👩⚕️ How AI is Helping Detect Autism
According to WHO, approximately one in 100 children in the world has autism spectrum disorder (ASD). Often diagnosed relatively late, these children do not receive the support they need to develop and learn. Artificial intelligence can help solve this problem as a potential screening tool.
🇸🇪 The AutMedAI model developed by researchers from Karolinska University (Sweden) diagnoses childhood autism with nearly 80% accuracy. The model was trained using a dataset of 30,000 children of different ages, half of whom had ASD.
The AI model analyzed 28 indicators based on data from parents and doctors, such as the age of the first smile, difficulties with eating, speech problems, and other behavior patterns.
🇦🇺 Australian scientists used AI to analyze the electroretinogram — the response of the eye's retina to flashes of light — of children aged 5-16. The algorithms found that the retinas of children with RAS responded differently.
🇰🇷 Researchers in South Korea used AI to analyze retinal photographs of about 1,000 children aged 7-8 years, half of whom were also diagnosed with RAS. The AI tool correctly selected the retinal images of the neurodiverse children in nearly 100% of cases.
🇺🇸 The AI model developed by researchers at the University of Louisville in Kentucky, also detects autism in children 24 to 48 months of age at a 98.5% accuracy rate based on the brain MRIs’ analysis.
More on how AI is helping advance medicine:
AI to search for new antibiotics
#news @hiaimediaen