According to the CDC, 1 in every 54 children are diagnosed with some form of Autism, making it one of the most common developmental delay disorders in the world. Although Autism does not have a "cure" per se, early treatment during key developmental stages can drastically improve quality of life for those affected, including treatment starting as early as age 2. The current method to diagnose Autism Spectrum Disorder requires the child to answer between 30 and 150 questions during an interview with a trained behavior specialist, a process that can take upward of 3 hours. Using the current method, the average age of diagnosis is 5.7 years old.

Knowing that early treatment requires early diagnosis, the President and Fellows at Harvard College developed a method to accurately diagnose an Autism disorder earlier then the standard tests.  This newly patented method (U.S. Pat. No. 10687751) uses AI to analyze a short home-recorded video of a child answering 7-10 questions to provide a preliminary diagnosis, and treatment plan. The AI software compares the subject's physical responses, as well as any sort of verbal response to a library of responses from both those with and without Autism. The test is able to provide a preliminary diagnosis to children as young as 2 years old, and can be done from the comfort of home with a computer and video camera.  The patent also discloses potential use for early dementia diagnosis in the elderly.  

Stay up-to-date with the most recent AI Neurology diagnostic patent activity and related news with the Magic Number® AI Biotech/Diagnostics: Neurology Patent Forecast®.