Development of an Algorithm for Classifying the State of the Oral Mucosa Based on the Electrochemical Parameters of Saliva

The aim of the study is to develop an algorithm for classifying the state of the oral mucosa based on the electrochemical parameters of saliva. The algorithm is a trained binary decision tree. As the initial data for training, the indicators of the work of the current (Q1-Q4) in saliva obtained by the joulmetric method, as well as the class of the state of the object under study, were used. Fields Q1-Q4 represent the value of the current work obtained in the joulemetric study of saliva. Parameters of impact on the biological object: pulse duration 3 s, current strength 0.01 mA, 0.02 mA, 0.03 mA, 0.05 mA for Q1-Q4, respectively. The class of the state of the object under study refers the set of parameters Q1-Q4 to one of two classes: the study group (people with an inflammatory disease of the oral cavity), the control group (people with a normal state of the oral mucosa). As a result of the study, 100 decision trees were built. Of these, two turned out to be the most accurate: a tree classifying the state of a biological object depending on the value of the parameter Q1, and a tree classifying the state of a biological object based on the parameters Q3, Q4.