R. Harold Burton Foundation
Scientific Computing and Imaging (SCI) Institute, Guido Gerig
Interim Report, October 2013
IntroductionMany mental illnesses are thought to have their origins in early stages of development, encouraging increased research effort related to early neurodevelopment. Both structural and diffusion tensor magnetic resonance imaging (MRI) provide a noninvasive view of the brain in vivo and are well suited for pediatric studies as they enable safe longitudinal scans of children. Longitudinal neuroimaging studies thus provide a unique opportunity for understanding brain maturation by taking repeated scans over a time course within individuals.
Student orientationAt the beginning of his one year project, the student learned how the imaging team led by Dr. Gerig within the Scientific and Computing Institute operates, and gained many technical skills and knowledge related to computer sciences and image processing in general. More specifically, he learnt the use of open-source software libraries and programming environments being used in medical image analysis, applied such skills to process adult and pediatric brain MRIs. The student is now specifically involved in an Infant Brain Imaging Study, as part of an NIH-funded Autism Center of Excellence.
Pilot projectUsing healthy control subjects from a longitudinal autism study, the student focused on a pilot sub-project on Down Syndrome (DS), the most common genetic cause of intellectual disability, with an incidence of 1 in 800 live births. DS remains the most common genetic defect associated with intellectual and developmental disability and contributes to about 30% of all moderate-to-severe cases. In recent years, there has been an increase in the number of children with Down syndrome who are also being diagnosed as having autism or autistic spectrum disorder. While life expectancy has significantly improved for individuals with DS, parallel progress has been very limited in understanding the basic underlying neuropathophysiology. All subjects were recruited and scanned at 6 months with follow-up assessment and scans at 12 and 24 months (Figure 1). The goal of the pilot study was to demonstrate feasibility and investigate early differences in brain development in a DS population.
|Figure 1: T1-weighted (top) and T2-weighted (bottom) MRI images at 6 (left), 12 (middle) and 24 (right) months of age.||Figure 2: Left: Tissue segmentation (white matter (red); gray matter (green) and CSF (blue)). Right: lobar parcellation|
As can be seen in Figure 2, the student segmented brain tissues (identifying white matter from gray matter) and defined brain lobes (anatomical classifications being related to different brain functions) for all subjects via automated software. Taking full advantage of longitudinal information in a newly state-of-the-art methodology, the student used results of later time-points as priors (or initializations) for successful and more accurate processing of earlier brain scans. Group analysis was then performed between 5 subjects with Down Syndrome and 13 Healthy Controls (HC). Figure 3 displays development of average white and gray matter volumes over time for both groups, where significant differences can be found. Infants with DS appear to show smaller tissue volumes as early as 6 months of age than healthy controls, particularly in gray matter. By normalizing with total brain volumes (sum of white and gray matter), we found that brains are not simply scaled down but present distinct patterns among lobes and tissues (Figure 4). These early brain differences mirror findings observed in an older sample of 8-35 year olds and suggest that brain growth trajectories differ from normative development and may be evident from infancy.
ConclusionsThis small pilot study helped define the nature and characteristics of very early brain development in Down Syndrome in comparison to healthy controls using multimodal magnetic resonance imaging methods. Future work will include image analysis and processing of brain MRI images from infants participating directly in the autism neuroimaging study. This study of very early brain development in autism indeed has the potential to provide important clues relevant to early detection of autism and discovering the early changes in the brain for young children with autism.
|Figure 3: Average white matter (left) and grey matter (right) volumes over time for Down Syndrome (DS (red)) and Healthy Controls (HC (blue)) groups.|
|Figure 4: Volume differences observed in infants with DS compared to healthy controls, after brain normalization.
Left: P-value maps between controls and Down's subjects (please note the very small number of infant DS subjects: 3 at 6months, 4 at 12months, and 2 at 24months, versus 13 controls in all age groups).
Right: Percentage volume differences between DS and controls for brain volumes normalized by total brain volume (TBV), illustrating relative differences of brain growth per lobe, tissue category and across time. Blue indicates DS<controls, and orange DS>Controls in TBV-normalized scans.