27Diffusion MRI allows non-invasive assessment of white matter maturation in typical development and 28 of white matter damage due to brain injury or pathology. Reliably attributing diffusion metrics to 29 specific white matter pathways either requires use of lengthy acquisition protocols with numerous 30 diffusion directions, which may be problematic in certain cohorts (e.g. children or adults with mild 31 cognitive impairment), or probabilistic white matter atlases, which allow delineation of white matter 32 tracts without the need to perform tractography, thus eliminating the need for the extensive scans 33 required for modern tractography algorithms. However, given the known age-dependency of 34 developmental change in white matter it may not be optimal to use an adult template when assessing 35 data acquired from children.
36This study develops an age-specific probabilistic white matter atlas for delineation of 12 major white 37 matter tracts in children aged 6-8 years. By comparing to fibre tracking in individuals, we 38 demonstrate that this age-specific atlas gives better overall performance than simply registering to the 39 Johns Hopkins University (JHU) adult white matter template in both data acquired from a single 40 cohort on a single scanner (age-specific r = 0.72; JHU r = 0.54) and from a cohort taken from the 41 ABIDE dataset (age-specific r = 0.75; JHU r = 0.72). Accuracy was assessed by comparing estimates 42 of tract-level diffusion metrics, using the age-specific and adult templates, to results of subject-43 specific tracing. To our knowledge, this is the first publicly available probabilistic atlas of white 44 matter tracts for this age group.
45We then use the age-specific atlas to provide evidence for reduced fractional anisotropy in several 46 tracts in children who were treated with therapeutic hypothermia for neonatal encephalopathy at birth 47 and did not have cerebral palsy, compared with controls matched for age, sex and socio-economic 48 status.