July 26, 2017 from 2:45 p.m. to 3:45 p.m.
TMII 1st Floor
TMII Large Conference Room 117
Hess Center for Science & Medicine
1470 Madison Avenue, New York, NY
- Yuri Levin-Schwartz, PhD
- Postdoctoral Fellow
- Icahn School of Medicine at Mount Sinai
Abstract: Due to the ability of different sensors to provide complementary views of complicated systems, the collection of data from multiple sources has become common in neurological studies. Therefore, full utilization of the common information, while placing few assumptions on the data, forms the fundamental goal of performing a joint analysis on this data and motivates the use of multivariate data-driven methods. It is expected that each dataset will not contribute equally to the final result, but defining this contribution for real neuroimaging data is difficult. In this talk, I address this question in a few ways. First, through the use of a technique that combines principal component analysis and canonical correlation analysis, links between different neuroimaging datasets are determined. Second, a unique set of imaging datasets consisting of functional magnetic resonance imaging (fMRI) data, structural MRI data, and electroencephalogram data is used in order to assess the effects of including each dataset in an analysis. Finally, through the use of a classification rate-based procedure, the performance of different fusion methods on real multi-task fMRI data is quantified and the "value added" by each dataset to a joint analysis is determined.