Neuroimaging study identifies three distinct ADHD biotypes through unsupervised clustering of brain scan data, with no clinical labels used as input.
Key Takeaways
Three biotypes emerged from brain imaging alone and mapped onto clinically recognized ADHD presentations, suggesting neurobiological grounding for subtypes.
One biotype appears linked to emotional dysregulation, a symptom absent from formal DSM diagnostic criteria despite clinical relevance.
Brain imaging for subtyping remains impractical at scale: too expensive and not yet precise enough at the individual patient level.
The clustering used no clinical information, making the alignment with known presentations a notable validation signal.
Hacker News Comment Review
Commenters broadly agree the DSM is lagging behind neuroscience research, with criticism that it categorizes surface symptoms while ignoring underlying causes like executive function deficits.
The “activation energy” framing for ADHD resonated: knowing what to do but being unable to initiate is a common lived pattern that “attention deficit” fails to capture.
Skepticism about current clinical practice cuts both ways: imaging is expensive and imprecise, but existing questionnaire-based diagnosis and trial-and-error medication are also low-precision tools.
Notable Comments
@bshaughn: describes psychiatry as mapping a high-dimensional object using only its 2D shadow, with very few allowed category labels.
@sudosteph: notes balance and motor control problems are highly correlated with ADHD but also absent from DSM criteria.
