Listening 2 Faces: Improving perception of speech in noise in children with communication disorders.

Project Summary Speech communication in noise seems effortless because of the rapid back and forth nature of the process. For children in particular, communication often occurs in the presence of noise (e.g. classrooms, playgrounds and cafeterias). Children with typical development appear to do this seamlessly, however, children with developmental disabilities such as autism spectrum disorders (ASD) and hearing loss (HL) have significant difficulty comprehending speech in noise, exacerbating existing social and communication deficits and negatively impacting academic outcomes. To counter this key deficit in speech comprehension, we will develop a prototype of an innovative therapeutic training system to increase perceptual sensitivity to speech by using videos of human faces producing the target speech in a specific noise type. The app will explicitly train children with ASD and HL on interrupted noise (which has silent intervals that allow for fragments of the target to be heard) and to attend to information from the speaking face via videos that focus on human faces/mouths. Recent evidence indicates that listening in interrupted noise can provide perceptual benefits, such as remapping the auditory environment and learning to use acoustic cues. The proposed product, Listening 2 Faces (L2F), will use adaptive training procedures that automatically increases difficulty as performance improves, and includes customizable rewards to maintain interest, as well as explicit reminders to attend to the speakerandapos;s face (and, in particular, the mouth). L2F will include an app that administers training via the childandapos;s own mobile device, and can be used by a clinician to track multiple children, or by a caregiver to track an individual childandapos;s progress.Project Narrative Children with autism spectrum disorders and hearing loss exhibit significant delays in speech perception and comprehension in noise. The proposed Listening 2 Faces system uses automated, adaptive auditory and audiovisual training delivered via mobile device administered by a clinician or caregiver, to improve speech perception in noise. The proposed system will benefit public health by providing cost-effective, evidence-based training for improving speech perception in naturalistic, noisy environments in children with developmental disabilities.