2022-192 Software Simulation of Brain-Machine Interfaces

Summary:

UCLA researchers in the Department of Electrical and Computer Engineering have developed an entirely software-based brain-machine interface (BMI) simulator that provides tools for the acceleration and optimization of BMI development. 

Background:

Brain-machine interfaces (BMIs) provide a direct communication pathway between the brain’s electrical activity and an external device. BMIs have remained in pilot clinical trials for years due to the lack of technology that enables larger communities to design and optimize BMIs. Current technologies to study BMIs require expensive equipment and lengthy experiments, which reduces the pace of BMI research and limits accessibility to only a specific set of laboratories. Therefore, there is a need for tools that accelerate BMI development. 

Innovation:
 
UCLA researchers in the Department of Electrical and Computer Engineering have developed a method that utilizes deep learning to create a BMI simulator. This method uses human-like artificial intelligence to efficiently interact with decoders in a simulated BMI environment, replacing the need for a human or macaque test subject. It enables rapid and efficient optimization of BMI decoders, a process which traditionally leads to the high cost and lengthy experimental times associated with BMI research. The method incorporates timely advances in systems neuroscience and deep learning to solve fundamental challenges in BMI simulation. Since this method requires software rather than physical equipment, it would help remove the constraints of costly equipment and increase the accessibility of BMI research to more laboratories. 

Patent:

Systems and methods for simulating brain-computer interfaces

Potential Applications:

•    Robotics simulation
•    Virtual laboratories 
•    Human-computer interaction (HCI)

Advantages:

•    Based on open-source software
•    Cost-effective
•    Rapid research development 

Development to Date:

First description of complete invention.

Related Papers: 

Liang K-F, Kao JC. Deep Learning Neural Encoders for Motor Cortex. IEEE Trans Biomed Eng. 2020;67: 2145–2158. DOI: 10.1109/TBME.2019.2955722

Cunningham JP, Nuyujukian P, Gilja V, Chestek CA, Ryu SI, Shenoy KV. A closed-loop human simulator for investigating the role of feedback control in brain-machine interfaces. J Neurophysiol. 2011;105: 1932–1949.   DOI: 10.1152/jn.00503.2010

Patent Information:
For More Information:
Joel Kehle
Business Development Officer
joel.kehle@tdg.ucla.edu
Inventors:
Jonathan Kao
Ken-Fu Liang