A wireless, 60-channel, AI-enabled neurostimulation platform

Area of Science:

• Neuroscience
• Biomedical Engineering
• Neuromodulation

Background:

• Closed-loop neuromodulatory therapies necessitate advanced devices capable of real-time brain state decoding and multi-site stimulation.
• Current technologies often lack the integration required for seamless operation in freely moving subjects.

Purpose of the Study:

• To introduce and evaluate the Smart Neurostimulation System (SNS), a novel implantable device for closed-loop neuromodulation.
• To demonstrate the device's capability for high-density neural recording, brain state decoding, and programmable stimulation in vivo.

Main Methods:

• The Smart Neurostimulation System (SNS) is a cranially mounted implant featuring 60 configurable recording/stimulation channels, inductive power, and onboard spectral-feature classification.
• Experiments were conducted in three freely-moving sheep, involving streaming of local-field potentials and parameter-sweep stimulation protocols.

Main Results:

• Movement classifiers achieved high accuracy (average AUC > 0.95) through cross-validation.
• Stimulation with varying amplitude and frequency resulted in observable increases in alpha and gamma band power in targeted cortical areas.

Conclusions:

• The SNS successfully integrates high-density sensing, real-time brain state decoding, and closed-loop stimulation in a single, implantable device.
• The study demonstrates the potential for biomarker-guided stimulation targeting distributed cortical networks crucial for memory and cognition.