A study of more than 7800 children has found that those with attention deficit hyperactivity disorder (ADHD) can have a smaller outer layer of the brain and specific neural connections not seen in those without the condition. The findings suggest that brain imaging could help diagnose ADHD.
Brain features of ADHD have previously been identified in small studies involving participants that aren鈥檛 very representative of the wider population.
Now, at the Yale School of Medicine and her colleagues have confirmed that certain brain features involved in impulse control, short-term memory and concentration located in the prefrontal cortex are linked to ADHD. They did this by analysing MRI brain imaging data from more than 7800 children aged 9 or 10 who are representative of the general US population.
Advertisement
The researchers trained a machine-learning artificial intelligence to form links between brain features and whether someone had ADHD. This involved giving it the brain imaging data of around 80 per cent of the children, including numerical data such as the size, surface area and neural connectivity between different brain regions, along with their ADHD status.
The team found that when the trained AI looked at data from the remaining 20 per cent of participants, it was able to accurately identify 60 per cent of children with ADHD and 56 per cent of children who didn鈥檛 have ADHD. The results will be presented at the annual meeting of the Radiological Society of North America on 27 November in Chicago.
鈥淲e show that certain brain features linked to ADHD are distinct enough for an AI to make predictions for who has the condition,鈥� says Lin.
鈥淭hese approaches are a great step forward in understanding the brain features underlying the disorder,鈥� says at Deakin University in Melbourne, Australia. 鈥淚t is great having such a large sample, particularly with a high number of females with ADHD, who tend to be underrepresented in ADHD research.鈥�
However, the predictive ability of the AI in this study is little better than the toss of a coin, which is quite common when using AI to diagnose complex neurological conditions, says Silk. It is also unclear whether the findings apply to other ages beyond those aged 9 and 10, he says.
Still, the better evidence of brain signatures linked to ADHD might help us diagnose the condition, says Lin.聽 鈥淐urrently, ADHD diagnosis is highly subjective, relying on self-reported symptoms through surveys. Adding brain imaging to this process could help provide a more objective assessment,鈥� she says.
Sign up to our free Health Check newsletter that gives you the health, diet and fitness news you can trust, every Saturday
