DENTAL CARIES PROGNOSIS BY NEURAL NETWORK COMPUTER TECHNOLOGIES

Computer technologies are widely implemented in clinical dental practice. The use of computer neural network programs in predicting dental caries as the most common dental disease is quite relevant.

The aim – to study the effectiveness of using the “CariesPro” computer program developed using neural network technologies in the individual prediction of dental caries in persons of all ages.

Materials and methods. We examined 73 persons aged 6–7, 12–15 and 35–44 years, in which the intensity of dental caries was determined taking into account the number of cavities, the hygiene condition of the oral cavity, the structural and functional acid resistance of the enamel of the teeth according to the enamel resistance test and its functional component. The data were added to a neural based computer software program “CariesPro” designed to predict dental caries. After 1 year, a second examination was performed and the dental caries obtained were compared with the individually predicted computer program.

Results. The highest intensity of dental caries was found in persons aged 35–44 – 6.69±0.38, in children 6–7 and 12–15 years it was 3.85±0.27 and 2.15±0.24, respectively (p <0.05). After 1 year, the corresponding intensity indices for persons of these age categories were 8.92±0.52; 6.27±0.35 and 4.23±0.2. The growth rates of caries intensity were, respectively, 2.23±0.25; 2.42±0.15 and 2.09±0.15. After comparing the re-survey data with the computer-programmed estimate, the probable number of carious cavities was found to be 61 true and 12 false predictions from the entire sample, the prediction accuracy of the constructed and trained neural network was 83.56 %.

Conclusion. The “CariesPro” computer program, developed using neural network technologies, allows to predict the number of carious lesions in a year with a probability of 83.56 %.

FULL PAPER

How to cite paper:

Udod, O., & Voronina, H. (2019). DENTAL CARIES PROGNOSIS BY NEURAL NETWORK COMPUTER TECHNOLOGIES. EUREKA: Health Sciences, 0(6), 15-21. doi:http://dx.doi.org/10.21303/2504-5679.2019.001070