About the Project

BioSense-AI is a software-hardware project focusing on disease prediction using ECG-PPG, temperature, and text prompts. Our pocket-size electronic device, combined with a mobile app, displays and analyzes body signals with a custom multi-model deep learning architecture.

Signals undergo amplification and filtering via custom analog circuits. Amplified signals are then transmitted to an Orange Pi Zero 2W SBC from an ADC for digital processing. The system utilizes a custom multi-model deep learning architecture with interpretability. Ongoing development includes the creation of a mobile app for user access to the AI model. Users can interact with the app by scanning ECG reports and other prompts, catering to hospitals without the electronic device.

Tools & Technologies Used

• PyTorch
• GradCAM
• lime
• Analog Filter Design
• Setting up Orange Pi zero 2W SBC for custom tasks
• I2C protocol
• ADC
• Flutter