Reading Ability

PI Name: Guinevere F. Eden
Institution: Georgetown University


Use LONIR software development and computer resources to assess the performance of various feature selection and classification methods in predicting reading ability based on a combination of sMRI and fMRI measures. We will assess the performance of four different computational methods in identifying good versus poor readers, based on a combination of anatomical measures and patterns of brain activation. Specifically, we will make use of three existing data sets (acquired under previous NIH grants) to (1) compare and contrast different types of algorithms in adults (using two data sets) and then (2) training the algorithm for pediatric data.