Creativity in vitro

Creativity in Vitro is a brain-computer interface prototype that translates neural signals into text using EEG data and machine learning. We aimed to explore whether imagined mental states (like visualization or relaxation) could be classified and turned into prompts for AI communication. The target audience includes BCI researchers, accessibility advocates, and creative technologists. We built a complete ML pipeline: from raw EEG data acquisition using OpenBCI, through filtering, artifact removal, and feature extraction (band power, Hjorth, wavelets), to classification using CNNs and classical models (LightGBM achieved 98% accuracy). We also developed an AI-powered literature review tool to process EEG-related scientific articles. Next steps include expanding the vocabulary beyond binary mental states and integrating real-time symbolic feedback for inclusive communication.