Cooling And Temperature Control Algorithms For Brain Hypothermia

A large number of clinical and experimental trails demonstrated that brain hypothermia is neural protective for brain injury therapy. To maximize the protection and reduce the side effects, how to accurately control the brain temperature during treatment by using brain surface cooling is a significantly urgent problem now. In this paper, by means of theoretical analysis, simulations and animal experiments, an adaptive brain temperature control method for brain hypothermia was proposed.A thermal model of brain was first established based on the Pennes bio-heat transfer equation. The thermal characterization of brain thus could be analyzed through simulation. The simulation results indicated that the brain hypothermia tend to have nonlinear, time-delay and time varying characteristics. An adaptive control method, which combines a fuzzy neural network with model reference control for brain hypothermia, was consequently proposed to control this particular system, possibly avoiding the shortages of conventional control methods such as PID and fuzzy controls. Fuzzy neural network is a fuzzy system constructed with neural network; possibly possess both expert-knowledge-based characteristics of fuzzy logic, and self-learning ability of neural network. The control system used a first-order linear system as the reference model and a variable step size BP algorithm as the learning algorithm, and can online learn and adjust the weight coefficients and membership function parameters.The control method was furthermore extended to develop an experimental setup for brain hypothermia and furthermore performed the evaluation through animal experiments of brain temperature control via this system. The SD rats were used as model animal and brain temperature control experiments were performed by using of both the conventional control methods and the adaptive control method proposed. Experimental results showed that the adaptive control method can reach better precision, stability, as well as robustness compared to the conventional control methods, such as PID and fuzzy controls. Clearly the proposed method can adapt to the nonlinear, time-delay and time varying characteristics of the brain hypothermia. Therein, the theoretical and experimental obtaining from this work proved the application of adaptive control method in the clinical treatment of brain injury.