| |
DOG Congress Home
Invitation
Organization, Deadlines
Overview of the Congress
Scientific Programme
Opening Ceremony
Ceremony 150 Years of DOG
Thursday, 20.September
Friday, 21.September
Saturday, 22.September
Sunday, 23.September
Poster Sessions
Symposia
Courses
Satellite Programme
Information
Social Programme
Sponsors, Exhibitors
DOG Homepage
|
|
AbstractP 091 Use of information technologies in the solution of classification problems Dragomiretskaya E., Kovylina I.
Filatov Institute of eye diseases and tissue therapy, Odessa, Ukraine Objective: The solution of classification problems of the research objects by many signs let us possibility to order them so, that the objects most similar among themselves made one class. The purpose of carried out investigation is to pick out the main kinds of pupil reaction (PR) by means of an automatic multivariate grouping in the group of patients with cataract and with cataract complicated pseudoexofoliative syndrome (PES).
Methods: Results of 248 pupillogramm (18 amplitude und times charackteristic) of patients with cataract and cataract complicated pseudoexofoliative syndrome are included in the investigation. The data are received before cataract phackoemulsification was worked out. The classification is carried out with use of cluster method - method k-means.
Results: Cluster analysis have given us the possibility to find 4 cluster, which have different kind of PR. The first cluster differs by higher parameters of amplitude and speed of pupil miosis, and also low time of restoration of a pupil (n=30). In the second cluster all values of the standardized parameters are near to a zero line, i.e. are close to average values of a studied cohort (n=87). In the third cluster - the maximal area of the pupil is a little above average value, however are marked the highest values of the minimal area (after a flare) and the lowest amplitude of the area of a pupil (n=34). In the fourth cluster the both maximal, and minimal area of a pupil are lowered and the greatest time of restoration of a pupil is marked (n=97). The 1 and 2 kind of PR we have found more often (69,5%) at patients with cataract not complicated with PES. The 3 and 4 kind of PR are more often (63%) at patients with cataract and PES (c2=50,03; p=0,00001).
Conclusions: Four kinds of PR are allocated and described with use of cluster analyses at patients with cataract with and without PES. These kinds of PR are significantly connected with presence of PES.
|
|
