Previous binaural experience supports compensatory strategies in hearing-impaired children's auditory horizontal localization

Abstract

This study investigates auditory localization in children with a diagnosis of hearing impairment rehabilitated with bilateral cochlear implants or hearing aids. Localization accuracy in the anterior horizontal field and its distribution along the angular position of the source were analyzed. Participants performed a localization task in a virtual environment where they could move their heads freely and were asked to point to an invisible sound source. The source was rendered using a loudspeaker set arranged as a semi-circular array in the horizontal plane. The participants' head positions were tracked while their hands pointed to the auditory target; the preferred listening position and the onset of active strategies involving head movement were extracted. A significant correlation was found between age and localization accuracy and age and head movement in children with bilateral hearing aids. Investigating conditions where no, one, or both hearing devices were turned off, it was found that asymmetrical hearing caused the largest errors. Under this specific condition, head movement was used erratically by children with bilateral cochlear implants who focused on postures maximizing sound intensity at the more sensitive ear. Conversely, those with a consolidated binaural hearing experience could use dynamic cues even if one hearing aid was turned off. This finding may have implications for the clinical evaluation and rehabilitation of individuals with hearing impairments.

Fig 1. The 3D scene with the beam pointing to a guessed sound source position.

Fig 2. Medians across conditions of signed and unsigned errors.

Fig 3. Medians across conditions of head rotation and head distance.

Fig 4. Signed errors, unsigned errors, head rotations, and distances of CI and HA listeners in the ON condition across positions.

The data are normalized. Filled circles represent the HA ON medians, empty circles represent the HA ON minima and maxima. Squares represent the CI ON medians, and diamonds represent the CI ON minima and maxima.

Fig 5. Signed errors, unsigned errors, head rotations, and distances of CI listeners in every condition across positions.

The data are normalized. Circles represent the medians, triangles pointing down represent maxima, and triangles pointing up represent minima. Squares represent the CI ON medians, and diamonds represent the CI ON minima and maxima. Pluses represent the CI L medians, and xs represent the CI L minima and maxima. Hexagons represent the CI R medians. When rotated, they represent the CI R minima and maxima.

Fig 6. Signed errors, unsigned errors, head rotations, and distances of HA listeners in every condition across positions.

The data are normalized. Filled circles represent the HA ON medians, and empty circles represent the HA ON minima and maxima. Pluses represent the HA L medians, and xs represent the HA L minima and maxima. Hexagons represent the HA R medians. When rotated, they represent the HA R minima and maxima. Triangles pointing up represent the HA NO medians, and triangles pointing up represent the HA NO minima and maxima.

Fig 7. Heat maps reporting the correlations occurring in each condition of the CI listeners.

Each asterisk indicates the statistical significance of the correlation. UE stands for unsigned error, HD for head distance, SE for signed error, and HR for head rotation.

Fig 8. Heat maps reporting the correlations occurring in each condition of the HA listeners.

Each asterisk indicates the statistical significance of the correlation. UE stands for unsigned error, HD for head distance, SE for signed error, and HR for head rotation.