Machine learning approach for hemorrhagic transformation prediction: Capturing predictors' interaction.

Background and purpose: Patients with ischemic stroke frequently develop hemorrhagic transformation (HT), which could potentially worsen the prognosis. The objectives of the current study were to determine the incidence and predictors of HT, to evaluate predictor interaction, and to identify the optimal predicting models. Methods: A prospective study included 360 patients with ischemic stroke, of whom 354 successfully continued the study. Patients were subjected to thorough general and neurological examination and T2 diffusion-weighted MRI, at admission and 1 week later to determine the incidence of HT. HT predictors were selected by a filter-based minimum redundancy maximum relevance (mRMR) algorithm independent of model performance. Several machine learning algorithms including multivariable logistic regression classifier (LRC), support vector classifier (SVC), random forest classifier (RFC), gradient boosting classifier (GBC), and multilayer perceptron classifier (MLPC) were optimized for HT prediction in a randomly selected half of the sample (training set) and tested in the other half of the sample (testing set). The model predictive performance was evaluated using receiver operator characteristic (ROC) and visualized by observing case distribution relative to the models' predicted three-dimensional (3D) hypothesis spaces within the testing dataset true feature space. The interaction between predictors was investigated using generalized additive modeling (GAM). Results: The incidence of HT in patients with ischemic stroke was 19.8%. Infarction size, cerebral microbleeds (CMB), and the National Institute of Health stroke scale (NIHSS) were identified as the best HT predictors. RFC (AUC: 0.91, 95% CI: 0.85-0.95) and GBC (AUC: 0.91, 95% CI: 0.86-0.95) demonstrated significantly superior performance compared to LRC (AUC: 0.85, 95% CI: 0.79-0.91) and MLPC (AUC: 0.85, 95% CI: 0.78-0.92). SVC (AUC: 0.90, 95% CI: 0.85-0.94) outperformed LRC and MLPC but did not reach statistical significance. LRC and MLPC did not show significant differences. The best models' 3D hypothesis spaces demonstrated non-linear decision boundaries suggesting an interaction between predictor variables. GAM analysis demonstrated a linear and non-linear significant interaction between NIHSS and CMB and between NIHSS and infarction size, respectively. Conclusion: Cerebral microbleeds, NIHSS, and infarction size were identified as HT predictors. The best predicting models were RFC and GBC capable of capturing nonlinear interaction between predictors. Predictor interaction suggests a dynamic, rather than, fixed cutoff risk value for any of these predictors.

Neutrophil-lymphocyte ratio as a marker for disability and activity in multiple sclerosis.

BACKGROUND: There is limited data regarding the association of Neutrophil-lymphocyte ratio (NLR) inflammatory marker with multiple sclerosis (MS) disability and activity. OBJECTIVE: The aim of the present study was to evaluate validity of NLR as an inflammatory marker for MS disability and activity. METHODS: A case-control study including 140 MS patients and 140 age, sex and body mass index matched healthy controls was performed. All participants were subjected to detailed history taking, complete general and neurological examination, laboratory and radiological investigations. Assessment of disease disability was performed using Expanded Disability Status Scale. RESULTS: The NLR levels was significantly higher in MS patients compared to the controls and in patients with relapse compared to remission. Logistic regression analyses showed that NLR was significantly associated with disease disability (odds ratio (OR): 2.568; confidence interval (CI): 1.377 - 4.788; P: 0.003) and activity (OR: 3.603; CI: 2.086 - 6.226; P: 0.02). The cutoff value for the NLR to predict MS disability and activity was 3.12. CONCLUSION: NLR was significantly increased in MS patients compared to controls. The significant association of NLR with MS disability and activity suggest that it could be utilized as a simple, rapid and inexpensive inflammatory marker.

Metabolic syndrome components and disease disability in egyptian multiple sclerosis patients.

BACKGROUND: There is limited and inconsistent data on metabolic syndrome (MetS) in multiple sclerosis (MS) patients. The aim of this study was to estimate the frequency of MetS and its components in MS patients and to evaluate their association with disease disability in Egyptian MS patients. METHODS: A cross-section study was carried out on 60 patients (19 males and 41 females) with relapsing remitting MS. All patients were subjected to full general and neurological examination, laboratory and radiological investigations. Assessment of disease disability was performed using Expanded Disability Status Scale (EDSS) and MetS was diagnosed according to National Cholesterol Education Program Adult Treatment Panel Ш (NCEP-ATP III). RESULTS: The frequency of MetS in MS patients was 36.7%. Our findings show that 53.3% of MS patients had abdominal obesity, 21.7% had hypertension, 38.3% had diabetes mellitus, 43.3% had elevated triglycerides level and 56.7% had dyslipidemia. Linear regression analysis revealed that body mass index (BMI), dyslipidemia, current medication and disease duration were significantly associated with disease disability. CONCLUSION: High frequency of MetS and its component was observed in MS patients. Disease duration and current medication as well as some MetS component such as BMI, dyslipidemia, were significantly associated with disability in MS patients.

Prevalence and Associated Risk Factors of Cerebral Microbleeds in Egyptian Patients with Acute Ischemic Stroke and Atrial Fibrillation.

BACKGROUND: Few studies addressed the prevalence of cerebral microbleeds (CMB) and associated risk factor profile in Egyptian ischemic cerebral stroke patients with atrial fibrillation (AF). METHODS: The prevalence of CMB was estimated in 150 cases of AF ischemic stroke patients and compared to the prevalence in 150 age- and sex-matched controls of ischemic stroke patients without AF. CMB-associated risk factors were identified by comparing AF ischemic stroke patients with and without CMB. All participants were subjected to complete general, neurological examination, and magnetic resonance imaging. RESULTS: The prevalence of CMBs in ischemic stroke with and without AF was 40.7% and 49.3%, respectively. Age, hypertension, diabetes mellitus, past history of stroke, antiplatelet, anticoagulant, National Institutes of Health Stroke Scale, CHA2DS2VASc, and white matter lesions (WML) were significant risk factors associated with CMB on univariate analysis. On multivariable logistic regression analysis, age (odds ratio [OR] 1.1, confidence interval [CI] 1.02-1.13), hypertension (OR 3.2, CI 1.19-8.81), anticoagulant (OR 3.3, CI 1.17-9.40), and WML (OR 9.6, CI 3.49-26.3) were the only independent risk factors associated with the presence of CMBs. CONCLUSIONS: AF in ischemic stroke patients was not associated with higher prevalence of CMBs. Old age, hypertension, anticoagulant treatment, and WML were the independent risk factors associated with CMB in AF ischemic stroke patients. Our results suggest that elderly hypertensive AF ischemic stroke patients maintained on anticoagulant therapy should be screened for the incidence of CMBs and monitored regularly for the development of intracerebral hemorrhage.

Early post-stroke seizures in a sample of Egyptian patients with first-ever stroke.

BACKGROUND: Cerebrovascular stroke is considered as the most common cause of symptomatic seizures. Post-stroke seizures herald a poor functional outcome, a high mortality rate, and a long in-hospital duration of stay. AIM: To assess the incidence and risk factors responsible for early seizures after the first-ever cerebral stroke, and its relation to the functional outcome and the in-hospital mortality following stroke as well as the duration of stay. PATIENTS AND METHODS: 150 patients with first-ever stroke, who were admitted in an intensive care and stroke unit of the Neurology Department of Zagazig University Hospitals, Egypt, within the first 24 hours of stroke onset were included in the study. Early seizures (within 7 days of stroke onset), stroke severity (National Institutes of Health stroke scale; NIHSS), radiological data, functional outcome (Barthel index), in-hospital mortality and duration of stay were collected and analyzed. RESULTS: The incidence of early seizures associated with first-ever stroke was 9.3%. Risk factors for early seizures were intracerebral hemorrhage (ICH) [odds ratio (OR) = 3.2, P = <0.001), cortical lesions (OR = 2.8, P = <0.05), and a large lesion size (OR = 2, P= <0.05). Patients with early seizures had lower scores on the Barthel index than those without seizures (7.5 ± 4.1 versus 10.5 ± 3.5), had higher discharge National Institutes of Health Stroke scale (NIHSS) scores, had a higher in-hospital mortality rate (35.7% versus 6.6% in patients without seizures) and duration of stay (19.4 ± 6.5 versus 10.7 ± 4.2 days respectively). CONCLUSION: The incidence of early post-stroke seizures was high in our study and was positively associated with a poor functional outcome, a higher in-hospital mortality rate and a longer duration of in-hospital stay. The most important risk factors were ICH, the cortical site of involvement and a large size of the lesion.

Infarction Size, Interleukin-6, and Their Interaction Are Predictors of Short-Term Stroke Outcome in Young Egyptian Adults.

BACKGROUND: A trend of increasing incidence of first-ever cerebral ischemic stroke in young adults has been recently reported. The current study was conducted with the objective of identifying independent predictors of short-term outcome of first-ever cerebral ischemic stroke affecting young Egyptian adults. METHODS: The present hospital-based study included 50 patients, 23 males and 27 females, aged 18-45 years, with first-ever ischemic stroke confirmed by computed tomography (CT) and magnetic resonance imaging. Twenty healthy age- and sex-matched random control subjects were included to set the reference laboratory values. Detailed medical, neurological, and laboratory data were collected. Stroke severity and short-term stroke outcome were assessed using the Canadian Neurological Scale and the National Institutes of Health Stroke Scale (NIHSS), respectively. RESULTS: High prevalence of modifiable risk factors was observed in young Egyptian adults affected with first-ever ischemic cerebral stroke. Although all studied risk factors were significantly correlated with NIHSS score, multiple regression analysis revealed that only infarction size (CT size), interleukin-6 (IL-6), and their synergistic interaction were the most important predictors of NIHSS stroke outcome. CONCLUSIONS: IL-6 and infarction size were independent predictors of short-term stroke outcome in young Egyptian adults. Synergistic interaction of IL-6 with infarction size suggests an investigative value for assessing serum IL-6 level and a therapeutic benefit for its reduction during the course of early ischemic stroke treatment.