Activity Number:
|
528
- Contributed Poster Presentations: Section on Statistics in Imaging
|
Type:
|
Contributed
|
Date/Time:
|
Wednesday, July 31, 2019 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Section on Statistics in Imaging
|
Abstract #307349
|
|
Title:
|
Population-Level Representational Mapping Based on Intracranial EEG Subjects with Varying Spatial Sampling
|
Author(s):
|
Peter W. Elliott* and Max G'Sell
|
Companies:
|
Carnegie Mellon University and Carnegie Mellon University
|
Keywords:
|
Neuroimaging;
Multiset CCA;
Factor Models
|
Abstract:
|
Intracranial EEG (iEEG) has shown promise as a tool for neuroscientific research in clinical settings due to its high signal to noise ratio and spatio-temporal resolution. However, conducting group analyses of iEEG experiments has proven challenging as individual subjects typically have limited and idiosyncratic brain coverage. In this work we develop a method for group analysis of representational content across space. We use Multiset Canonical Correlation Analysis (MCCA) for inter-subject alignment, aggregating analysis across subjects with relevant spatial sampling while accounting for individual variation in signal strength. Spatial contributions to classification within each subject are assessed using overlapping cluster modeling of electrodes. We apply this method to a visual category experiment to reveal organizing principles of the human visual processing system.
|
Authors who are presenting talks have a * after their name.