Artificial Intelligence and detection and management of Autism

Artificial Intelligence (AI) has shown tremendous promise in healthcare. Already care providers are using AI to detect cancer and improve care outcomes. Now AI has another use case, detection and management of Autism Spectrum Disorder (ASD).

Researchers could predict with precision which risk babies ended up autism by using AI  to decode brain scans. Detecting autism has been hard, in part since the disease is diagnosed based on traits, including a difficulty react to their name or to make eye contact. The classical signs are not present in kids that young, especially babies younger than one year old. The scans show areas that are colored that represent areas with expansion of surface region. Researchers believe that signs develop in the uterus itself, though doctors who monitor risk children with ASD do not intervene until clinical signs appear.

In accord with the Center for Disease Control (CDC) in the US , roughly one in 68 children have ASD. Babies with a sibling possess an one in five chance of ASD. In 24 months, 15 of the 106 infants were diagnosed with ASD. In the MRI outcomes, researchers noticed that their mind volume grew from that in 12 to 24 month period compared to people without ASD. In this time, behavioural signs of autism started to surface as well. Detectable changes started happening even earlier. In those infants, between the six to twelve months, researchers found growth of the cortex, which are the folds of the brain’s surface region.

While researchers have known because the 1990 that kids with autism may have bigger brains, suggesting that it can be a telltale sign of that the disorder. The timing and direct correlation of that the enhancement to ASD behavioural symptoms haven’t been confirmed. The AI evaluation of the MRI scans in six and twelve months allowed the researchers to predict, with 81% precision, which high risk babies would develop autism by age two. Team members in that the study, led by medical researchers at the University of North Carolina at Chapel Hill and such as 10 other associations, credited their precision to a customized algorithm which was applied to the data in identifying kids more than likely to meet criteria for autism by age 2. Joseph Piven, that the study’s senior author, said that the results open a brand new window into the identification of autism.

This is indeed great news for the detection and treatment of Autism. As seen in the movie ‘Tare Zameen Par’ the identification of Autism can be a challenge and often it remains under diagnosed leading to difficult consequences to the child and its parents. If AI can solve this problem than doctors can counsel parents, teachers and school administrators alike on the approach to be taken with these children.

It is believed one in 250 children have ASD. Interestingly, the government of India only recognized the disorder in 2001, till 1980s, there were reports that Autism didn’t exist in India. So we have come a long way and with AI we have the opportunity to focus on management and helping children with ASD live up to their potential.

(Photo by Dmitry Ratushny on Unsplash)

Dr. Vikram Venkateswaran

Management Thinker, Marketer, Healthcare Professional Communicator and Ideation exponent

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.