A $300,000 grant from the Louisiana Department of Health will help University of New Orleans researchers to bridge the gap in understanding between caregivers and autistic children by using artificial intelligence-powered insights.
Researchers will leverage state-of-the-art AI technologies to understand, monitor and predict behavioral patterns and outcomes in autistic children between the ages of 2-5.
The innovative project combines 色色研究所鈥檚 expertise in both computer science and psychology. The principal investigator on the grant is Shreya Banerjee, assistant professor of computer science. The co-principal investigators are Md Tamjidul Hoque, associate professor of computer science, and Tracey Knaus, associate professor of psychology.
The researchers will develop techniques to empower caregivers鈥攊ncluding parents and educators鈥攖o better understand the reasons behind difficult behavior in children with Autism Spectrum Disorder. In collaboration with Applied Behavioral Analysis therapists, the researchers will utilize AI-powered tools that can analyze and interpret the speech and behavior of individuals with autism. The project is geared toward translating this analysis into actionable insights for caregivers and healthcare professionals, thereby helping to reduce the behavioral issues in children stemming from the frustrations of being unable to communicate their needs, according to Banerjee.
鈥淏etter management and knowledge about the children鈥檚 misunderstood cues, frustrations and underlying triggers also reduce the burnout of the parents and caregivers,鈥 Banerjee said. 鈥淪uch assistance in autism support increases the overall value of the provided care.鈥
For evaluation, researchers will use evidence-based techniques in psychology to test improvement in the children鈥檚 behavior, and language and communication functions. The approach encompasses a wide range of tools and technologies in AI, including machine learning, deep learning algorithms, natural language processing techniques, computer vision, and vision and sensor-based monitoring.
