Hot Yoga Dash – EMG-Based Rehab

(BE 3100 Final Project)

What:

Designed and built a low-cost, interactive rehabilitation system that uses real-time EMG signal processing to control an on-screen game character. The system gamifies physical therapy to encourage consistent participation and engagement.

How:

• Designed and implemented an analog filter circuit to extract heart rate and respiratory rate from ECG signals as part of preliminary testing.

• Built custom hardware to receive and process Bluetooth EMG data using a Raspberry Pi, enabling responsive detection of muscle flexes.

• Developed and tested the EMG signal acquisition pipeline, including filtering, thresholding, and flex classification to drive a game interface.

• Supported integration of hardware with a Python GUI, enabling character movement based on live EMG data.

  
  

Results:

• Achieved 97.95% accuracy for flex detection across 146 trials.

• Created a functional and engaging rehab game prototype, with future expansion opportunities such as historical logging and character selection.

• Demonstrated reliable muscle activation detection and system responsiveness, validating the feasibility of hardware-based EMG control in interactive rehabilitation tools.