A Distributed Artificial Immune Network for Optimizing Tracer Kinetic Models with MATLAB Distributed Computing Engine

Artificial immune networks (AIN) as one of the new intelligent soft computing methods have been widely used in many application fields. The AIN shows good ability of global optimization, especially in parameter optimization of pharmacokinetic models. The AIN search global optimum based on the principles of clone selection and immune network. However, as one of the heuristic-based optimal algorithms, the evolution of memory cells in the AIN is time consuming compared with gradient-based optimal algorithms. In this paper, a distributed AIN with distributed clone selection evolutionary strategy is proposed to improve the efficiency of the AIN. Then a distributed artificial immune network is implemented with MATLAB Distributed Computing Engine. One of the advantages of MDCE is that it is easy to run optimal algorithms programmed in MATLAB platform. In the experiments, parameters of the [¹⁸F] Fluoro-2-deoxy2D-glucose (FDG) tracer kinetic model are optimized with the distributed AIN. Experimental efficiency of the algorithm is discussed.