A hardware implementation of a qEEG-based discriminant function for brain injury detection

This paper presents a feature extraction engine based on using Electroencephalogram (EEG) as a tool for Traumatic-Brain-Injury (TBI) detection. The design focuses on the development of hardware accelerator components integrated onto an FPGA platform. Utilizing a combination of four key quantitative-EEG (qEEG) features, the hardware design can perform a discriminant function (DF) based on 20 variables used for predicting TBI. Since the design is intended to operate in real-time and needs to perform intensive EEGprocessing tasks, the emphasis is on the architectural aspects and speed capabilities of the feature extraction work.