Neuromorphic Chip: Artificial Neurons Identify Biosignals in Real Time
Scientists from Zurich have created a portable, energy-efficient gadget made from synthetic neurons that can decode brainwaves. The chip uses information recorded from the brainwaves of epilepsy patients to determine which regions of the brain cause epileptic seizures. This opens brand-new perspectives for treatment.
Current neural network algorithms produce excellent results that help address an extraordinary variety of issues. Nonetheless, the electronic devices utilized to run these algorithms still need way too much processing power. This artificial intelligence (AI) system can not compete with an actual brain when it concerns processing sensory information or communications with the environment in real-time.
Neuromorphic chip discovers high-frequency oscillations
Neuromorphic engineering is a promising brand-new approach that bridges the gap between artificial and natural intelligence. An interdisciplinary study team at the College of Zurich, the ETH Zurich, and the university hospital Zurich used this method to develop a neuromorphic chip based on neuromorphic technology that reliably and accurately recognizes complex biosignals. Scientists used this technology to find previously recorded high-frequency oscillations (HFOs) successfully. These particular waves, determined using an intracranial electroencephalogram (iEEG), have been verified to be appealing biomarkers for determining the brain tissue that creates epileptic seizures.
Complicated, compact, and power-efficient
The researchers first made a formula that identifies HFOs by imitating the brain’s natural neural network: a tiny supposed spiking neural network (SNN). The second step involved applying the SNN in a fingernail-sized item of hardware that gets neural signals through electrodes and, unlike conventional computers, is incredibly energy efficient. This makes calculations with a high temporal resolution feasible without relying upon the internet or cloud computing. “Our design allows us to identify spatiotemporal patterns in biological signals in real-time,” claims Giacomo Indiveri, professor at the Institute for Neuroinformatics of UZH and ETH Zurich.
Gauging HFOs in operating theaters and beyond healthcare facilities
The scientists are now preparing to use their findings to create an electronic system that reliably identifies and observes HFOs in real-time. When utilized as an extra diagnostic tool in operating theaters, the system might boost the results of neurosurgical interventions.
Nevertheless, this is not the only field where HFO recognition can play an essential part. The team’s long-term goal is to develop a device for monitoring epilepsy that could be utilized outside the hospital, which would certainly make it feasible to evaluate signals from a large number of electrodes over several weeks or months. “We wish to incorporate low-energy, wireless data communications in the design– to connect it to a mobile phone, for instance,” states Indiveri. Johannes Sarnthein, a neurophysiologist at UniversityHospital Zurich, specifies: “A mobile or implantable chip such as this might identify periods with a greater or reduced percentage of incidence of seizures, which would allow us to supply tailored medicine.” This study on epilepsy is being carried out at the Zurich Facility of Epileptology and Epilepsy Surgery, which is run as part of a collaboration between UniversityHospital Zurich, the Swiss Epilepsy Center, and the College Kid’s Healthcare facility Zurich.
Originally published on Scitechdaily.com. Read the original article.
Reference: “An electronic neuromorphic system for real-time detection of high frequency oscillations (HFO) in intracranial EEG” by Mohammadali Sharifshazileh, Karla Burelo, Johannes Sarnthein and Giacomo Indiveri, 25 May 2021, Nature Communications.
DOI: 10.1038/s41467-021-23342-2
