This repository includes the code used for the paper titled Approximate Zero-Crossing: A new interpretable, highly discriminative and low complex feature for EEG and iEEG seizure detection
Diffusion AZC Aproximate Zero-Crossing Feature for EEG/iEEG (master)
Recent Commits
Recent Commits
| Commit | Author | Details | Committed | ||||
|---|---|---|---|---|---|---|---|
| 1d2a71304a70 | renatoznt | Scripts for the paper titled 'Approximate Zero-Crossing: A new interpretable… | Oct 3 2022 |
readme.md
readme.md
This repository will include the code used for the paper entitled AZC: A new feature for EEG and iEEG seizure detection]{Approximate Zero-Crossing: A new interpretable, highly discriminative and low complex feature for EEG and iEEG seizure detection
License: LGPL
We have used Anaconda toolset, including the following packages (original and added later):
- Python (v3.9.7)
- numpy (v1.20.3)
- pandas (v1.3.5)
- scikit-learn (v1.0.2)
- scipy (v1.7.3)
- seaborn (v0.11.2)
- matplotlib (v3.5.0)
- pyedflib (v0.1.25)
- antropy (v0.1.4)
- PyWavelets (v1.1.1)
Script descriptions:
- script_AZC_datasetPreProcessing.py: to be executed first, including feature extraction and preparation for TSCV execution.
- script_AZC_FeatDivergence.py: calculates the KL divergence and generate related plots.
- script_AZC_Classification.py: used for seizure classification (TSCV approach).
- script_AZC_ConsolidateResults.py: consolidation of results and its plots.
All scripts include few expected user setups before execution (e.g., the path of the folder containing the datasets), also observing the parameters set in parametersSetup.py.
It's important to note that the scrpit script_AZC_datasetPreProcessing.py employs a pool of processing cores in case the variable parallelize is set. Similar behaviour can be achieved for script_AZC_Classification.py script, but in this case, you should launch the execution via bash using the bash_run_python_variosPat.sh script.
c4science · Help