[The multivariate analysis of clinico-laboratory data in stomatology]

The authors demonstrate the potentialities of multidimensional analysis methods: component, cluster, and discriminant as exemplified by clinico-laboratory examinations of 26 patients with medium-severe periodontitis and concomitant diseases. A two-stage modification of multidimensional grouping of biomedical objects is suggested. The first stage implies detection of the principal components and their estimation for each examinee. The second stage consists in realization of multidimensional grouping method according to G. S. Kil'dishev and Iu. I. Abolentsev in relation to individual component values. Different typologically homogeneous groups of patients with a specific microbiologic and biochemical status evidently requiring a differentiated approach to treatment may be singled out within the frames of a current classification of periodontal diseases basing on clinical and laboratory data.