A Device to Spot Autism EarlyBy Kremen, Rachel; Technology Review,
Publication Date: September 15, 2009
Article describes a system that can automatically identify autistic children as young as 24 months using the vocalizations they make in their home environment. The system, called LENABaby, was developed at the LENA Foundation which takes its name from its principal product, the Language ENvironment Analysis System. LENABaby consists of a questionnaire regarding the child’s development, a digital audio recorder, and a shirt with a pocket on the front. The child to be diagnosed wears the shirt with the turned-on recorder inside the pocket during a normal day at home, enabling the capture of up to 16 hours of audio. The recorder is then returned to the foundation, where LENABaby software analyzes the data. Building on knowledge that autistic children vocalize differently from their nonautistic peers, the software assesses vocalizations from the child using complex algorithms analyzing factors such as the phonological composition of each sound and how sounds are clustered and paired. Using LENA’s database of previously analyzed audio, the software considers how these characteristics compare to those of typically developing children, children with delayed language development, and autistic children. The foundation claims a 91-percent diagnosis accuracy with the system, which can also be used in tracking a child's language development.
Published by: Technology Review, Inc. (Website:http://www.technologyreview.com)
Massachusetts Institute of Technology (MIT) (Web Site: http://www.mit.edu )
Link to text: http://www.technologyreview.com/biomedicine/23466/