March 31, 2010
A team of Carnegie Mellon researchers recently won the Biomag Data Analysis Competition for movement decoding of the brain-computer interface. Assistant Research Professor of ECE Xin Li, ECE PhD student Jinyin Zhang and Gustavo Sudre, a PhD student in the university's Center for the Neural Basis of Cognition, developed a new decoding algorithm to accurately predict movement directions from magnetoencephalography signals (i.e., the magnetic field generated by the human brain).
Movement decoding is a key signal processing step for the brain-computer interface that aims to develop an alternative communication path that will enable human subjects to directly control external devices (like a computer mouse or wheelchair) through brain signals. Such a technology would substantially improve the life quality of human patients with either limb amputation or neurological disorders. For the Biomag competition, the Carnegie Mellon team's algorithm was facilitated by exploring the unique spatial correlation of a magnetic field, and it resulted in the best decoding accuracy among all participating teams.
The data analysis competition is associated with the 17th International Conference on Biomagnetism (Biomag 2010). The winners will receive a 500 Euro prize and present their results at Biomag 2010, which is being held this week in Croatia. For more on Biomag 2010, visit www.biomag2010.org.
Movement decoding for brain computer interface: predict intended movement behavior by measuring brain signals from human subjects.