Language Impairment Classifier Validation (MEG / Autism)

Regulatory-style validation of a speech-discrimination biomarker for language impairment in autism using blinded ROC and mixed models.

Role: Post-Doctoral Researcher, Lurie Family Foundations MEG Imaging Center, CHOP (2008-2011)

Highlights:

• AUC 0.86; sensitivity 82.4%; specificity 71.2%; Cohen’s d = 3.11
• 78 pediatric participants; 301 observations
• Bonferroni correction; pre-specified criteria; nonparametric linear mixed modeling

Outcomes: Roberts et al. (2011). Biological Psychiatry – highly cited clinical biomarker paper

MEG ROC analysis linear mixed models MATLAB