Abstract
This poster describes a completed NIH Phase I study, a prototype of the Human Action Recognition Engine (HARE) utilizing the three-dimensional body mapping technology of the X-box Kinect to automate the extraction of infant postural and motor data of the infant during floor play with the parent in the natural environment of the home. This demonstrates the feasibility of an automated developmental risk screener to achieve state-of-the-art accuracy in the assessment of infant motor development.
Primary Author and Speaker: Teresa Fair-Field
Contributing Authors: Bharath Modayur
The emergence of motor control is critical to an infant’s development due to its demonstrated effect on social relationships, cognition and memory, and the emergence of functional communication and language. These building blocks may become compromised when motor development is disrupted; associated with generalized and specific developmental delays, speech/language impairment, cerebral palsy, autism spectrum disorder, developmental coordination disorder, and intellectual disability (Kim et al., 2016).
Early intervention targeting motor behavior can also impact growth and development in other domains. Children who receive developmental screening are more likely to be identified with delays and receive early intervention services compared to those that only receive age-appropriate milestone checks as part of pediatrician-led well-child office visits (Flick & Zachry, 2018). Complicating issues of access, a comprehensive motor screening is labor intensive and requires specialized training in administration and assessment. The prevalence of wait-and-see approaches, appointment access, issues of reimbursement, and a shortage of trained personnel can lead to missed opportunities for early intervention which contribute negatively in lifelong outcomes for infants at risk for developmental and behavioral health disorders.
This poster describes the completed NIH Phase I study, a prototype of the Human Action Recognition Engine (HARE) utilizing the 3-dimensional body mapping technology of the X-box Kinect to automate the extraction of infant postural and motor data of the infant during floor play with the parent in the natural environment of the home. This technical engine, which leverages recent advances in machine learning, demonstrates the feasibility of an automated developmental risk screener to achieve state of the art accuracy in the assessment of infant motor development.
Importantly, the poster focuses on clinical considerations in the areas of usability, accessibility, and justice. Following extensive conversations with clinical experts, the Launch Bottle team is refitting the existing technology to be a widespread deployable, inexpensive, developmental screening tool to be available as a smartphone app. This maneuver brings the power of comprehensive motor screening, usually administered only to at-risk children and utilizing specialized resources, to the underserved and economically-fragile communities, many of which, despite economic hardship, rely upon a smartphone for communication and community access. By converting to a smartphone platform, the widely-available tool will identify children with motor data falling outside the range of typical norms who may benefit from further clinical follow-up, then directing parents to community and professional resources for a comprehensive assessment. Providing education and activities to encourage motor growth at all ages and stages is an essential component of the app to increase the value of such a tool beyond that of motor screening to engage parents in the holistic development of their child.
The tool retains the family-centered focus of the platform—an infant participating in parent-child floor play in the natural environment of the home, a data collection method supported in the literature for developmental monitoring of pre-term infants (Spittle et al., 2016). This poster will articulate next steps in the research and development of the technology, as well as the logical fit of occupational therapists to participate in the technical space, to articulate our distinct value in clinical assessment to establish the validity of the database, and to ensure that the tool meets the occupational needs of parents with an inclusive and equitable lens.
Flick, J. & Zachry, A. H. (2018, April). Occupational therapy in a pediatric primary care setting. Presented at AOTA National Conference & Expo, Salt Lake City, Utah.
Kim, H., Carlson, A. G., Curby, T. W., & Winsler, A. (2016). Relations among motor, social, and cognitive skills in pre-kindergarten children with developmental disabilities. Research in Developmental Disabilities, 53-54, 43-60. doi:10.1016/j.ridd.2016.01.016
Spittle, A. J., Olsen, J., Kwong, A., Doyle, L. W., Marschik, P. B., Einspieler, C., & Chong, J. L. Y. (2016). The Baby Moves prospective cohort study protocol: using a smartphone application with the General Movements Assessment to predict neurodevelopmental outcomes at age 2 years for extremely preterm or extremely low birthweight infants. British Medical Journal Open, 6, e013446. doi:10.1136/bmjopen-2016-013446