In adolescent idiopathic scoliosis, the selection of an optimal instrumentation configuration for correcting a specific spinal deformity is a challenging combinatorial problem. Current methods mostly rely on surgeons' expertise, which has been shown to lead to different treatment strategies for the same patients. In this work, a mathematical model of the human spine derived from in-vitro experimentally-obtained data was used to simulate the biomechanical behavior of the spine under the application of corrective forces and torques. The corrective forces and torques were optimized based on the particle swarm optimization algorithm for each combinatorially possible instrumentation strategy. Finally, a multi-criteria decision support for optimal instrumentation in scoliosis spine surgery has been proposed and applied to five patient data sets exhibiting similar spinal deformities according to two commonly used classification systems. Results indicated that the classification of the spinal deformities based on the current standardized clinical classifications systems is not a sufficient condition for recommending selective fusion of spinal motion segments. In addition, the particle swarm optimization algorithm was successfully applied to solve a realistic interdisciplinary clinical problem in a patient-specific fashion. The proposed method enables a better understanding of the biomechanical behavior of spinal structures and has the potential to become a standard tool in preoperative planning.