Skip Navigation
Search

Back to Search Innovations

A Closed-Loop Stimulation Controller for Exploring Neural States


A brand-new type of reinforcement learning algorithm

Tech Image

Free Photos, Pixabay, CC0

Background


Current day algorithms governing how medical devices deliver stimulation to patients have not changed in well over 50 years and cannot account for neurological variabilities across individuals that come about from differences in genetics, lifestyle factors, and sustained injuries. As a result, clinicians are experiencing highly inconsistent patient outcomes, with 30%-90% of individuals not responding to treatment at all, depending on the disease. A large quantity of research in the fields of computational and systems neuroscience has demonstrated that the clinical state-of-the-art is insufficient to remedy many pathologies and must be re-invented for better results.

Technology


Escape-the-Maze (EtM), is a brand-new type of reinforcement learning algorithm. EtM is the first of its kind and is installed on existing medical devices in a closed-loop recording-stimulating setup with the brain circuit of interest, where it leverages the patient’s unique neural activity to guide the electrical activity out of the unwanted state e.g coma. Due to EtM’s versatility, there are a wide array of applicable markets associated with neuropsychiatric and movement disorders such as obsessive compulsive disorder (OCD), as well as disorders of consciousness.

Advantages


Personalized algorithms to treat various disorders - Automatically modulates stimulus applied by medical devices - Hardware agnostic (both invasive and non-invasive treatments)

Application


Treatment of neuropsychiatric, movement, and consciousness disorders.

Inventors

Ian Jordan, , Department of Applied Mathematics and Statistics
Il Park, Assistant Professor, Neurobiology and Behavior
Josue Nassar, , Electrical and Computer Engineering

Licensing Potential


Development partner - Commercial partner - Licensing

Licensing Status


Available

Licensing Contact

Valery Matthys, Licensing Associate, Intellectual Property Partners, valery.matthys@stonybrook.edu,

Patent Status


Patent Application Published: WO2023/196280

Stage of Development


In-silico data

Tech ID

050-9281