Application of self-learning evolutionary algorithm for optimal design of a porous polymethylmethacrylate scaffold fabricated by laser drilling process

Fabrication of scaffolds that closely mimic the biomechanical properties of the surrounding bone is one of the main issues in designing an orthopedic implant. This study is aimed to develop an optimal predefined three-dimensional polymethylmethacrylate scaffold as a potential bone substitute via laser drilling technique, which is a novel method of fabricating bone scaffold. To achieve this goal, using the obtained experimental data, the authors have tuned an adaptive neuro fuzzy inference system, which predicts the hole diameter and its depth as two process outputs for any given values of laser power and laser irradiation time. In addition to this adaptive neuro fuzzy inference system model, some experimental models were also proposed to estimate the scaffold porosity and the mechanical strength regarding the drilled hole geometry. To design an optimal scaffold with the highest amount of strength and porosity, a Pareto-based self-learning evolutionary algorithm called SOPEA was proposed. To enhance its ability in finding the optimal Pareto front, it was equipped with the adaptive self-organizing map. Based on the obtained Pareto front, a trade-off has been made in order to find a solution that yields deliberate operational properties. The scaffold morphology was observed by scanning electron microscopy images and its mechanical property was determined using a compressive test. The versatility provided by this technique allows the fabrication of scaffolds with controllable amount of porosity, pore size, and mechanical properties while providing fully interconnected channel network. The results also indicate that the proposed model predicts the process outputs with acceptable accuracy and the proposed optimization algorithm is able to find the optimal solution effectively.