Otimização da orientação de mini-implante para minimizar a possibilidade de falha devido à perda óssea : um procedimento usando algoritmo genético e análise de elementos finitos

Abstract

From the need for replacement, or repair of human bone structures, the use of implants has become an increasingly common practice in humans. Even with the development and research of methods and materials more technologically advanced, some procedures still become difficult to apply. The analysis performed in the patient itself becomes unfeasible and invasive. This work presented a methodology for the creation of three-dimensional computational models of bone structures from computed tomography. Stress distribution studies of von-Mises were performed in a global model of a mandible to identify regions subject to stress. Later, the methodology was implemented using the genetic algorithm method to find the best orientation of osseointegrated mandibular mini-implants, aiming at minimizing the mean value of bone deformation, recurring high levels of stress. With the optimization algorithm it was possible to find optimal insertion angles of the mini-implant in a computational model generated by means of computational tomographies decreasing the risk of failure of the mini-implant due to bone loss. Finally, a comparison between the genetic algorithm method and the quadratic sequential programming method was performed, in order to demonstrate the efficacy of the genetic algorithm. The genetic algorithm method presented low efficiency in relation to the quadratic sequential programming method due to the computational cost used to implement the routine, but when compared to the efficacy of each method, the genetic algorithm method had a better objective function value, showing an efficacy of 42%, higher.