A0866
Title: Statistical challenges for large neuroimaging cohort studies
Authors: Haochang Shou - University of Pennsylvania (United States) [presenting]
Abstract: With the increasing need for big data analytics in medical imaging, integrating data from multiple studies and various biological domains has become critical to better understanding complex human diseases. Analysis of such data is challenging due to the existence of site differences that could mask the biological associations of interest. The different data property of various modalities further poses difficulties in conducting integrative analyses to evaluate the joint relationship among modalities. Several most recent developments in statistical harmonization methods in large neuroimaging studies under various data modalities will be discussed. These approaches are designed to mitigate site differences that exist in mean, variance, and covariance structures in structural and functional imaging outcomes. A novel distance-based regression model, referred to as Similarity-based Multimodal Regression (SiMMR), will be discussed, enabling simultaneous regression of multiple modalities through their distance profiles. The proposed method can detect associations of differing properties and dimensionalities in multimodal data, even with modest sample sizes. The methods are motivated by the iSTAGING (Imaging-based coordinate SysTem for AGing and NeurodeGenerative diseases) consortium, which brings together multisite neuroimaging studies to understand the complex and heterogeneous process of normal ageing and AD pathology.
