BCI-MCP Server

# BCI Features This document outlines the key features of the Brain-Computer Interface (BCI) component of the BCI-MCP system. ## Supported Devices The BCI-MCP system is designed to work with a variety of BCI hardware devices, including: ### OpenBCI - **Cyton Board**: 8-channel EEG acquisition - **Ganglion Board**: 4-channel EEG acquisition - **Cyton + Daisy**: 16-channel EEG acquisition - **WiFi Shield**: Wireless data transmission ### Emotiv - **EPOC+**: 14-channel EEG headset - **EPOC Flex**: Advanced EEG acquisition with flexible positioning - **Insight**: 5-channel mobile EEG headset ### NeuroSky - **MindWave**: Single-channel EEG headset ### Custom Hardware - Support for custom and DIY EEG hardware through configurable device interfaces ## Data Acquisition ### Sampling Capabilities - Adjustable sampling rates (up to 1000 Hz depending on hardware) - Multi-channel data acquisition - Real-time impedance checking - Signal quality monitoring ### Data Formats - Standard EDF/EDF+ format support - CSV export functionality - Integration with common EEG data formats - Raw data access for custom processing ## EEG Monitoring ### Real-time Visualization - Time-domain signal plotting - Frequency spectrum analysis - Topographical mapping - Custom visualization components ### Impedance Testing - Real-time electrode impedance monitoring - Visual feedback for connection quality - Electrode status indicators ## Supported Paradigms ### P300 - Oddball paradigm implementation - P300 speller matrix - Target detection ### Steady-State Visual Evoked Potentials (SSVEP) - Frequency-coded stimulation - Phase-coded stimulation - Multi-target detection ### Motor Imagery - Left/right hand imagery - Multiple body part classification - Continuous control paradigms ### Passive BCI - Cognitive workload monitoring - Attention level tracking - Emotional state detection ## Markers and Events ### Event Annotation - Precise timestamp synchronization - Custom event markers - Experimental protocol design tools ### Trigger I/O - External trigger input/output - Hardware synchronization - Integration with stimulus presentation software ## Extension Capabilities ### Plugin Architecture - Custom signal processing plugin support - Protocol extension framework - Device driver extensibility ### API Access - Comprehensive Python API - WebSocket streaming for web applications - Network data transmission ## Example Usage ```python from bci_mcp.devices import OpenBciDevice from bci_mcp.visualization import SignalViewer # Connect to an OpenBCI Cyton board device = OpenBciDevice(port="/dev/ttyUSB0", board_type="cyton") device.connect() # Start data streaming device.start_stream() # Create a real-time signal viewer viewer = SignalViewer(device) viewer.show() # Add an event marker device.add_marker(code=1, description="Stimulus onset") # Access raw data data = device.get_data(seconds=10) # Stop streaming when done device.stop_stream() device.disconnect() ``` ## Next Steps To understand how these BCI features integrate with the Model Context Protocol, see the [MCP Integration](mcp-integration.md) documentation.