ACTA FAC. MED. NAISS. 2007; 24 (1): 3-7

   Original article

Dariusz Świetlik¹, Cezary Kłosiński²
 

¹Laboratory of Medical Informatics and Neural Networks, Medical University, Gdańsk, Poland
²Department of Prosthodontic Dentistry Medical University, Gdańsk, Poland
 

   SUMMARY

    The aim of the study was to demonstrate the usefulness of artificial neural networks for predicting the number of follow-up visits (number of corrections).We used clinical data of 82 patients 48-90 years of age, mean age 66 ±9 years (38 males and 44 females) for whom a total of 164 complete dentures was performed (82 upper complete dentures and 82 lower complete dentures). The actual number of corrections was compared with the number obtained using a neural network. Pearson's linear correlation coefficient (r) was 0.86 ( < 0.05) and the coefficient of determination was 0.74. Artificial neural networks are a useful tool which employs artificial intelligence techniques for the prediction of the numberof follow-up visits.
 

Key words:artificial neural networks, complete denture, prosthetic treatment, edentulous jaws, adaptation