Technology to predict drowsiness while driving has been developed at GIST

Drowsiness while driving is predicted by measuring EEG and cerebral hemodynamic signals

GIST Professors Jae Gwan Kim and Sung Chan Jun publish their research in Scientific Reports

Professor Jae Gwan Kim of the Department of Biomedical Science and Engineering and Professor Sung Chan Jun of the School of Electrical Engineering and Computer Science at the Gwangju Institute of Science and Technology (GIST • President Seung Hyeon Moon) have developed a technique that can predict drowsiness by simultaneously measuring brain waves and hemodynamic signals. This technology has the potential to reduce auto accidents caused by driver drowsiness and fatigue.

Previous attempts to gauge the relative fatigue of drives have included monitoring the vehicle's operation or by observing the driver's behavior. However, predicting driver drowsiness by measuring bio-signals is more effective it does not have the disadvantage of being affected by external environmental factors.

Electroencephalogram (EEG) signals are widely used as a reliable measurement of brain function due to its high temporal resolution, portability, and low costs. Functional near-infrared spectroscopy (fNIRS), which can acquire oxygen saturation and blood volume information in organs by measuring  light in the near-infrared (650 to 1,000 nm) wavelength transmitted into the body, has been used to measure cerebral hemodynamics signals for various applications, such as use in brain-computer interfaces as well as use in epilepsy seizure detection.

The research team created a monitoring system that integrated EEG and fNIRS measurements of brain waves and cerebral hemodynamics signals to predict when a driver would become drowsy when operating a motor vehicle.

From the results of their experiments, the research team discovered that driver drowsiness can be accurately predicted whenever there is a decrease of the beta EEG signal with a simultaneous increase in the oxidative hemoglobin concentration in the brain. By using the Drowsiness Detection Index (DDI), this approach can predict the first eye closures 3.6 seconds before it occurs.

Their research, entitled "Utilization of a combined EEG/NIRS system to predict driver drowsiness," was authored byThien Nguyen, Sangtae Ahn, Hyojung Jang, Sung Chan Jun, and Jae Gwan Kim, and it was published on March 7, 2017, in Scientific Reports.

Professor Jae Gwan Kim and Professor Sung Chan Jun said, "By simultaneously measuring and using other types of brain signals such as brain waves and brain hemodynamics, our research establishes the underlying technology to prevent traffic accidents caused by driver drowsiness and fatigue. We will continue to develop our research findings so that it can have real-world application by using wireless technology and by miniaturizing the overall system."

Professor Jae Gwan Kim and Professor Sung Chan Jun will be participating in the "Center for Optimal Dementia Management Technology Research Center," which is being sponsored by the Ministry of Science, ICT and Future Planning to help accelerate the development of patient-friendly dementia monitoring technology using EEG and fNIRS.

Their research was supported by Hyundai Motors, a Small Grant Exploratory Research (SGER) from the National Research Foundation, and the Brain Science Development Project from the Ministry of Science, ICT and Future Planning.