Beijing, September 21
Chinese researchers have developed a neural sign evaluation system with memristor arrays, paving the way in which for high-efficiency brain-machine interfaces.
Brain-machine interfaces are promising instruments for rehabilitation medication and medical electronics.
Most typical neural sign evaluation techniques utilized in brain-computer interfaces are composed of silicon CMOS circuits.
However, with the growing variety of recording electrodes, the techniques face nice challenges when it comes to energy consumption and delays.
Researchers from Tsinghua University have developed a memristor-based neural sign evaluation system and used the system to implement the filtering and identification of epilepsy-related neural indicators, attaining an accuracy of over 93 per cent.
The energy consumption of the system is lower than one four-hundredth of that of typical neural sign evaluation techniques.
The analysis demonstrates the feasibility of utilizing memristors for high-performance neural sign evaluation in next-generation brain-machine interfaces.
According to Wu Huaqiang, memristors are a brand new sort of knowledge processing machine. Their working mechanism is just like that of synapses and neurons within the human mind. With their low energy consumption, memristors have promising prospects for future information storage and neuromorphic computing.
The system attracts upon work from researchers within the fields of electronics and medical analysis, and is the product of two years of interdisciplinary cooperation, mentioned Wu.
The analysis was revealed within the journal Nature Communications. IANS