Local shape analysis in children at high risk for autism (IBIS) (2015)

Undergraduates: Kirsten Nicole Consing, Beatriz Paniagua, Joseph Piven, Mark Shen, Martin Styner

Faculty Advisor: Martin Styner
Department: Psychology & Neuroscience

---

Statistical shape analysis allows us to specifically pinpoint morphological changes in the subcortical structures of the hippocampus, caudate, putamen, and thalamus of infants at high familial risk (HR) for autism spectrum disorder (ASD) at 12 and 24 month time points. In the first study of its kind, we perform shape analysis on the subcortical structure data from the Infant Brain Imaging Study Network (IBIS). Subcortical structures were computed on a large sample using the SPHARM-PDM framework. Statistical analysis measured medial profiles of local thickness, volume, and surface area at each time point. We also analyzed SPHARM-PDM surface vertices to perform quantitative morphological assessment at all surface locations. Medial and surface thickness analysis show that HR infants with ASD compared to those without ASD show significant group interactions in the left hippocampus at 12 months and 24 months, right hippocampus at 12 months, left putamen at 24 months and both thalami at 12 months. Surface coordinate analysis show that significant regional mean differences in the left and right caudate, putamen, and thalamus at 12 months as well as significant regional mean differences in the left caudate, left putamen, and both hippocampi at 24 months. This suggests that at the pre-diagnostic stage, HR infants with ASD have certain areas of local size reduction in the subcortical structures particularly in the caudate, hippocampus, and thalamus compared to HR infants without ASD.