May 29, 2012 – 1:22 am
Researchers at the National Institute for Physiological Sciences (NIPS) in Japan have accurately reconstructed intracortical local field potentials (LFPs) by capturing electrocorticogram (ECoG) signals from the brain surface. The technique may lead to the development of a high-performance brain-machine interface that is less invasive than current methods.
Neural signals from individual ECoG channels tend to be coarse and limited in their capability to estimate deep-layer LFPs. To probe neural activity in the brain using conventional technology, medical practitioners implant electrodes in the brain, which carries the risk of causing injury.
The research team led by NIPSs Hidenori Watnabe has developed a high-density, 32-channel, micro-ECoG array and applied a sparse linear regression algorithm to reconstruct LFPs at various depths of a monkeys primary motor cortex as it performs a reach-and-grasp task. A time-frequency analysis of the reconstructed LFP shows a clear transition between the monkey’s resting and movement phases. The result heightened expectations that neural activity in the brain could be measured without inserting electrodes.
The study was published in the online edition ofJournal of Neural Engineering on 9 May. This NIPS-led project was funded by the Ministry of Education, Culture, Sports, Science and Technology, as part of the ministrys Strategic Research Programme for Brain Sciences.