Studies show that 85% of deaths from cardiovascular diseases are due to heart attacks and strokes. Early detection of possible hemodynamic problems allows timely measures to be taken to prevent such complications.
As part of the scientific research of the Laboratory of Physiology of Biofeedback of the Institute of Experimental Medicine and the completion of her master's thesis, SUAI student Elizaveta Simonova created a new algorithm based on the machine learning method for recognizing human conditions based on recorded hemodynamic parameters. The algorithm allows you to predict changes in human health under various conditions more accurately.
The study showed that the indicators of the cardiovascular system (systolic blood pressure, diastolic blood pressure, cardiac stroke volume, heart rate) have different ranges of normal values in subjects in various hemodynamic conditions. A special Python program was developed to evaluate the results of the study and automate the diagnosis of the development of disorders in the functioning of the cardiovascular system. The program implements the Random Forest machine learning method to train a classification model based on the individual data of each subject.
The study is important for the further development of methods for monitoring and diagnosing cardiovascular diseases, it allows detecting early circulatory disorders and taking personalized prevention and treatment measures depending on the health status of patients. The program can be integrated into medical monitoring systems to improve the quality of diagnosis and treatment of patients.
