sEEG-Action
Investigating goal-specific processing of observed actions in human brain utilizing intracranial sEEG recordings
About the project
- We examined the goal-specific processing of observed actions utilizing sEEG recordings from the human brain
- Intracranial sEEG data from 40 epilepsy patients, 9000+ electrodes
My Contributions:
- Led data cleaning and preprocessing initiatives. Managed the project, coordinating team efforts and milestones.
- Developed automatic artifact rejection techniques utilizing methods such as kurtosis and detection of high-frequency oscillations.
- Mapped electrode coordinates using CT and structural MR images and executed transformations between coordinate systems.
- Employed statistical and machine learning techniques including General Linear Models (GLM), Support Vector Machines (SVM), and Linear Discriminant Analysis (LDA).
- Utilized a variety of tools for analysis and visualization, including MNE, Brainstorm, Slicer, Caret, Nilearn, Nipype, HDF5, scikit-learn, NumPy, pandas, and SciPy.
- Enhanced computational efficiency through parallel processing using Dask.