News travels fast on Facebook. You post a YouTube video, and by the next day, your cousin’s roommate’s friend’s ex-boyfriend in Australia is watching it.
Facebook is a prime example of what is known as a small-world network. Each node (in this case, a person) is only connected to a few others within the network (their “friends”), but because the way that interconnected nodes cluster together, no node is more than a few steps away from any other. It’s a virtual demonstration of the classic six degrees of separation paradigm.
Now, scientists are investigating whether neural networks use similar strategies to efficiently transmit complex information. For example, in a recent study published in the Royal Society of Chemistry journal Integrative Biology, an international team of researchers examined how the nanostructure of silicon surfaces affected the way neural networks formed on it.
The team grew neuroblastoma brain cancer cells on two different silicon substrates: one smooth, the other etched with nanoscale pores. (Neuroblastoma cells display many of the same properties as ordinary neurons but are easier to grow in culture.) The cells grew much more quickly on the etched silicon than on the smooth surface. Furthermore, the porous, etched silicon induced the cells to form a more clustered network with a small-world topology. It appeared that the nanoscale-level constraints induced the cells to form more efficient network structures.
Porous silicon has shown promise in biomedical applications. These results suggest that biomedical engineers could influence the way neural networks form on silicon by modifying its surface. And while it has not been experimentally demonstrated, the researchers suggest that similar nanoscale cues within the brain could influence the formation of neural networks in the human brain and guide them towards more efficient configurations.
The full paper by Marinaro et al is free* to access until 9th March 2015. Download now by clicking the link below:
* Access is free through a registered RSC account – click here to register