Assessing hippocampal development and language in early childhood: Evidence from a new application of the Automatic Segmentation Adapter Tool

Joshua K. Lee, Christine W Nordahl, David G Amaral, Aaron Lee, Marjorie Solomon Friedman, Simona Ghetti

Research output: Contribution to journalArticle

19 Scopus citations

Abstract

Volumetric assessments of the hippocampus and other brain structures during childhood provide useful indices of brain development and correlates of cognitive functioning in typically and atypically developing children. Automated methods such as FreeSurfer promise efficient and replicable segmentation, but may include errors which are avoided by trained manual tracers. A recently devised automated correction tool that uses a machine learning algorithm to remove systematic errors, the Automatic Segmentation Adapter Tool (ASAT), was capable of substantially improving the accuracy of FreeSurfer segmentations in an adult sample [Wang et al., 2011], but the utility of ASAT has not been examined in pediatric samples. In Study 1, the validity of FreeSurfer and ASAT corrected hippocampal segmentations were examined in 20 typically developing children and 20 children with autism spectrum disorder aged 2 and 3 years. We showed that while neither FreeSurfer nor ASAT accuracy differed by disorder or age, the accuracy of ASAT corrected segmentations were substantially better than FreeSurfer segmentations in every case, using as few as 10 training examples. In Study 2, we applied ASAT to 89 typically developing children aged 2 to 4 years to examine relations between hippocampal volume, age, sex, and expressive language. Girls had smaller hippocampi overall, and in left hippocampus this difference was larger in older than younger girls. Expressive language ability was greater in older children, and this difference was larger in those with larger hippocampi, bilaterally. Overall, this research shows that ASAT is highly reliable and useful to examinations relating behavior to hippocampal structure.

Original languageEnglish (US)
JournalHuman Brain Mapping
DOIs
StateAccepted/In press - 2015

Keywords

  • Autism
  • Development
  • Freesurfer
  • Hippocampus
  • Language
  • Methods
  • Segmentation
  • Segmentation adapter

ASJC Scopus subject areas

  • Clinical Neurology
  • Anatomy
  • Neurology
  • Radiology Nuclear Medicine and imaging
  • Radiological and Ultrasound Technology

Fingerprint Dive into the research topics of 'Assessing hippocampal development and language in early childhood: Evidence from a new application of the Automatic Segmentation Adapter Tool'. Together they form a unique fingerprint.

  • Cite this