UNAIR NEWS – Drowsiness is one of the factors causing road accidents. Traffic accident data showed that 83% of events were caused by fatigue; drowsiness; and speeding over the limit.
Therefore, Osmalia Nur Rahma ST, M.Si. with Akif Rahmatillah S.T., M.T., a lecturer of Faculty of Science and Technology UNAIR studying drowsiness with an artificial intelligence method, Extreme Learning Machine (ELM). Beside ELM, wavelets are needed as a processing method converting signals from time format to frequency format.
“With an artificial neural network Extreme Learning Machine method, this device can distinguish drowsiness and awake condition automatically. So that it can be used to detect drowsiness when driving,” said Osmalia.
Previously there had been many alarm systems developed to prevent drowsiness. Most use the camera to monitor eye movements or sensors placed on tires to prevent slippage when the driver is drowsy. However, there is no optimum system responding to the intensity of light or weather.
Based on the results of the study, brain waves recorded by Electroencephalograph (EEG) based on Brain Computer Interface (BCI) to become an alternative to overcome this. “BCI is not like an EEG in the hospital with lots of connected cables, but it is like wireless headset and connected to bluetooth so it’s easy to use, ” she said.
BCI uses the EEG principle and has been widely developed commercially at an affordable price for both games and medical applications. BCI apparently can also be used to record the electrical activity of the brain, when the individual is sleepy or awake.
When they are sleepy, she continued, alpha waves and theta waves’ intensity in the brain increase compared to when they are awake. Alpha waves are brain waves with a frequency of 8-12 Hz, while theta waves have a frequency of 4-8 Hz. “Both of these appear in normal people when they feel relaxed and tend to be drowsy,” she said.
He continued, there are differences in frequency when they are sleepy and awake. When they are sleepy, alpha and theta waves increase. However, these differences are not statistically significant, thereby reducing the accuracy of ELM testing. “Therefore, we need to add the Common Spatial Pattern method that able to increase the differences in the values of the two groups,” she said.
The results showed that ELM supported by the Common Spatial Pattern can be used to distinguish EEGs in drowsiness and wakefulness. “In order to work optimally, this device needs to be combined with the Common Spatial Pattern method,” she said.
Osmalia hoped that the system can be useful in developing alarm systems, especially a brainwave signal based, to detect sleepy motorcyclists to reduce the risk of accidents. (*)
Author: Erika Eight Novanty
Editor: Nuri Hermawan
Reference : Osmalina Nur Rahma, Akif Rahmatillah. Drowsiness Analysis Using Common Spatial Pattern and Extreme Learning Machine Based on electroencephalogram Signal. Journal of Medical Signals and Sensors, ISSN : 2228-7477