Direct Brain Interface-Quality of Life Technology Center - Carnegie Mellon University

Direct Brain Interface

The overall goal of this project is to develop the prototype of a brain-computer interface (BCI) system based on electrocorticography (ECoG) and test its feasibility on patients with severe motor disabilities through the collective effort of investigators across multiple institutions, such as University of Pittsburgh, Carnegie Mellon University, Washington University in St. Louis and University of Wisconsin-Madison.

This project is one of the major research themes in the Human-System Integration Thrust, which focuses on dynamically adjustable human-system interface.  This interface needs to have the capability of inferring the state or desire of a human being based on various physical and physiological measurements and dynamically adjusting its behavior.  The ECoG-based BCI system being developed by this project is a perfect example of this dynamic human-system interface.  It directly records ensemble neuronal activity through ECoG recording.  By processing and analyzing the recorded ECoG signals in real-time, the state of the brain, which is the desired movement that a user wants to generate, is decoded by this BCI system and used to drive a computer cursor, a prosthetic arm, or a wheelchair.  This BCI system is also dynamic, because it periodically evaluates its performance against recorded cortical activity and adjusts its decoding parameters that it uses to extract the intended movement from cortical activity. This project is translational in nature, as it aims to develop a practical BCI device and bring the technology from laboratories to users with severe motor disabilities.  BCI technology will enable those users to interact with computers and assistive devices intuitively, easily, and efficiently.  This will significantly improve their quality of life. 

People with severe motor disabilities, such as high-level spinal cord injury (SCI), amyotrophic lateral sclerosis (ALS), and brainstem stroke, have a very limited means to communicate with outside world.  BCI technology attempts to establish direct communication between the brain and external devices to provide faster communication rate and to eventually restore volitional limb movement.  The core of this technology is to extract reliable control signals from neural activities.  Single neuron activity can be recorded by inserting microelectrodes into the cortex, but this is invasive, and the recorded signal lacks long-term stability.  Alternatively, electroencephalography (EEG) can be recorded non-invasively, but it has very limited information content due to the low-pass filtering property of the skull.  Recently, ECoG has received attention as a promising modality for BCI application.

Project Team

  • Wei Wang, Lead
  • Doug Weber, Lead
  • Mike Boninger
  • Andrew Schwartz
  • Elizabeth Tyler-Kabara
  • Anto Bagic
  • Gary Fedder
  • Xin Li
  • Jeyanandh Paramesh
  • Tina Harrison
  • Ramana Vinjamuri
  • Sam Clanton
  • Jennifer Colinger
  • Alan Degenhart
  • John Kelly
  • Gustavo Sudre