AI model detects early autism signs
Using limited information, a new machine learning model has been created to help predict autism in children aged under two.
A research team from Karolinska Institutet examined a US database, otherwise known as SPARK, which included information on approximately 30,000 individuals with and without autism. From here, they created a model called ‘AutMedAI’, which was able to identify around 80% of children, who were yet to reach 24-months old, with the disorder.
This discovery can be described as nothing short of exceptional as children who are this young often can’t undergo extensive medical assessments, however the model was able to detect autism signs by simply looking at information such as the age of an infant’s first smile, first short sentence and any indications of eating difficulties.
‘The results of the study are significant because they show that it is possible to identify individuals who are likely to have autism from relatively limited and readily available information,’ said Shyam Rajagopalan, the study’s first author.
‘This can drastically change the conditions for early diagnosis and interventions, and ultimately improve the quality of life for many individuals and their families.’
Kristiina Tammimies, associate professor at KIND and last author of the study, added: ‘With an accuracy of almost 80% for children under the age of two, we hope that this will be a valuable tool for healthcare.’
As well as highlighting the excellence of the model, Tammimies has also outlined her teams next steps.
She said: ‘To ensure that the model is reliable enough to be implemented in clinical contexts, rigorous work and careful validation are required. 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.’
The research team is now planning further improvements and validation of the model in clinical settings. Work is also underway to include genetic information in the model, which may lead to even more specific and accurate predictions.
This study was published in JAMA Network Open and funded by the Swedish Foundation for Strategic Research, Hjärnfonden and Stratneuro.
Image: Mindy Olson P
More on autism:
‘Avalanche’ of demand for autism services has surpassed NHS capacity
