Assessment of Listening Effort by Eye Blinks and Head Tilt Angle Using a Glasses-Type Wearable Device.

Abstract

Objectives: Listening effort is the mental effort that increases in situations where listening is challenging. Objective indicators are needed to assess listening effort, but no established testing methods can be performed in a daily environment. We used a glasses-type wearable device (JINS MEME, JINS Inc., Tokyo, Japan) equipped with an electrooculography sensor and an acceleration/angular velocity sensor to measure the number of eye blinks and changes in head tilt angle during listening under noise and investigated its use as an objective indicator of listening effort.

Design: The study included 16 normal-hearing individuals (mean = 27.94 years, SD = 7.18 years). They wore a glasses-type wearable device and were asked to repeat a passage presented at a sound pressure level of 60 dB sound pressure level. Three conditions were performed with signal to noise ratios (SNRs) of 0, -5, and -10 dB SNR. The number of eye blinks and head tilt angle were measured during the time spent listening to the conversation (listening period) and the time spent repeating it after listening (response period). After each task, the effort and motivation required for that trial were evaluated subjectively on a scale. Friedman tests were performed on the percentage of correct words repeated as well as subjective scores for effort and motivation based on the SNR. A linear mixed model was used to evaluate the effects of SNR and interval (listening period and response period) on the number of eye blinks and head tilt angle. In addition, correlation analysis was performed on each indicator.

Results: As the SNR decreased, the correct answer rate and motivation score decreased, and the effort score increased. These changes were significantly greater at -10 dB SNR than in the other 2 conditions. The eye blink rate was significantly higher in the -5 dB SNR condition than at 0 dB SNR, and was significantly higher in the response period than in the listening period, regardless of SNR. The head tilt angle was tilted forward when the SNR decreased in the listening period and response period sections. No significant correlation was observed between the indicators.

Conclusions: The number of eye blinks increased during listening in noise, but decreased with decreased subjective motivation. The head tilt angle tilted forward when the noise load increased, indicating that the participant tilted more toward the sound source. The changes in the number of eye blinks and head tilt angle during listening in noise may be objective indicators related to listening effort that can be detected quantitatively and simply using a glasses-type wearable device.