Absolute Eye Gaze Estimation With Biosensors in Hearing Aids
Level 4 - case-series / case-control
Small observational laboratory experiment and model evaluation without a control group.
OpenAlex W2993032014 · doi:10.3389/fnins.2019.01294
What was done
Eleven hearing-impaired participants performed a visual tracking task mimicking a multi-talker scenario with speech targets positioned at -30°, 0°, and +30° azimuth. Participants tracked LED-indicated targets under two conditions: head fixed in a chinrest and head free to move. In-ear electrooculography (EarEOG) biosensors and ear-level motion sensors were recorded, and a mathematical model was developed to estimate relative and absolute eye gaze by accounting for saccades, fixations, head movement, and electrode-skin half-cell drift.
What was found
The relative eye-gaze model explained 90.5% of the EarEOG variance in the head-fixed condition and 82.6% in the head-free condition. Absolute eye-gaze estimation was reliable when the head was fixed; however, during the head-free condition, hardware limitations caused excessive variance that prevented reliable identification of the attended target.
Why it matters
This study shows the technical feasibility of capturing eye movements via in-ear biosensors to potentially steer hearing aid beamformers, while identifying key motion-sensor hardware hurdles that must be resolved for practical use.
Limits
The sample size was small (n = 11), and the primary objective of reliably estimating absolute gaze during unconstrained head movement failed due to hardware issues. Testing was limited to three discrete target angles in a controlled laboratory setting rather than continuous, real-world conversation.
Cited by
- supports A sensor placed in the ear concha can measure heart rate, pulse, blood oxygen levels, electrooculogram (EOG) eye movement signatures, and EEG signatures related to attention.