Systems, methods, and devices for intracranial measurement, stimulation, and generation of brain state models

Inventors

Pradeep, Anantha K.

Assignees

Stimscience Inc • StimScience Inc

Abstract

Provided are systems, methods, and devices for intracranial measurement, stimulation, and generation of brain state models. Systems include a plurality of intracranial electrodes configured to be coupled to a brain of a user. Systems further include an interface configured to obtain measurements from the plurality of intracranial electrodes. Systems include a first processing device including one or more processors configured to generate a plurality of brain state parameters characterizing one or more features of at least one brain state of the user, and a second processing device including one or more processors configured to generate at least one model of the brain of the user based, at least in part, on the plurality of brain state parameters and the measurements. Systems include a controller including one or more processors configured to generate a control signal based on the plurality of brain state parameters and the at least one model.

Core Innovation

The invention relates to a closed-loop system using a plurality of intracranial electrodes configured to obtain a plurality of measurements from a brain of a user, wherein the measurements include an oscillation or resonance frequency. An interface obtains the plurality of measurements from at least the intracranial electrodes, and one or more processors generate a plurality of brain state parameters characterizing one or more brain states of the user.

The processors further generate at least one model of the brain of the user using machine learning, based at least in part on the brain state parameters and the plurality of measurements. The at least one model is configured to model input-output behavior of the brain, and the system generates a control signal based on the plurality of brain state parameters and the at least one model to obtain a desired brain state by changing the oscillation or resonance frequency.

In the described system and device variants, brain-state characterization and modeling can incorporate additional measurement inputs, including co-registered scalp-level measurements. The model generation can use functional and structural brain models in combination, and adaptive operation is described by updating at least one model using a plurality of additional measurements received through a plurality of electrodes.

Claims Coverage

The independent claim set comprises three independent claims that cover: a complete system with intracranial electrodes, brain-state parameter generation, machine-learning input-output modeling, and control signal generation based on changing an oscillation or resonance frequency; a device architecture separating processors and controller to generate parameters, models, and a control signal; and a method that performs these steps. Across the independent claims, at least three core inventive features recur: intracranial measurements including oscillation or resonance frequency, machine-learning input-output behavior modeling driven by brain state parameters and measurements, and control signal generation to obtain a desired brain state by changing the oscillation or resonance frequency.

The independent claims collectively cover obtaining intracranial measurements including an oscillation or resonance frequency, generating brain-state parameters, training/generating a machine-learning model configured to model brain input-output behavior from the parameters and measurements, and producing a control signal that targets a desired brain state by changing the oscillation or resonance frequency.

Stated Advantages

Documented Applications