Machine learning helps experts identify ADHD biomarkers in MRI scans

Machine Learning


A total of 51 patients diagnosed with ADHD and 60 neurotypical control patients were included.

Model 1 had an area under the curve (AUC) of 0.67, while model 3, which utilized both time point and yearly relative changes, achieved a slightly better AUC of 0.73. At both time points, we observed several white matter features that appear to be common in ADHD patients.

Alterations in the superior longitudinal fasciculus, frontal oblique tract, stria terminalis, inferior fronto-occipital tract, thalamic and striatal tracts, and other tracts involving sensorimotor regions were also indicative of ADHD. Similarly, higher rates of change over time in generalized fractional anisotropy (GFA) in these regions correlated with improvements in visual attention, short-term memory, and spatial working memory.

The authors noted that the findings related to accelerated microstructural development, which is more common among ADHD patients, are consistent with previous neuroimaging studies on this subject.

“White matter microstructural properties that indicate deviations from typical developmental patterns and the rate of developmental changes may serve as important biomarkers for ADHD,” the research group wrote.

A summary of the study is available here.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *