Chip can decode brain waves

A research group from the University of Zurich (UZH), the Swiss Federal Institute of Technology (ETH) and the University Hospital Zurich has developed a chip that can decode brain waves. To do so, it first designed an algorithm that simulates the natural neuron network in the brain. This spiking neural network (SNN) was then implemented in hardware the size of a fingernail, which can receive neuron signals via electrodes.

The researchers’ algorithm is able to recognize previously recorded high-frequency oscillations (HFO) according to a press release from the UZH. These are specific waves in an intracranial electroencephalogram (iEEG) that are promising biomarkers for the brain tissue that causes epileptic seizures.

According to the information provided by the UZH, the chip is particularly energy-efficient, unlike conventional electronic devices running algorithms. It enables calculations with very high time resolution and does not need the internet or cloud computing to do so.

The chip from Zurich could on the one hand be used for real-time sensing and monitoring of HFOs in operating theaters. The UZH states that it could improve the outcome of neurosurgical procedures.

On the other hand, the chip could also serve as an epilepsy monitoring device. Johannes Sarnthein, neurophysiologist at the University Hospital of Zurich, explains in the press release: “A portable or implantable chip such as this could identify periods with a higher or lower rate of incidence of seizures, which would enable us to deliver personalized medicine.”  There are also plans to integrate low-energy wireless data communication into the chip. This would allow data to be recorded on a cell phone, for example.