M.V. Voitikova - Ph.D. (Phys.-Math.), Leading Research Scientist , Institute of Physics, National Academy of Sciences, , Minsk, Belarus. E-mail: voitikova@imaph.bas-net.by
R.V. Khursa - Ph.D. (Med.), Assoсiate Professor, Belarusian Medical State University, Minsk, Belarus. E-mail: Rvkhursa@tut.by

This paper presents a nomogram for classifying of the hemodynamic states, based on linear regression modeling of blood pressure (BP) parameters and Data Mining algorithm called Support Vector Machine (SVM). We analyzed the BP recordings for day, night and 24-h periods. The regression coefficients as the information patterns are compared with a library of hemodynamic samples of the patients with known diagnoses. The so-called feature vector, whose coordinates are the linear regression coefficients of the systolic and diastolic pressure on pulse pressure, is applied to the nomogram. Determined position of the vector in limited area on the nomogram corresponds to normal hemodynamics of the cardiovascular system. The pathological changes of hemodynamics inherent in hypertension, hypotension or clinically latent hemodynamic disturbances can be diagnosed according a position of the feature vector on the nomogram.