2014 Summer Project Week:Stroke-ImagingGenetics

• 

  Projects List

• 

  Clinical Stroke Image

• 

  Leukoaraiosis Model

Key Investigators

• Adrian Dalca, Ramesh Sridharan, Polina Golland, MIT
• Natalia Rost, Jonathan Rosand, MGH

Project Description

This is one of a series of projects on a large clinical dataset of stroke patients, including T2-FLAIR and DWI imaging, and genetic information. The ultimate goal is to investigate clinical and genetic causes affecting stroke outcome and small vessel disease. We recently (MICCAI 2014) developed a leukoaraiosis (small vessel disease) probabilistic co-variation model for T2-FLAIR, which we use to differentiate between small vessel disease and acute or chronic stroke. In this work, we wish to investigate the possible genetic and clinical effects related to this model.

Average Leukoaraiosis projection

Explore the first component of the model

Objective

• Here, we wish to investigate the parameters of the model as potential phenotypes against for clinical and genetic analyses.

Approach, Plan

• We will create the phenotype projections for ~80 stroke patients and experiment with correlating the genetic and clinical factors.

Progress

• We've gotten some feedback on looking at imaging genetics, and run the projection on 100 projections. The genetics, however, needs much more processing.