Browsing by Author "Vukčević, Vladan (15741934700)"

[No abstract available]

Introduction. Atrial fibrillation (AF) is the major cause of stroke, particularly in older patients over 75 years of age. European Society of Cardiology guidelines recommend chronic anticoagulation therapy in patients with atrial fibrillation if CHA2DS2-VASc score is ≥ 1 [CHA2DS2-VASc score for estimating the risk of stroke in patients with nonrheumatic AF consisting of the first letters of patients condition: C – congestive heart failure; H – hypertension; A2 – age ≥ 75 years; D – diabetes mellitus; S2 – prior stroke, transitory ischaemic attack (TIA) or thrombolism; V – vascular disease; A – age 65–74 years; Sc – sex category]. However, a significant number of patients have a high bleeding risk, or are contraindicated for chronic oral anticoagulation, and present a group of patients in whom alternative treatment options for thromboembolic prevention are required. Transcatheter percutaneous left atrial appendage closure (LAAC) devices have been recommended in patients with contraindications for chronic anticoagulant therapy. Case report. We present our first three patients with nonvalvular AF and contraindications for chronic anticoagulant therapy who were successfully treated with implantation of LAAC Watchman device in Catheterization Laboratory of the Clinic for Cardiology, Clinical Center of Serbia in Belgrade Conclusion. Our initial results with Watchman LAAC device are promising and encouraging, providing real alternative in patients with non-valvular AF and contraindication for chronic anticoagulant therapy and high bleeding risk. © 2017, Institut za Vojnomedicinske Naucne Informacije/Documentaciju. All rights reserved.

Objective: Optimizing timing of defibrillation by evaluating the likelihood of a successful outcome could significantly enhance resuscitation. Previous studies employed conventional machine learning approaches and hand-crafted features to address this issue, but none have achieved superior performance to be widely accepted. This study proposes a novel approach in which predictive features are automatically learned. Methods: A raw 4s VF episode immediately prior to first defibrillation shock was feed to a 3-stage CNN feature extractor. Each stage was composed of 4 components: convolution, rectified linear unit activation, dropout and max-pooling. At the end of feature extractor, the feature map was flattened and connected to a fully connected multi-layer perceptron for classification. For model evaluation, a 10 fold cross-validation was employed. To balance classes, SMOTE oversampling method has been applied to minority class. Results: The obtained results show that the proposed model is highly accurate in predicting defibrillation outcome (Acc = 93.6 %). Since recommendations on classifiers suggest at least 50 % specificity and 95 % sensitivity as safe and useful predictors for defibrillation decision, the reported sensitivity of 98.8 % and specificity of 88.2 %, with the analysis speed of 3 ms/input signal, indicate that the proposed model possesses a good prospective to be implemented in automated external defibrillators. Conclusions: The learned features demonstrate superiority over hand-crafted ones when performed on the same dataset. This approach benefits from being fully automatic by fusing feature extraction, selection and classification into a single learning model. It provides a superior strategy that can be used as a tool to guide treatment of OHCA patients in bringing optimal decision of precedence treatment. Furthermore, for encouraging replicability, the dataset has been made publicly available to the research community. © 2020 Elsevier B.V.

Objective: Optimizing timing of defibrillation by evaluating the likelihood of a successful outcome could significantly enhance resuscitation. Previous studies employed conventional machine learning approaches and hand-crafted features to address this issue, but none have achieved superior performance to be widely accepted. This study proposes a novel approach in which predictive features are automatically learned. Methods: A raw 4s VF episode immediately prior to first defibrillation shock was feed to a 3-stage CNN feature extractor. Each stage was composed of 4 components: convolution, rectified linear unit activation, dropout and max-pooling. At the end of feature extractor, the feature map was flattened and connected to a fully connected multi-layer perceptron for classification. For model evaluation, a 10 fold cross-validation was employed. To balance classes, SMOTE oversampling method has been applied to minority class. Results: The obtained results show that the proposed model is highly accurate in predicting defibrillation outcome (Acc = 93.6 %). Since recommendations on classifiers suggest at least 50 % specificity and 95 % sensitivity as safe and useful predictors for defibrillation decision, the reported sensitivity of 98.8 % and specificity of 88.2 %, with the analysis speed of 3 ms/input signal, indicate that the proposed model possesses a good prospective to be implemented in automated external defibrillators. Conclusions: The learned features demonstrate superiority over hand-crafted ones when performed on the same dataset. This approach benefits from being fully automatic by fusing feature extraction, selection and classification into a single learning model. It provides a superior strategy that can be used as a tool to guide treatment of OHCA patients in bringing optimal decision of precedence treatment. Furthermore, for encouraging replicability, the dataset has been made publicly available to the research community. © 2020 Elsevier B.V.

Introduction The metabolic syndrome and its influence on coronary artery disease development and progression remains in focus of international research debates, while insulin resistance, which represents its core, is the key component of hypertension, dyslipidaemias, glucose intolerance and obesity. Objective The aim of this study was to establish relationship between basal glucose and insulin levels, insulin sensitivity and lipid panel and the degree of coronary atherosclerosis in non-diabetic patients. Methods The coronary angiograms were evaluated for the presence of significant stenosis, insulin sensitivity was assessed using the intravenous glucose tolerance test with a minimal model according to Bergman, while baseline glucose (G0), insulin (I0) and lipid panel measurements (TC, HDL, LDL, TG) were taken after a 12-hour fasting. Results The protocol encompassed 40 patients (19 men and 21 women) treated at the Institute for Cardiovascular Diseases of the Clinical Centre of Serbia, Belgrade. All were non-diabetics who were divided into 3 groups based on their angios: Group A (6 patients, 15%, with no significant stenosis), Group B (18 patients, 45%, with a single-vessel disease) and Group C (16 patients, 40%, with multi-vessel disease). Presence of lower insulin sensitivity, higher I0 and TC in the group of patients with a more severe degree of coronary atherosclerosis (insulin sensitivity: F=4.279, p=0.023, A vs. C p=0.012, B vs. C p=0.038; I0: F=3.461 p=0.042, A vs. B p=0.045, A vs. C p=0.013; TC: F=2.572, p=0.09), while no significant difference was found for G0, LDL, HDL and TG. Conclusion Baseline insulinaemia, more precisely, fasting hyperinsulinaemia could be a good predictor of significant coronary atherosclerosis in non-diabetic patients, which enables a more elegant cardiometabolic risk assessment in the setting of everyday clinical practice.