ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login


Journal: NAUKA MOLODYKH (Eruditio Juvenium) (Vol.7, No. 4)

Publication Date:

Authors : ;

Page : 533-540

Keywords : hemodynamically significant stenosis; predictors; neurological deficits; internal carotid arteries; vertebral arteries; machine learning technology;

Source : Downloadexternal Find it from : Google Scholarexternal


Aim. The study aims to describe possible predictors of neurological deficits in operated patients with hemodynamically significant stenosis of internal carotid and vertebral arteries based on clinical, instrumental and laboratory findings as well as machine learning technologies. Material and Methods. The study involved 49 patients with hemodynamically significant atherosclerotic stenotic lesions of main blood vessels of the neck and head: 34 subjects underwent carotid endarterectomy, and 15 patients underwent reimplantation of the vertebral arteries. Vascular ultrasound scanning was performed to assess the degree of stenosis by area, linear blood velocity in common carotid, internal carotid, and vertebral arteries on both sides; echocardiography estimated systolic function scores (ejection fraction), left ventricular end-diastolic and end-systolic diameters. All patients underwent a standard 12-lead electrocardiography. Haemostatic parameters, including fibrinogen levels, prothrombin time, partial thromboplastin time, international normalized ratio, prothrombin index, and platelets were evaluated. Biochemical risk factors of impaired cerebral circulation such as blood glucose and cholesterol were also evaluated. Results. Based on the neurological assessment and neuroimaging findings, the patients were alloca-ted into 2 subgroups: with neurological deficit due to ischemic brain lesions and without such disorders. We created an artificial neural network, which is capable of classifying patients into subgroups on the basis of the hemodynamic parameters in main blood vessels of neck, laboratory data, and systolic cardiac function. The artificial neural network has satisfactory classification abilities. Patterns were established between impaired vertebrobasilar cerebral circulation and neurological deficits. Statistically significant trends were observed for hemostatic parameters in terms of hypercoagulation as well as systemic hemodynamic parameters characterized by cardiac pumping function. Statistically significant results were also obtained for the degree of carotid artery stenosis. Conclusion. The most significant predictors of neurologic deficit in patients with stenosis of blood vessels of the neck undergoing surgical treatment are hemodynamic parameters in vertebral arteries, hemostatic parameters, and impaired pumping function of the heart. Prognosis-based machine learning technologies allow effective prediction of neurological deficits in patients with hemodynamically significant stenosis of main arteries of the neck.

Last modified: 2020-01-09 16:12:30