Experiment using optical stimulation of a cortical network


Coupling within a cortical network using optical stimulation in a multi-electrode array. Spontaneous firing at source electrode #26 (Row 6, Column 2) is used to trigger transmitter uncaging at target electrode #77 (Row 7, Column 7).



Beagle Supercomputer

We often use this 151 teraflops, Cray XE6 system that supports computation, simulation, and data analysis for our biomedical research. You may read more about this supercomputer here.

Results of modeling epileptic activity

This is a summary movie of a simulation. The top part depicts the activities of individual neurons. The bottom traces show EEG and local field potential contributions of different cell types.




This movie shows activities during a simulated seizure in a neocortical network. The trace in the bottom is the EEG showing onset of oscillations. Each plane represents one cell type. From top to bottom: superficial and deep pyramidal cells, three types of basket cell and the chandelier cell.


A simulation of a focal patch that entrains three other areas in the brain. The individual EEGs for each patch is depicted in the bottom panel and the activity in each patch is depicted in the four quadrants in the top display.




Verdandi is a C++ parallel network simulator based on both distributed- (MPI) and shared-memory (OpenMP) protocols. It is a general purpose simulator described in Visser et al. (2010). It was utilized for  simulations up to 400 thousand neurons and 600 million synapses. It has been show to scale very well up to a few hundred processors even if the scalability of its shared memory approach is far from perfect, even if still useful. In its current implementation is lacks both gap-junctions and long range connections.

Visser, S., Meijer, H. G. E., Lee, H. C., Van Drongelen, W., Van Putten, M. J. A. M., & Van Gils, S. a. (2010). Comparing epileptiform behavior of mesoscale detailed models and population models of neocortex. Journal of clinical neurophysiology : official publication of the American Electroencephalographic Society, 27(6), 471–8.

Download Code here - Verdandi_jan_2013.tar_.bzip2



pNeo is a C parallel network simulator based on distributed-memory (MPI) protocol. It is derived from p-genesis by stripping all the functionalities that are not relevant to simulate the neocortex.  It was utilized for simulations up to 400 thousands neurons and 1.5 Billion synapses. In its current implementation is lacks both gap-junctions and long range connections. Its scalability has been shown to be nearly perfect for up to a few hundred processors.

Hereld, M., Stevens, R., Van Drongelen, W., & Lee, H. C. (2004). Developing a petascale neural simulation. Conference proceedings : Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference, 6, 3999–4002.

Download Code here - pNeo_jan_2013.tar_.bzip2