Fully Closed Loop Test Environment for Adaptive Implantable Neural Stimulators Using Computational Models

Implantable brain stimulation devices continue to be developed to treat and monitor brain conditions. As the complexity of these devices grows to include adaptive neuromodulation therapy, validating the operation and verifying the correctness of these systems becomes more complicated. The new complexities lie in the functioning of the device being dependent on the interaction with the patient and environmental factors such as noise and artifacts. Here, we present a hardware-in-the-loop (HIL) testing framework that employs computational models of pathological neural dynamics to test adaptive deep brain stimulation (DBS) devices prior to animal or human testing. A brain stimulation and recording electrode array is placed in the saline tank and connected to an adaptive neuromodulation system that measures and processes the synthetic signals and delivers stimulation back into the saline tank. A data acquisition system is used to detect the stimulation and provide feedback to the computational model in order to simulate the effects of stimulation on the neural dynamics. In this study, we used real-time computational models to emulate the dynamics of epileptic seizures observed in the anterior nucleus of the thalamus (ANT) in epilepsy patients and beta band (11-35 Hz) oscillations observed in the subthalamic nucleus (STN) of Parkinson's disease (PD) patients. These models simulated neuronal responses to electrical stimulation pulses and the saline tank tested hardware interactions between the detection algorithms and stimulation interference. We tested and validated the operation of adaptive DBS algorithms for seizure and beta band power suppression embedded in an implantable DBS system (Medtronic Summit RC+S). This study highlights the utility of the proposed hardware-in-the-loop framework to systematically test the adaptive DBS systems in the presence of system aggressors such as environmental noise and stimulation-induced electrical artifacts. This testing procedure can help ensure correctness and robustness of adaptive DBS devices prior to animal and human testing.

Automated summary

This study presents a hardware-in-the-loop testing framework that uses computational models to simulate pathological neural dynamics for testing adaptive deep brain stimulation (DBS) devices. The framework allows for the validation of device functionality by simulating the effects of stimulation on neural dynamics in a saline tank. The proposed testing procedure helps ensure the correctness and robustness of adaptive DBS devices before animal and human testing.