Spatial speech perception benefits in young children with normal hearing and cochlear implants

OBJECTIVES: Several studies have demonstrated better speech perception performance in children using two rather than one cochlear implant (CI). The extent to which bilaterally implanted children benefit from binaural cues to segregate speech and noise in a spatial configuration is less clear. Although better-ear effects are expected to be similar to adults, it is unknown whether electrical stimulation allows true binaural processing of speech signals in noise. Moreover, little data are available on the binaural hearing abilities of normal-hearing children. This study aimed at (1) developing and evaluating a speech test based on numbers to determine speech reception thresholds (SRTs) fast and accurately in young children, (2) evaluating a setup for measuring benefits of speech perception in a spatial configuration in young children and determining normative values of normal-hearing children, and (3) measuring spatial speech benefits in cochlear-implanted children with good sound localization abilities. DESIGN: The speech test was conducted using the Leuven Intelligibility Number Test (LINT) data base. The test was limited to the numbers 1 to 10 spoken by one female speaker ("LittleLINT"). The LINT speech-weighted noise was used as a masker. Perception of this speech material was evaluated at fixed signal-to-noise ratios (SNRs) through monaural presentation via headphones in 34 normal-hearing children of 4 and 5 yrs of age and 20 normal-hearing adults. Subsequently, spatial speech perception benefits were measured in 50 normal-hearing children between 4 and 8 yrs of age, 15 normal-hearing adults, and eight children with bilateral CIs. An adaptive procedure was used for estimating unilateral and bilateral SRTs for different spatial configurations of speech and noise. Speech was always presented at 0 degrees azimuth (the front) and noise at the front, 90 degrees to the right, or 90 degrees to the left. RESULTS: Unilateral headphone SRTs for the LittleLINT were higher for children (-9 dB SNR) than for adults (-13 dB SNR) and were lower than those for the LINT (-10 dB SNR for adults). Slopes (12 to 14%/dB) were comparable with that of the LINT (15%/dB), suggesting equal efficiency for the limited set of numbers. Normal-hearing subjects demonstrated several benefits of two-ear listening in spatial configurations (spatial release from masking [SRM], head shadow, summation, and squelch). Only SRM was influenced by age. Implanted children clearly benefited from bilateral implantation, as shown by SRM (3 dB) and head shadow effects (4 to 6 dB) comparable with normal-hearing children, but no summation or binaural squelch was established. The first CI seemed to contribute most to spatial speech perception. CONCLUSIONS: The steep slope, the familiarity to children, and the repeatability of lists make the LittleLINT suitable for fast and accurate SRT estimation in children. Spatial speech perception benefits were observed in normal-hearing subjects from the age of 4 yrs. Cochlear-implanted children showed better-ear effects but there was no evidence of true binaural processing.status: publishe