Researchers Leverage YouTube To DiagnoseAutism In Minutes
Lindsay ONeal April 27, 2012
Researchers at at Harvard Medical School have found a way to reduce the time ittakes to accurately detect autism in young children from hours to minutes usingparent-submitted videos.A recent study revealed that one in 88 children suffer from Autism, a disorder of neural development typically characterized by impairedsocial interaction and
communication and restricted and repetitive behavior. Signs typically appearduring the first three years of life, but the current diagnosis process is lengthy,complex and frustrating.With hours of data, physician observations, and parental acumen all needed for anaccurate diagnosis, the average time between initial evaluation and diagnosis for achild living in a large metropolitan area is over one year. Not only is the delayfrustrating for families, but it prevents many children from receiving medicalattention until they are beyond developmental time periods when targetedbehavioral therapy may have had maximal impact.The growing number of incidents and the need for simpler more accurate testing iswhat led Dennis Wall, associate professor of pathology and director of theComputational Biology Initiative at the Center for Biomedical Informatics, to work on developing algorithms and deployment mechanisms to detect autism morerapidly and effectively.Wall and his team found that the current behavioral test ADI-R could be paireddown from 93 questions to a seven-point parent questionnaire and a home videoclip.
Wall’s new method could
reduce the time for autism diagnosis by nearly 95percent, from hours to minutes. The testing also could help to reach significantly