A multimodal drowsiness monitoring ear-module system with closed-loop real-time alarm

A closed-loop electroencephalogram (EEG) - near-infrared spectroscopy (NIRS) ear-module system for wearable drowsiness system is proposed. It is implemented on the earpiece and ear hook. Thanks to the ear-module SoC in the earpiece, the proposed system monitors 1-ch EEG and 1-ch NIRS simultaneously with small form-factor. Since the NIRS signal can be easily distorted because of ambient light and extra-cerebral artifact, dedicated noise reduction scheme is proposed for ear-NIRS. After the measured physiology signal is transferred to MCU in the ear hook, it is preprocessed and classified. With the classified result, alarm makes user feel awake. The proposed system is fully implemented and verified by Oxford Sleep Resistance Test. It shows the improved classification accuracy up to 20%. As a result, the proposed system is expected to be effectively utilized for closed-loop drowsiness monitoring with the multimodal signal monitoring and auditory alarm as well as convenient and non-intrusive form factor. It has only 40g weight and operates at least 5 hours with 3.6V 200mAh rechargeable battery.