Repository for the study of consciousness using graph theory
Comparing between different ways to build a graph for the graph-theoretical analysis of patients undergoing propofol-induced anesthesia.
- 9 Healthy adult undergoing anesthesia and recovering : subset of this bigger study
- Amplitude Enveloppe Correlation
- Weighted Phase Lag Index
- Directed Phase Lag Index
Each of these graphs will go through the following transformation:
- Keep the top X% of connection
- Minimal Spanning Graph of top connections
- Orthogonal Minimum Spanning Tree thresholding
Describe statistically the difference in the following feature for the graphs studied:
- global efficiency
- clustering coefficient
- modularity
- small worldness
over the following epochs:
- Baseline
- Induction of the anesthetic protocol
- Unconscioussness
- -5 minutes before recovery of consciousness
- +30 minutes after recovery of consciousness
- +60 minutes after recovery of consciousness
- +90 minutes after recovery of consciousness
- +120 minutes after recovery of consciousness
- +150 minutes after recovery of consciousness
- +180 minutes after recovery of consciousness
see the preprint titled Brain network motifs are markers of loss and recovery of consciousness for more information on both the participants and the calculation of these features.
The graph that best describe awareness is a permutation of these parameter. By doing a search in this space using a model we will be able to create this graph that may or may not be an hybrid of different graphs type. This graph can then be used in patient with a disorder of consciousness undergoing the same anesthethic pertubration protocol to characterize their awareness state. This is important because in this population the standard is to use behavioral sign of awareness to assess its presence.
- The "extra" folder contains:
- Some markdown files explaining the original aim of the project and some work done prior to the experiments, in folder "1_experiment_preparation"
- Some incomplete python code attempting to recreate the effect of the matlab code used for this experiment, in folder "2_python_code". Documentation for this code is provided in the same folder
- The "steps" folder contains the code neccessary for this experiment, separated in the relevent steps and a folder for utulity functions, "utils". Documentation is also provided in this folder.
Documentation for the code is provided in the "steps" folder. After cloning this repository, one needs to download the NeuroAlgo library, found here. One then needs to adjust the code to their needs, as described in the documentation provided in the "steps" folder.