Infant Brain Imaging Study (IBIS)

The Infant Brain Imaging Study (IBIS) is a longitudinal MRI and behavioral study run by the IBIS Network, an NIH Autism Center of Excellence. It follows infants who have an older sibling with autism, and so a raised familial likelihood of autism themselves, together with low-risk infants, from 6 to 24 months of age. The aim is to find brain changes that come before the behavioral signs of autism. Imaging and behavioral data are shared through NIMH Data Archive collection 19, which lists 1039 subjects with shared data.

Composition

Infants were enrolled at 6 months and returned at 12 and 24 months. Exclusion criteria included prematurity, low birth weight, major medical or sensory conditions, contraindications for MRI, adoption and a first-degree relative with intellectual disability, psychosis, schizophrenia or bipolar disorder. Autism outcome is assessed at 24 months with the ADOS, and development at each visit with the Mullen Scales of Early Learning. Published analyses use subsets: Hazlett et al. (2012) compared 98 high-risk and 36 low-risk infants at 6 months, and Wolff et al. (2012) followed 92 high-risk infants with diffusion data, 28 of whom met ASD criteria at 24 months. The NDA record does not break the shared subjects down by group, age, sex or visit.

Acquisition

Four clinical sites (UNC Chapel Hill, University of Washington, Children's Hospital of Philadelphia and Washington University in St. Louis) used identical Siemens 3T TIM Trio scanners with 12-channel head coils, with data coordinated at the Montreal Neurological Institute. Infants were scanned in natural sleep without sedation. The protocol has 1 mm isotropic 3D T1w MPRAGE and 3D T2w FSE scans and a 25-direction DTI at 2 mm isotropic with b-values up to 1000 s/mm². Monthly phantom scans and yearly traveling human phantoms tracked scanner stability across sites.

Annotations

No manual image labels are described. A pediatric neuroradiologist reviews the scans for incidental findings, and diffusion data pass automated quality control.

Known limitations

• The high-risk group is defined by family history, so most infants are siblings of children with autism; low-risk controls are fewer in the published analyses.
• Later IBIS follow-ups (school age, adolescence) and newer IBIS studies are not described by NDA collection 19.
• Release dates and per-release contents are not published.