanth ◽  
Shabnam Mohamed Aslam ◽  
Seifedine Kadry
Keyword(s):  
Feature Extraction ◽  
Ischemic Stroke ◽  
Deep Learning ◽  
Brain Mri ◽  
Brain Abnormality ◽  
Deep Feature ◽  
Feature Based ◽  
Stroke Lesion ◽  
Binary Classifiers ◽  
Feature Based Classification

Ischemic stroke lesion (ISL) is a brain abnormality. Studies proved that early detection and treatment could reduce the disease impact. This research aimed to develop a deep learning (DL) framework to detect the ISL in multi-modality magnetic resonance image (MRI) slices. It proposed a convolutional neural network (CNN)-supported segmentation and classification to execute a consistent disease detection framework. The developed framework consisted of the following phases; (i) visual geometry group (VGG) developed VGG16 scheme supported SegNet (VGG-SegNet)-based ISL mining, (ii) handcrafted feature extraction, (iii) deep feature extraction using the chosen DL scheme, (iv) feature ranking and serial feature concatenation, and (v) classification using binary classifiers. Fivefold cross-validation was employed in this work, and the best feature was selected as the final result. The attained results were separately examined for (i) segmentation; (ii) deep-feature-based classification, and (iii) concatenated feature-based classification. The experimental investigation is presented using the Ischemic Stroke Lesion Segmentation (ISLES2015) database. The attained result confirms that the proposed ISL detection framework gives better segmentation and classification results. The VGG16 scheme helped to obtain a better result with deep features (accuracy > 97%) and concatenated features (accuracy > 98%).

U-Net Supported Segmentation of Ischemic-Stroke-Lesion from Brain MRI Slices

2021 ◽  
Author(s):  
Seifedine Kadry ◽  
Robertas Damasevicius ◽  
David Taniar ◽  
Venkatesan Rajinikanth ◽  
Isah A. Lawal
Keyword(s):  
Ischemic Stroke ◽  
Brain Mri ◽  
Stroke Lesion

Abstract WP190: Clinical Features of Sarcoid Patients With Ischemic Stroke

Stroke ◽  
2016 ◽  
Vol 47 (suppl_1) ◽  
Author(s):  
Ahmed Z Obeidat ◽  
Heidi Sucharew ◽  
Charles J Moomaw ◽  
Dawn O Kleindorfer ◽  
Brett M Kissela ◽  
...  
Keyword(s):  
Ischemic Stroke ◽  
Current Knowledge ◽  
Case Reports ◽  
Brain Mri ◽  
Sporadic Case ◽  
Prior History ◽  
Stroke Patients ◽  
Stroke Subtype ◽  
Stroke Event ◽  
History Of

Background: Current knowledge on ischemic stroke in sarcoid patients stems from sporadic case reports. The mechanism is thought to be related to granulomatous involvement of brain vasculature. However, clinical, demographic, and radiographic features of sarcoid patients with ischemic stroke are lacking. If sarcoid patients are at higher risk for ischemic stroke event, we hypothesized that the risk factors for ischemic stroke and stroke subtype distribution would differ between sarcoid and non-sarcoid ischemic stroke patients. Methods: Cases of ischemic stroke were identified for the years 2005 and 2010 from the population-based Greater Cincinnati/Northern Kentucky Stroke Study (population 1.3 million). Ischemic stroke cases were physician study confirmed and patients with a history of sarcoid were identified through medical chart review. Clinical variables were compared between stroke patients with history of sarcoid and those with no prior sarcoid history. Results: A total of 4258 cases of ischemic stroke were identified; of them, only 18 had prior diagnosis of sarcoid (0.04%). Brain MRI showed diffusion restriction in 14 out of 15 (93%) MRIs performed in sarcoid patients. The table presents risk factor and subtype data on sarcoid patients compared with non-sarcoid patients. Conclusions: We identified only a few cases of prior sarcoid history in our two-year ascertainment of ischemic stroke patients in our population. In comparison with stroke patients with no prior history of sarcoid, the sarcoid patients tended to be of younger age at presentation, female, have a history of diabetes and hyperlipidemia, and more likely of African descent, perhaps related to the diagnosis of sarcoid itself. We were unable to detect differences in stroke subtype distributions between sarcoid and non-sarcoid ischemic stroke patients.

Abstract T P188: High Prevalence Of Igm Anti-phosphatidylserine-prothrombin Complex Antibodies And Lupus Anticoagulant In Patients With Non-lacunar Ischemic Stroke Without A Determined Cause

Stroke ◽  
2015 ◽  
Vol 46 (suppl_1) ◽  
Author(s):  
Eisuke Furui ◽  
Ryo Itabashi ◽  
Yukako Yazawa ◽  
Masahide Yamazaki ◽  
Satoru Fujiwara
Keyword(s):  
Ischemic Stroke ◽  
Lupus Anticoagulant ◽  
Brain Mri ◽  
Arterial Stenosis ◽  
Enzyme Linked Immunosorbent Assay ◽  
Lacunar Stroke ◽  
Stroke Patients ◽  
Carotid Ultrasonography ◽  
Glycoprotein I ◽  
High Prevalence

Background and Purpose: The cause of ischemic stroke remains undetermined in about 25% of patients even after appropriate studies. Antiphospholipid syndrome (APS) is an autoimmune disease in which vascular thrombosis occurs in patients with antiphospholipid (aPL) antibodies. Anti-phosphatidylserine-prothrombin complex (aPS/PT) antibodies are reported to be associated with symptoms of APS and lupus anticoagulant (LAC) activity. Correlations between ischemic stroke and IgM aPS/PT antibodies have not been well investigated. We tried to explore possible correlations between non-lacunar ischemic stroke without a determined cause and aPL antibodies. Methods: Diagnostic studies consisted of brain MRI/CT, extracranial and intracranial MRA, carotid ultrasonography, 12-lead ECG, 24-hour ECG, transthoracic echo cardiography, transesophageal echo cardiography and venous ultrasonography. After excluding lacunar stroke, proximal arterial stenosis and caridoembolic sources with all these diagnositc studies, we selected 22 non-lacunalr ischemic stroke patients without a determined cause (age 48.5 +/- 14.4 years; 9 woman). We preliminarily measured IgG and IgM anticardiolipin (aCL) antibodies, IgG and IgM aPS/PT antibodies, anti-beta2-glycoprotein I-dependent cardiolipin (anti-beta2 GPI/CL) antibodies and LAC. aCL, aPS/PT, and anti-beta2 GPI/CL antibodies were measured by enzyme-linked immunosorbent assay. LAC was determined according to the Subcommittee on Lupus Anticoagulant/Phospholipid Dependent Antibody guidelines. Results: IgM aPS/PT antibodies, LAC, and both of them were detected in 10, 7, and 3 patients, respectively. Neither aCL nor anti-beta2 GPI/CL antibodies were identified in all patients. Conclusion: Our data demonstrate a high prevalence of IgM aPS/PT antibodies and LAC in patients with non-lacunar ischemic stroke without a determined cause. IgM aPS/PT antibodies and LAC might play some role in the development of ischemic stroke of these patients.

Brain microbleeds, anticoagulation, and hemorrhage risk

Neurology ◽  
2017 ◽  
Vol 89 (23) ◽  
pp. 2317-2326 ◽  
Author(s):  
Andreas Charidimou ◽  
Christopher Karayiannis ◽  
Tae-Jin Song ◽  
Dilek Necioglu Orken ◽  
Vincent Thijs ◽  
...  
Keyword(s):  
Ischemic Stroke ◽  
Meta Analysis ◽  
Oral Anticoagulants ◽  
Brain Mri ◽  
Stroke Patients ◽  
Nonvalvular Atrial Fibrillation ◽  
Random Effects Models ◽  
Level Data ◽  
Hemorrhage Risk

Objectives:To assess the association between cerebral microbleeds (CMBs) and future spontaneous intracerebral hemorrhage (ICH) risk in ischemic stroke patients with nonvalvular atrial fibrillation (AF) taking oral anticoagulants.Methods:This was a meta-analysis of cohort studies with >50 patients with recent ischemic stroke and documented AF, brain MRI at baseline, long-term oral anticoagulation treatment, and ≥6 months of follow-up. Authors provided summary-level data on stroke outcomes stratified by CMB status. We estimated pooled annualized ICH and ischemic stroke rates from Poisson regression. We calculated odds ratios (ORs) of ICH by CMB presence/absence, ≥5 CMBs, and CMB topography (strictly lobar, mixed, and strictly deep) using random-effects models.Results:We established an international collaboration and pooled data from 8 centers including 1,552 patients. The crude CMB prevalence was 30% and 7% for ≥5 CMBs. Baseline CMB presence (vs no CMB) was associated with ICH during follow-up (OR 2.68, 95% confidence interval [CI] 1.19–6.01, p = 0.017). Presence of ≥5 CMB was related to higher future ICH risk (OR 5.50, 95% CI 2.07–14.66, p = 0.001). The pooled annual ICH incidence increased from 0.30% (95% CI 0.04–0.55) among CMB-negative patients to 0.81% (95% CI 0.17–1.45) in CMB-positive patients (p = 0.01) and 2.48% (95% CI 1.2–6.2) in patients with ≥5 CMBs (p = 0.001). There was no association between CMBs and recurrent ischemic stroke.Conclusions:The presence of CMB on MRI and the dichotomized cutoff of ≥5 CMBs might identify subgroups of ischemic stroke patients with AF with high ICH risk and after further validation could help in risk stratification, in anticoagulation decisions, and in guiding randomized trials and ongoing large observational studies.

Cortical Superficial Siderosis and Risk of Bleeding after Thrombolysis for Ischemic Stroke

2015 ◽  
Vol 40 (3-4) ◽  
pp. 191-197 ◽  
Author(s):  
Thomas Gattringer ◽  
Sebastian Eppinger ◽  
Markus Beitzke ◽  
Gerit Wuensch ◽  
Kurt Niederkorn ◽  
...  
Keyword(s):  
Ischemic Stroke ◽  
Clinical Data ◽  
Brain Mri ◽  
Tertiary Care ◽  
Medical Documentation ◽  
Superficial Siderosis ◽  
Amyloid Angiopathy ◽  
Stroke Patients ◽  
Cerebral Mri ◽  
Cortical Superficial Siderosis

Background: Cortical superficial siderosis (CSS) is a neuroimaging marker of cerebral amyloid angiopathy and has been associated with a high risk for early subsequent major intracranial hemorrhage (ICH). Therefore, many experts recommend withholding of antithrombotic medication to patients with CSS. In this study, we sought to investigate the prevalence of CSS and the associated risk of ICH in the setting of intravenous thrombolysis (IVT) for ischemic stroke. Methods: We retrospectively searched the medical documentation system of our primary and tertiary care university clinic for all patients with ischemic stroke that received IVT from 2009 to December 2014. All available imaging data were reviewed in a standardized manner and blinded to any clinical data for the presence of CSS and ICH. CSS was defined as linear signal loss along the cerebral cortex on gradient echo T2*-weighted sequences. A stroke neurologist, who was blinded to the neuroimaging data, extracted the corresponding clinical data including follow-up information. Results: We identified 298 patients that received IVT and had undergone brain MRI (mean age 67.6 ± 12.6 years, 59.4% male). Cerebral MRI was performed in 116 patients (38.9%) before and in 182 patients (61.1%) after IVT (median time from stroke symptom onset to MRI: 1 day; range 0-7 days). Only 3 patients (2 females and 1 male aged 90, 76 and 73 years, respectively) had CSS (1%). All of them had a middle cerebral artery (MCA) stroke with a corresponding vessel occlusion. The 76-year-old female patient had extensive CSS and numerous cerebral microbleeds and received another IVT treatment for recurrent MCA stroke 8 months after the first event. After both IVTs, she had clinically asymptomatic small ICH outside the ischemic infarct and distant from CSS. The 2 other patients had only mild to moderate CSS and did not experience any ICH on postthrombolytic imaging. Conclusions: The prevalence of CSS in a clinical cohort of stroke patients that received IVT was low and thus does not appear to pose a substantial risk for symptomatic ICH although this may occur in individual patients. However, such analysis also needs to be extended to the very old stroke patients in whom IVT is increasingly used.

Sign in / Sign up

Export Citation Format

Share Document