Design and Implementation of a Hybrid GA-ANN System for Predecting the Size of Unerupted Canines and Premolars in the Mixed Dentition Period Other biomedical engineering topics

In this paper a hybrid GA - ANN algorithm is developed for prediciting the size unerupted teeth in both madible and maxillo. The algorithm is built upon the measurements carried out in 106 dental casts. The GA and ANN work simultaneously . The role of the GAis to select the best possible refrerences for the prediction task and its outputs are fed into the ANN, while the ANNseeks the best possible mapping of the reference inputs to the target values. It is illustrated that the algorithm is capable of predicting the size of unerupted teeth in mandible and maxilla with mean error rates of 2.7% and 3% respectively . Agraphical user interface is also developed for utilizing the algorithm capabilities, using which the dentist may upload digital photographs of the casts into the system, instead of handling the manual measurements, and proceed through different phases of the software.

Keywords:

Genetic algorithm , artificial neural network , prediction , unerupted teeth size , hybrid algorithm , graphical user interface