Statistical forecasting when creating the quick technique of determining a level of regulatory and adaptive status of the human body

Abstract

In order to improve the existing incremental principle of determining the cardiac and respiratory rhythm synchronization (CRS) range the 'quick technique' of determining the index of regulatory-adaptive status (IRAS) of the human body has been proposed. In order to reduce the number of samples we proposed to apply the statistical forecast of regulatory and adaptive status (RAS) (from 'high' to 'low') on the basis of its least-consuming measurements or by fixing the CRS parameters. The following factors/arguments have been examined as the basis of the forecast: the minimum boundary of the synchronization range, duration of evolution synchronization at the minimum boundary, the following categories also have been taken into account: initial respiratory rate, initial heart rate, age, height, weight, gender, menstrual cycle phase, diastolic and systolic blood pressure values. Linear discriminant analysis and neural network (NN) classification techniques - multilayer perceptron and network of radial basis function type have been analyzed. Neural network classification has shown the best accuracy. Assessment of accuracy and confidence level of the forecast has been performed. On the basis of the NN-classification RAS level has been proposed curtailed incremental definition of IRAS technique: possible to complete the examination in 1 stage (probability 70%), 2 stages (probability 90%), or 3 (control) stages (samples), it is depending on the subject's observed RAS level.