Development and In Silico Evaluation of a Model-Based Closed-Loop Fluid Resuscitation Control Algorithm

Objective : To develop and evaluate in silico, a model-based closed-loop fluid resuscitation control algorithm via blood volume feedback. Methods: A model-based adaptive control algorithm for fluid resuscitation was developed by leveraging a low-order lumped-parameter blood volume dynamics model, and then, in silico evaluated based on a detailed mechanistic model of circulatory physiology. The algorithm operates in two steps: 1) the blood volume dynamics model is individualized based on the patient's fractional blood volume response to an initial fluid bolus via system identification; and 2) an adaptive control law built on the individualized blood volume dynamics model regulates the blood volume of the patient. Results: The algorithm was able to track the blood volume set point as well as accurately estimate and monitor the patient's absolute blood volume level. The algorithm significantly outperformed a population-based proportional-integral-derivative control. Conclusion: Model-based development of closed-loop fluid resuscitation control algorithms may enable the regulation of blood volume and monitoring of absolute blood volume level. Significance: Model-based closed-loop fluid resuscitation algorithm may offer opportunities for standardized and patient-tailored therapy and reduction of clinician workload.