A safe-enhanced fully closed-loop artificial pancreas controller based on deep reinforcement learning achieved an 87.45% median time in range in the Simglucose simulator.
Does a safe-enhanced fully closed-loop artificial pancreas controller based on deep reinforcement learning improve time in range and reduce hypoglycemia in simulated patients with type 1 diabetes?
A novel deep reinforcement learning-based artificial pancreas controller demonstrated high time in range and improved safety in a simulated type 1 diabetes environment.
Patients with type 1 diabetes and their physicians have long desired a fully closed-loop artificial pancreas (AP) system that can alleviate the burden of blood glucose regulation. Although deep reinforcement learning (DRL) methods theoretically enable adaptive insulin dosing control, they face numerous challenges, including safety and training efficiency, which have hindered their clinical application. This paper proposes a safe and efficient adaptive insulin delivery controller based on DRL. It employed ten tricks to enhance the proximal policy optimization (PPO) algorithm, improving training efficiency. Additionally, a dual safety mechanism of 'proactive guidance + reactive correction' was introduced to reduce the risks of hyperglycemia and hypoglycemia and to prevent emergencies. Performance evaluations in the Simglucose simulator demonstrate that the proposed controller achieved an 87.45% time in range (TIR) median, superior to baseline methods, with a lower incidence of hypoglycemia, notably eliminating severe hypoglycemia and treatment failures. These encouraging results indicate that the DRL-based fully closed-loop AP controller has taken an essential step toward clinical implementation.
Zhao et al. (Mon,) conducted a other in Type 1 diabetes. Safe-enhanced fully closed-loop artificial pancreas controller based on deep reinforcement learning vs. Baseline methods was evaluated on Time in range (TIR). A safe-enhanced fully closed-loop artificial pancreas controller based on deep reinforcement learning achieved an 87.45% median time in range in the Simglucose simulator.