Seoul National University Hospital researchers have developed an artificial intelligence (AI)-based system that can diagnose attention deficit hyperactivity disorder (ADHD).
|Professor Kim Bung-nyun|
Diagnosis of ADHD, characterized by poor concentration, distraction, and impulsiveness, is made by clinical diagnostic procedures that require long professional training, including various developmental assessments, use of assessment scales, and diagnostic interview tools.
However, there is a lot of reliance on the parent or teacher who fills out the evaluation forms. The problem is that such a process has room for subjective judgment of parents and teachers.
In the case of parents with high anxiety, it is possible to report the seriousness of the children's problems with high severity, while parents with low anxiety may downplay the children's symptoms.
Also, even if the child needs treatment, symptoms may worsen due to the false belief of parents or caregivers who may have a wrong idea of the disorder and lead to secondary complications.
Therefore it is necessary to consider the credibility of the information carefully, and the ADHD diagnosis process requires high proficiency of the clinician and additional follow-up with neuropsychological tests.
The AI program, based on the combined brain imaging data developed by the research team led by Professor Kim Bung-nyun, acquires data from a variety of brain images, including fMRI, DTI, and aMRI in 47 ADHD patients and 47 non-ADHD personnel. The team then used a repeated-learning model to determine whether the brain belongs to an ADHD patient or an average person.
Professors Jeong Beo-seok at Korea Advanced Institute of Science and Technology and Yoo Jae-hyun at St. Mary's Hospital also participated in the study.
The team's program focused on developmental abnormalities in several critical areas of the brain as the brains of ADHD patients had distinct structural defects in the network for screening important stimuli and the frontal lobe responsible for suppressing responses.
"The brain imaging big data opened the way to distinguish between normal developing children and ADHD patients," Professor Kim said. "Various brain structures and functional images can be explained more thoroughly with the AI-based platform to explain the cause of future ADHD behavior."
Such aspects have an excellent potential for ADHD diagnosis and development of a treatment, Kim added.
Brain Imaging & Behavior published the results of the study.