In a groundbreaking study published in JAMA Network Open, researchers at Karolinska Institutet unveiled an AI model capable of predicting autism in young children with nearly 80% accuracy, promising significant advancements in early diagnosis and intervention.
Researchers at Sweden’s Karolinska Institutet have developed a cutting-edge machine learning model that can accurately predict autism in young children, marking a significant stride in early childhood health care. This groundbreaking study, published in JAMA Network Open, unveils a promising tool for health care professionals that could revolutionize the early detection and treatment of autism spectrum disorders (ASD).
“With an accuracy of almost 80% for children under the age of two, we hope that this will be a valuable tool for health care,” Kristiina Tammimies, associate professor at the Department of Women’s and Children’s Health at Karolinska Institutet and the study’s last author, said in a news release.
The research team leveraged the extensive SPARK database from the United States, containing data on approximately 30,000 individuals with and without autism spectrum disorders.
By meticulously analyzing a combination of 28 different parameters accessible before the age of two, the team developed four distinct machine-learning models. The standout model, termed “AutMedAI,” demonstrated remarkable efficacy.
“Among about 12,000 individuals, the AutMedAI model was able to identify about 80% of children with autism,” Shyam Rajagopalan, the study’s first author and an affiliated researcher at Karolinska Institutet, said in the news release.
This model identified key predictors of autism, such as the age of first smile, the first short sentence and the presence of eating difficulties, based on simple, readily available information rather than exhaustive medical tests.
Early diagnosis, the researchers stress, is crucial for implementing timely and effective interventions that can significantly enhance the developmental outcomes for children with autism.
“This can drastically change the conditions for early diagnosis and interventions, and ultimately improve the quality of life for many individuals and their families,” Rajagopalan added.
The model particularly excelled in identifying children with greater difficulties in social communication and cognitive abilities, who also exhibit more general developmental delays.
Moving forward, the research team plans to fine-tune the model and assess its performance in clinical settings, with ambitions to incorporate genetic data for enhanced specificity and accuracy in predictions.
“To ensure that the model is reliable enough to be implemented in clinical contexts, rigorous work and careful validation are required,” added Tammimies. “I want to emphasize that our goal is for the model to become a valuable tool for health care, and it is not intended to replace a clinical assessment of autism.”