Eeg

🧠 Wearable EEG Device – Biomedical Instrumentation I

A wearable EEG system for real-time brain activity monitoring, designed with embedded systems and IoT principles.

---

📚 Project Overview

This project presents a wearable Electroencephalogram (EEG) device developed for the Biomedical Instrumentation I course. It aims to acquire and process low-amplitude human brain signals (5µV to 300µV) across standard EEG frequency bands:

• Delta (δ): 0.5 – 4 Hz
• Theta (θ): 4 – 8 Hz
• Alpha (α): 8 – 13 Hz
• Beta (β): 13 – 30 Hz
• Gamma (γ): 30 – 100 Hz

This is a full-stack biomedical system, integrating analog signal conditioning, high-resolution digital conversion, and wireless data transmission for real-time EEG visualization.

EEG Hardware Prototype

Real-Time EEG Interface in Qt

Physical Setup

---

⚙️ Hardware Architecture

🧩 Microcontroller Core

• STM32H730VBT SoC
• Features a 16-bit SAR ADC up to 3.6 MSPS, ideal for high-resolution EEG data acquisition.

🔌 Analog Front-End (AFE)

EEG signals are amplified ~12000x and filtered in a 3-stage analog pipeline:

1. INA333 Instrumentation Amplifier – 20x gain + initial passive filtering
2. OPA333 Operational Amplifier – 600x gain
3. 6th-Order Passive Bandpass Filter – 0.5 Hz (HP) to 45 Hz (LP)

📡 Communication & Storage

• External Flash: W25Q512JV QSPI NOR Flash for firmware and logs
• Wireless Co-Processor: ESP32-C6-MINI-1H4, communicating via UART

---

💾 Firmware Architecture

STM32H7 Side (Signal Acquisition)

• DFSDM Peripheral used for enhanced digital filtering
• Oversampling & Bit Shifting increase ADC ENOB
• Implements Wavelet Transform for real-time brain wave classification

ESP32 Side (Wireless Stack)

• Developed using ESP-IDF and FreeRTOS

• Handles:

  • Serial communication with STM32
  • Data packetization
  • Wi-Fi communication to local host
  • MQTT protocol for lightweight IoT data streaming

---

🧪 Real-Time EEG Visualization

On the host side:

• A Qt interface provides live graphical feedback of EEG signals.
• Visualizes brain wave patterns and band activity in real time.

---

🛠️ Technologies Used

---

🚧 Future Improvements

• ✅ Integrate live FFT for spectral analysis
• ✅ Mobile dashboard version
• 🔲 Dry electrodes support
• 🔲 Artifact rejection (e.g., blink/motion noise)
• 🔲 Bluetooth Low Energy (BLE) alternative

---

👨‍🔬 Authors & Acknowledgments

Stephan Costa Barros - Electrical Engineering Department – Federal University of Uberlândia (UFU), Brazil

Marcelo Barros de Almeida - Electrical Engineering Department – Federal University of Uberlândia (UFU), Brazil

Alcimar Barbosa Soares - Electrical Engineering Department – Federal University of Uberlândia (UFU), Brazil