
BBD Body Monitor
BBD Body Monitor is a single channel data acquisition system (DAS) to provide a bio-potential reading in the 0.25Hz - 125kHz frequency range. The main focus of the prototype is to use machine learning to detect different mental states (eg. focused wakefulness, relaxation, meditation, REM sleeping, deep sleeping or lucid dreaming) by digitizing potential changes on our bodies.
Last 30 days:
Project type
Data VisualizationHealthMachine LearningRaspberry PiToolsProject tech
C#Currently seeking
- Backend Developers
- Developers
- Frontend Developers
- Researchers
- Testers
- UX
Contribution overview
Most contributors are in Europe
Overview
This is the 8th generation Bio Balance Detector prototype that I work on.
The full description can be found here:
https://github.com/andrasfuchs/BioBalanceDetector/wiki/Proto-%238-Body-Monitor
Summary:
BBD Body Monitor is a single channel data acquisition system (DAS) to provide a bio-potential reading in the 0.25Hz - 125kHz frequency range. The main focus of the prototype is to use machine learning to detect different mental states (eg. focused wakefulness, relaxation, meditation, REM sleeping, deep sleeping or lucid dreaming) by digitizing potential changes on our bodies.
Requirements:
The device must record bio-potential data in a convenient and mobile way. The recorded data must be processed onboard and give audio and/or visual feedback depending on the measured data. The device must work for at least 8 hours to record sleep, and it must be battery powered to avoid the 50/60 Hz artifacts that are present in all devices that are powered from the typical power outlet. An algorithm will be developed to recognize the different states we usually have, like wakefulness, relaxation, sleeps etc. If that is successful it can be more gradual, potentially recognizing emotions like contentment, joyfulness, the flow state, anger or frustration. The long-term goal is to find correlation between these measurements and physical and mental medical conditions.
Contributing
Onboarding isn't straightforward because it needs a hardware component. The software is fairly hardware-agnostic, but currently it's optimized for Digilent Analog Discovery 2 which is a $400 ADC.
On the other hand this project has a huge potential, it's a exiting research and it has a ton of cool tech to learn and use!
If you feel exited about the project too and you would like to join, let me know!