Title: Multiple imputation methods for functional data with applications in mental health research
Authors: Adam Ciarleglio - George Washington University (United States) [presenting]
Abstract: In mental health research, the number of studies that include multimodal neuroimaging is growing. Often, the goal is to integrate both the clinical and imaging data to address a specific research question. Functional data analytic tools for analyzing such data can perform well, but these methods assume complete data. In practice, some proportion of the data may be missing. We present several approaches for imputation of missing scalar and functional data when the goal is to fit functional regression models for the purpose of estimating the association between a scalar or functional outcome and scalar and/or functional predictors. We present results from a simulation study showing the performance of various imputation approaches with respect to fidelity to the observed data and estimation of the parameters of interest. The proposed approaches are illustrated using data from a placebo-controlled clinical trial assessing the effect of SSRI on subjects with major depressive disorder.