Biomarkers in the Brain May Predict ADHD Diagnosis With 99% Accuracy
Specific communication among different brain regions, known as brain connectivity, could potentially serve as a biomarker for attention-deficit hyperactivity disorder (ADHD), according to a study published in Frontiers in Physiology.
Specific communication among different brain regions, known as brain connectivity, could potentially serve as a biomarker for attention-deficit hyperactivity disorder (ADHD), according to a study published in Frontiers in Physiology. The investigators used machine-learning classifiers to identify adults who had received a childhood diagnosis of ADHD with 99% accuracy.
According to the researchers, these findings have implications not only for increasing the ease with which ADHD is diagnosed, but also for assisting clinicians in properly targeting treatments to address each patient’s specific needs.
“Because certain pharmaceuticals react with certain pathways, understanding the different types of ADHD can help inform decisions about one medication versus others,” said Chris McNorgan, PhD, in a press release.
Identifying and diagnosing ADHD is often difficult, despite being the most commonly diagnosed psychological disorder among school-aged children. Diagnostics are complicated by multiple subtypes of the disorder and a clinical diagnosis of ADHD may change when the same patient returns for follow-up evaluations.
“A patient may be exhibiting behavioral symptoms consistent with ADHD one day, but even days later, might not present those symptoms, or to the same degree,” McNorgan said in the release. “It could just be the difference between a good day and a bad day. But the brain connectivity signature of ADHD appears to be more stable. We don't see the diagnostic flip-flop.”
The investigators used archival fMRI data gathered from 80 adult participants who had been diagnosed with ADHD as children. The researchers then applied machine learning classifiers to 4 snapshots of activity, taken while the participants were performing a task designed to test the subject's ability to inhibit an automatic response. The collective analysis approached 99% diagnostic accuracy, with focused analysis of individual runs reaching 91%.
“It's by far the highest accuracy rate I've seen reported anywhere—it is leagues beyond anything that has come before it, and well beyond anything that has been achieved with a behavioral assessment,” McNorgan said in the release. “Many factors likely contributed towards our superior classification performance.”
This unique accuracy may be attributable to the application of deep learning networks, which are far more capable of detecting conditional relationships than direct linear classification, according to the researchers. Because the current study was designed to predict ADHD based on the patterns of communication between groups of brain areas, and these connections often require the consideration of multiple factors as opposed to linear correlation, the use of deep learning classifiers may have strongly contributed to the success of this diagnostic model, according to the authors.
Detecting ADHD with near perfect accuracy [news release]. EurekAlert; January 27, 2021. Accessed July 13, 2021. https://www.eurekalert.org/pub_releases/2021-01/uab-daw012721.php