TU/e is working on a chip for eeg wearable that patients can use at home

Researchers at Eindhoven University of Technology are working on an economical EEG chip that epilepsy and Parkinson’s patients can use at home to register brain activity. The chip is to be used in a portable system that can analyze EEG signals.

A paper from the researchers reveals that the plan is to use the ARM Cortex-M0 processor design. The Cortex M0 will be integrated as a general-purpose CPU in a so-called BrainWave processor. The M0 is ARM’s smallest CPU design and must be adapted for the eeg system. The processor will be produced on a 28nm fd-soi cmos process from STMicroelectronics. The resulting chip would be four times more energy efficient than current EEG measurement systems, according to the researchers.

The goal of the project, called BrainWave, is to develop a wearable system that allows patients to make continuous EEG scans and record brain activity seven days a week, 24 hours outside the hospital. Acute symptoms, such as epileptic seizures, could thus be identified more quickly, after which action could be taken with, for example, medication.

The system would also be able to recognize freezing episodes in Parkinson’s patients. Patients feel as if their feet keep gripping the ground and they can no longer take a step. That could result in a fall. The BrainWave could sound an alarm when such an episode occurs.

Recognizing patterns in EEG signals that indicate, for example, an epileptic seizure is complicated, according to the researchers. Specialist staff often have to look at the EEG for a long time, which makes the examination expensive and patients have to travel to a hospital or specialist center for this. The EEG chip that the researchers are developing could analyze the signals itself and thus also be used for long-term EEG monitoring in hospital intensive care units.

The proposed BrainWave device should have an accuracy of greater than ninety percent with an error of less than one percent in determining seizures and freezing episodes. The research receives financial support from technology foundation STW.