NEURAL NETWORK ANALYSIS AS A WAY TO ASSESS SOME PATHOGENETIC ASPECTS OF NEUROINFLAMMATION IN HYPERTENSIVE DISEASE

Highlights Cardiovascular diseases are the leading cause of death in the population. The most common cardiovascular disease is hypertension, which can lead to damage to target organs, including the brain. Several studies have examined the role of specific neuroinflammatory molecules and tumor necrosis factor family proteins in the progression and outcome of vascular-related cerebral damage. The article highlights the possibilities of neural network analysis, which can be used to determine the ranked contribution of certain biomarkers to the pathogenesis of neuroinflammation that develops in hypertension. Abstract Background. Hypertensive disease is accompanied by involvement of target organs, including the brain, in the pathological process, leading to the development of acute and chronic cerebrovascular diseases. The development of modern methods for predicting the development and course of angio-cerebral pathology can be used to create new methods for managing patients at risk. Aim . Using the capabilities of neural network analysis, we determined the ranked contribution of biomarkers of neuroinflammation and proteins of the tumor necrosis factor family to the pathogenesis of neuroinflammation in the context of hypertension. Methods. The study involved 80 participants. The control group consisted of healthy individuals. The first group included patients with stage II–III hypertension, achieved target blood pressure levels, and a 3–4 risk of developing cardiovascular complications. The second group consisted of patients with stage II–III hypertension, unachieved target blood pressure levels, and a 3–4 risk of developing cardiovascular complications. Group 3 included patients with a verified diagnosis of ischemic (atherothrombotic) stroke. The data obtained from the study were used to train a multilayer perceptron and create a mathematical model that determines the ranked contribution of biomarkers of neuroinflammation and proteins of the tumor necrosis factor family to the pathogenesis of neuroinflammation in the context of hypertension. Results. According to the results of the conducted multiplex analysis and statistical data processing, similar shifts in the neuroimmune status were identified in both patients with ischemic stroke and individuals with hypertension. Modern developments in the field of predicting the development and course of angiocerebral pathology can be used to create new methods for managing patients at risk. Conclusion. The data obtained as a result of the conducted study indicate changes in the neuroimmune status of patients with hypertension. It is noteworthy that the content of biomarkers associated with neuroinflammation in patients with unattained target blood pressure levels is as close as possible to that of patients with ischemic stroke. At the same time, in the group of patients with achieved target blood pressure levels, there is also a shift in the studied indicators, but to a lesser extent, which may indicate the presence of a neuroinflammatory process that is present even in the presence of antihypertensive therapy and blood pressure control. The mathematical model developed using neural network analysis allowed us to determine the ranked contribution of the studied parameters to the pathogenesis of neuroinflammation in patients with hypertension. These results can be further used for a more personalized approach to patient treatment by influencing specific aspects of the pathogenesis, thereby moderating the ongoing process and influencing the future course of the disease.