Researchers at Karolinska Institutet have developed a machine learning model, AutMedAI, which can predict autism in children under two with nearly 80% accuracy. The model uses a set of 28 parameters easily gathered before the age of 24 months, including information about children that can be obtained without extensive assessments and medical tests.
The study, published in JAMA Network Open, highlights the model’s ability to identify key predictors like the age of first smile and presence of eating difficulties. This breakthrough promises to facilitate early interventions, enhancing the quality of life for affected individuals and their families.
According to the researchers, the results show that it is possible to identify individuals who are likely to have autism from relatively limited and readily available information. The AutMedAI model was able to identify about 80% of children with autism, and specific combinations with other parameters, such as age of first smile, first short sentence, and presence of eating difficulties, were strong predictors of autism.
The study’s authors emphasize the importance of early diagnosis in implementing effective interventions that can help children with autism develop optimally. They also highlight the potential benefits of using AI models like AutMedAI to aid in detecting early markers of autism, which could lead to earlier intervention and improved outcomes for affected individuals.
Overall, this breakthrough has significant implications for improving our understanding and management of autism, as well as enhancing the lives of individuals with autism and their families.
Source: https://scitechdaily.com/breakthrough-ai-predicts-early-autism-with-surprising-accuracy/