Physiological signal classification (Brainhack project)

[drombas]

Project goal

• Build a physiological signal classification algorithm

Important details

• The goal of the project is only to generate ideas and make them work. We do not expect to merge this into the main branch during hackaton (that would come later)
• Although challenging we are aiming for a 100% accuracy as misclassification can be highly problematic
• To maximize generalization and robustness, solutions that make use of intuitive features are prefered over fancy algorithms/features.
• Do not hesitate to contact us via mattermost. We are happy to collaborate and provide any extra info.

Dataset description (link)

• 4 type of signals (cardiac, respiratory chest, O2 and CO2)
• 240 time-series (60x4) recordings of 500 seconds long
• 2 files: time_series.csv (time-series data) and annotations.csv (time-series labels)

Suggested steps

1. Have a look at the docs to get an overall idea of what phys2bids is
2. Read the docs chapter about physiological signal types
3. Download the dataset
4. Visualize the signals to see how they look like
5. Think, share and test ideas about approaches/features that could work

[drombas]

I close this because there is already a specific repository and an issue related to this.