In-Ear System for Monitoring Lower Jaw Motion: A Novel Approach to Speech Intention Detection in Laryngectomy Patients.

OBJECTIVE

Restoring speech following laryngectomy presents a significant clinical challenge due to the reliance of existing artificial voice systems on manual operation. This limitation reduces their practicality and social acceptability, emphasizing the critical need for hands-free solutions. In this proof-of-concept study, we investigate an untapped biomarker: deformation of the outer ear canal induced by lower jaw motion, as a discreet and reliable indicator of speech intention.

METHODS

We propose a novel in-ear device designed to monitor these deformations and enable intuitive, hands-free control of artificial voice systems. The prototype device is equipped with four infrared proximity sensors, housed within a 3D-printed enclosure. Testing protocol assessed the device's ability to detect fundamental jaw movements (protrusion, retraction, depression/elevation, and lateral shifting), integrated movements (chewing and coughing), and speech patterns (isolated vowels, single words, and complete sentences).

RESULTS

Device performance was validated through comparison with surface electromyography and audio recordings, confirming its accuracy in detecting both activity onset and termination across all tested movements and speech patterns. Reproducibility of recorded signals was established across independent trials, with the device removed and repositioned within the ear canal between trials.

CONCLUSION

We demonstrated that outer ear canal deformation reliably captures mandibular movements, including those associated with speech, establishing its potential as a discreet and reproducible biomarker for intuitive, hands-free control of artificial voice systems.

SIGNIFICANCE: By offering a non-invasive, socially acceptable solution, this approach holds significant promise for enhancing the quality of life for laryngectomy patients and facilitating their reintegration into social and professional environments.