ClosedLoop

ClosedLoop provides real-time EEG-based neurofeedback processing and cognitive-state decoding to support neurofeedback research and interventions.

Key Features:

• Real-Time Signal Processing: Performs real-time signal preprocessing of scalp EEG recordings and includes adaptive artifact rejection to minimize noise and interference.
• Cognitive State Classification: Implements advanced algorithms for classification of cognitive states from EEG with continuous updating of classification models using recently recorded data to enable real-time feedback without prior recordings.
• Modular and Extensible Design: Implemented entirely in Python with a modular architecture that supports extensibility and customization for diverse analyses.
• Proof of Concept — Attention Training Paradigm: Demonstrated an attention training paradigm using a consumer-grade dry-electrode EEG system in a 22-participant, three-day study achieving a mean decoding error rate of 34.3% (chance 50%) for subjective attentional states.

Scientific Applications:

• Neurofeedback Training: Enables targeted training of brain activity through real-time EEG neurofeedback for experimental and clinical protocols.
• Cognitive State Monitoring: Supports real-time decoding of subjective attentional states for studying attention and other cognitive processes.
• Personalized Intervention Development: Facilitates development and evaluation of personalized neurofeedback interventions and translational therapeutic studies.

Methodology:

Real-time signal preprocessing of scalp EEG, adaptive artifact rejection, cognitive-state classification using advanced algorithms with continuous model updating, implemented in Python.