Monday, June 3, 2013

Can Brain Imaging Aid in the Diagnosis of ADHD?

Brain Cerebellum Highlighted in Purple
Brain imaging advances to date are predominantly within the domain of research with limited clinical utility.

However, brain imaging techniques such as sensitive structural imaging methods, functional connectivity imaging, combined EEG and structural brain imaging and diffusion tensor imaging hold promise for clinical applications.

An obvious application would be the use of brain imaging in confirmation of the diagnosis of brain disorders.  Such methods would be particularly helpful in disorders with atypical presentations or diagnostic uncertainty.

Lim and colleagues from the United Kingdom and Singapore recently published a study examining brain MRI with voxel-based morphometry in the classification of adolescents with ADHD.

This study had the following key components in design:

  • Study subjects: 29 male adolescents (primarily ADHD medication-naive) between the ages of 10 and 18 with ADHD combined type
  • Control subjects: 29 age-matched males without a diagnosis of ADHD and with low ADHD symptom scores
  • Psychiatric control subjects: 19 age-matched adolescents with a diagnosis of autism spectrum disorder (ASD)
  • Imaging technique: whole brain MRI imaging with a GE 3T scanner using voxel-based morphometry (VBM) with application of the DARTEL algorithm
  • Statistical analysis: multivariate pattern recognition comparing ADHD, controls and ASD groups

The study identified significant patterns in ADHD, ASD and control groups:

  • ADHD adolescents showed early development of ventral brain regions of the frontal, premotor, temporal, limbic and brain stem regions
  • Control adolescents showed later brain development in regions of the dorsolateral prefrontal cortex, anterior cingulate, dorsal striatum, thalamus and inferior parietal regions
  • ASD adolescents could be discriminated from ADHD adolescents by difference in the cerebellum, anterior cingulate cortex, caudate/thalamus, temporal and parietal regions
  • In the more traditional VBM analysis, the most statistically different findings between ADHD and controls was a reduction in grey matter involving the right and left cerebellum

The authors noted in the discussion that regions identified with the most discrimination power for ADHD included key brain regions known in "mediating the higher level cognitive control, attention and timing functions that are impaired in ADHD". 

Accuracy of diagnosis is a key statistical measure of the power of a diagnostic tool.  In this study, the algorithm achieved an overall accuracy of 79.3% in classifying cases and controls.

This study will need to be replicated in independent samples by independent laboratories to confirm specific brain region findings and to confirm the accuracy estimate.  Additionally, because imaging is an expensive tool, it's utility will need to be contrasted with other less expensive diagnostic tools such as rating scales and neuropsychological testing.

Nevertheless, this study points to the approaching utility of brain imaging as a diagnostic tool in the diagnosis of ADHD and other brain disorders.  Early accurate identification can provide clinicians with the opportunity to provide early interventions that may modify the course and severity of the disorder.

So what is the answer to the question "Can Brain Imaging Aid in the Diagnosis of ADHD?"  I would  propose the answer is yes, but not yet.  However, I believe further research including advances in cost-effectiveness will soon make structural brain imaging an important diagnostic tool in ADHD.

Image of brain cerebellum is from an iPad screen shot using the app 3D Brain.

Readers with more interest in this research can access the full free text article in the citation link below.

Lim L, Marquand A, Cubillo AA, Smith AB, Chantiluke K, Simmons A, Mehta M, & Rubia K (2013). Disorder-specific predictive classification of adolescents with attention deficit hyperactivity disorder (ADHD) relative to autism using structural magnetic resonance imaging. PloS one, 8 (5) PMID: 23696841

No comments:

Post a Comment