Shake, Rattle, and Roll: Detection of Parkinsonian Symptoms

This project was developed as part of the Embedded Challenge in Embedded System Design (ECE-UY 4144) at NYU Tandon during Spring 2025.
All rights reserved by the authors.

Overview

This embedded system detects Parkinsonian tremors and dyskinesia using a 3-axis accelerometer onboard the Adafruit Circuit Playground Classic. By applying a Fast Fourier Transform (FFT) to 3 seconds of motion data, the system distinguishes between two key frequency bands:

• 3–5 Hz: Parkinsonian tremors (displayed with blue LEDs)
• 5–7 Hz: Dyskinesia (displayed with red LEDs)

The number of LEDs lit corresponds to the intensity of the detected symptoms.

Features

• Real-time frequency analysis using FFT
• Blue LEDs (positions 0–4) indicate tremor intensity
• Red LEDs (positions 5–9) indicate dyskinesia intensity
• Both red and blue LEDs light up if intensities are equal
• A small red LED (pin 13) indicates data collection
• Left button (PD4) restarts the detection cycle

How It Works

1. Collects 128 samples over 3 seconds
2. Calculates the magnitude of acceleration using all three axes
3. Applies FFT to analyze the frequency spectrum
4. Measures energy in 3–5 Hz and 5–7 Hz bands
5. Displays detection results with onboard LEDs
6. Pressing the left button resets the test

License

This project is for academic use only.
Copyright © 2025, Simon Hu and contributors