Data Acquisition

  • Body slicing synchronization

Feature Calculators

  • Our system uses validated novel algorithms to calculate high-level clinical gait features from raw IMU data
  • Work in real-time to process incoming sample data from our BLE sensors and output useful features that can be used in our classifier

Classification System

  • Using the calculated features, the classifier identifies which gait deviations are present
  • It does this using an ensemble learning method

Feedback/Audio Engine

  • The feedback engine, used specifically in the ReLOAD project, classifies each step and works in tandem with the audio engine to map a deviation to an audio effect
  • The audio engine, built on top of Apple’s AVAudioEngine API, contains custom built audio effects used for real-time feedback