Epilepsy is a neurodegenerative disease identified by its characteristic synchronized high-frequency neuronal activity, also known as seizures, in specific parts of the brain and, even though around 50 million people around the world have to cope with the destructive effects of this neurological disorder, it is estimated that approximately 10% of people worldwide should have at least one brain seizure during their lifetime. It is known that neurons bind to each other in connections, known as synapses, that can excite or inhibit the response of the next cell down the network and this balance between excitation and inhibition may be the key not only for the encoding of information in the brain but also for the possibility of using brain cells as computing agents allowing these synthetically engineered neurons to act as digital logic gates and help to filter neuronal activity at abnormal frequency levels.
Figure 1: Synthetic logic gate connected inside a neuronal network.
In this work, we build neuronal logic gates by controlling and simulating computational models of neurons which can behave as traditional digital logic gates and, by positioning them inside a network of excitatory connected neurons concerning their respective specific connection probability, we simulate synthetic modifications to the synaptic connections between the cells composing a logic gate and evaluate the performance and accuracy of the gates by applying widely used concepts of queueing-theoretical analysis taking into account not only the type of gate that was built but also the approximation of neural activity with the manipulation of intra- and extra-cellular concentration of ions that has the potential of triggering seizure-like events in a neuronal network.
Figure 2: Schematic of the connection of the neuronal logic gates in between the N cells that forms the network with M layers. The placement of a logic gate in the network require the breakage of the natural connections between the cells.
Simulations have, visually and numerically, shown a decrease on the frequency of neural activity when placing the proposed models of neuronal logic gates in comparison with the network without any synthetic element in between the neurons to help decrease the average firing rate of the network.
Figure 3: Raster plots without any logic gate (top) in the network and with 16 gates (bottom) positioned inside the network. A clear decrease in the frequency of the neural activity can be visually assessed.
The approach that was taken in this work requires the cells used to build the logic gates to be synthetically engineered and strategically placed inside the network to achieve better results when smoothing out the effects of brain seizures. Even though the analysis was performed with the improvement of the quality of life of people affected by epilepsy, logic gates can play an important role in the enhancement of information processing by the brain.
Authors: Geoflly L. Adonias, Anastasia Yastrebova, Michael T. Barros, Yevgeni Koucheryavy, Frances Cleary and Sasitharan Balasubramaniam.
Journal: IEEE Transactions on NanoBioscience (Early Access)