Analysis & Results

In this study, we wanted to see how damage to certain areas of the brain was related to the development of REM Behavior Disorder.

In order to answer this question, we need to see how the brain injuries (shown on the MRI scans we collected from case studies) affect various brain networks. These networks area basically just groups of brain regions that tend to be active, or inactive, at the same times. So, if region A and region B tend to be working hard at the same time, or tend to be quiet at the same times, we would say those areas are "functionally connected", because their levels of functioning are generally similar to each other. Neural networks are what we call groups of "functionally connected" brain regions.

STEP 1

Find case studies where someone received a brain injury and then developed RBD.

STEP 2

Look at the picture of the individual's brain (from the medical case study).

Draw the injured area onto a picture of an average brain using a specialized software called 3D Slicer, which is used to visualize medical data in 3D models.

STEP 3

Save all those drawings, and put them in a supercomputer for analysis.

STEP 4

The supercomputer software, called FSLeyes, compares my drawings of brain injuries to a huge amount of data that it already has saved. For each individual drawing of one brain injury, the computer answers the question, "what brain networks is this injured area a part of?", or, "what other areas of the brain tend to be active when this area is active?"

Then, the software highlights all the areas that the injured region is connected to (ie., it highlights all the networks the injured region is a part of).

It does this for each drawing, creating 13 individual maps of affected networks (because there were 13 individuals in this study).

STEP 5

The software repeats this process, but this time, it highlights all the areas that tend to be inactive when the injured area is active. It creates 13 maps of areas of inactivity.

STEP 6

The computer layers the 13 positive and the 13 negative maps to create one map of all the areas that tend to be active / inactive at the same time that the injured area is active/inactive. Each of the 26 individual maps are different, so when they are layered, there are some areas where all 13 of the positive maps overlap, but there are other areas where only 2 or 3 of the 13 maps overlap.

STEP 7

We ask the software to create a final map showing only the areas where 12/13 or more of the maps overlap. This way, we know what areas were positively connected and negatively connected to the injured area for a full 12 of 13 individuals.

STEP 8

Here is the final map. The green shows areas that tend to be active when the injured area is active, and the red shows areas that tend to be inactive when the injured area is active.