Automated Vocal Analysis of Children With Hearing Loss and Their Typical and Atypical Peers

Abstract

Objectives This study investigated automatic assessment of vocal development in children with hearing loss as compared with children who are typically developing, have language delays, and autism spectrum disorder. Statistical models are examined for performance in a classification model and to predict age within the four groups of children. Design The vocal analysis system analyzed over 1900 whole-day, naturalistic acoustic recordings from 273 toddlers and preschoolers comprising children who were typically developing, hard of hearing, language delayed, or autistic. Results Samples from children who were hard-of-hearing patterned more similarly to those of typically-developing children than to the language-delayed or autistic samples. The statistical models were able to classify children from the four groups examined and estimate developmental age based on automated vocal analysis. Conclusions This work shows a broad similarity between children with hearing loss and typically developing children, although children with hearing loss show some delay in their production of speech. Automatic acoustic analysis can now be used to quantitatively compare vocal development in children with and without speech-related disorders. The work may serve to better distinguish among various developmental disorders and ultimately contribute to improved intervention.

Multiple Coordination Patterns in Infant and Adult Vocalizations

“…The study described here would be impractical to conduct without automatic labeling of event onsets. Many of the same limitations also apply to a number of studies that have also used the LENA system to study language development (e.g., Ambrose, VanDam, & Moeller, ; Caskey & Vohr, ; Greenwood et al., ; Johnson, Caskey, Rand, Tucker, & Vohr, ; Oller et al., ; VanDam et al., ; Warlaumont et al., , ; Warren et al., ); future studies and technological advances will be necessary to overcome these limitations.…”

HomeBank: An Online Repository of Daylong Child-Centered Audio Recordings

“…Research utilizing automated analysis of daylong audio recordings using the LENA system is advancing our understanding of autism spectrum disorders, [19][20][21][22] childhood hearing loss, 23,24 the consequences of premature birth, 25,26 the role of television viewing in development, [27][28][29][30] and more. These studies using the LENA system have utilized thousands of hours of audio data.…”

“…43 Some research groups have further developed custom software for audio extraction and playback, and human listener judgment collection, acoustic analysis, database management, and statistical analysis. 23,27,47 A major advantage to this type of specialized playback software is that it enables much more efficient human coding of the data. VanDam and Silbert reported on human judgments of over 90,000 exemplars of speech segments labeled by the automatic methods of the LENA system, with accuracy/validity compared for different label types.…”

A modular, extensible approach to massive ecologically valid behavioral data

“…Finally, in another study we compared four groups: typically developing, autistic, language delayed, and hard-ofhearing children (VanDam et al, 2015). Here we analyzed 1,913 daylong recordings from 273 children in four groups of known diagnoses: typically developing, autism spectrum disorder, language delayed, and hard of hearing.…”