The current ones neural network algorithms they achieve impressive results that help solve numerous problems. However, the electronic devices used to run these algorithms still require too much processing power. These artificial intelligence systems simply cannot compete with a real brain when it comes to processing sensory information or interactions with the environment in real time.
The neuromorphic engineering is a promising new approach that bridges the gap between artificial intelligence and the natural. An interdisciplinary research team from the University of Zurich, the Swiss Federal Institute of Technology (ETH) in Zurich and the Zurich University Hospital has used this approach to develop a chip based on neuromorphic technology that reliably and accurately recognizes complex biosignals.
The team led by Giacomo Indiveri, professor at the Institute for Neuroinformatics at the University of Zurich and at the ETH, managed to use this technology to successfully detect high frequency oscillations previously registered. These specific waves, measured by a intracranial electroencephalogram (iEEG), have been shown to be promising biomarkers for identifying brain tissue that causes epileptic seizures.
The researchers first designed an algorithm that detects high-frequency oscillations thanks to a good simulation of the brain’s natural neural network: the so-called impulse neural network.
The second step consisted in implementing the impulse neural network in a piece of hardware the size of a fingernail that receives neural signals through electrodes and that, unlike conventional computers, is extremely efficient in the use of resources. This makes calculations with very high temporal resolution possible, without relying on cloud computing or even the internet. “Our design allows us to recognize patterns of space and time in biological signals in real time,” summarizes Indiveri.
The neuromorphic chip reliably and accurately detects high-frequency oscillations in the previously recorded intracranial EEG. (Image: UZH, ETH Zurich, USZ)
The researchers now plan to use their findings to create an electronic system that reliably recognizes and monitors high-frequency oscillations in real time. When used as an additional diagnostic tool in operating rooms, the system could improve the outcome of neurosurgical interventions. (Source: NCYT from Amazings)