Another and perhaps more important application of the movable probe concept is
in so-called
chronic recordings, where electrodes are surgically inserted into
the brain tissue. The advantage of this method is that the microelectrodes
could be placed in the brain area of interest and they can gather large
volumes of data in a relatively short time. The major difficulties in chronic recordings include
the deterioration of performance over time, very often due to encapsulation
of the electrodes by scar tissue as well as inflammatory processes in the
brain. Sometimes, the receptive fields of the neurons that are recorded from
are non-optimal and not particularly tuned to the variable of interest.
Therefore, it might be advantageous to be able to move or reposition the
microelectrodes within the implant, break through encapsulation or simply look
for a cell with better tuning properties.
Here is a list of people involved in this project:
- Eddie Branchaud
- Joel Burdick
- Zoran Nenadic
I have developed a simulated environment for testing the movable probe
algorithm. The following slide summarizes the algorithm:
(please click on the slide for a full size image)
Take a look at some animations of movable probes
- Stochastic gradient animation (0.5MB, duration: 22 sec)
- Basis functions animation (0.5MB, duration: 17.5 sec)