VR-Driven Neuro-Symbiotic AI for Personalised Psychological Flow State Modulation in Parkinson’s Disease

People with Parkinson’s disease face fluctuating motor and cognitive symptoms that undermine engagement with conventional rehabilitation, which largely lacks real-time personalisation. This research investigates the development of a Neuro-Symbiotic AI framework that integrates virtual reality, EEG-based brain-computer interfaces, and adaptive machine learning to detect and modulate psychological flow states during VR therapy. The system aims to adjust the VR environment dynamically in response to real-time neural signals, sustaining the challenge-skill balance associated with therapeutic engagement. Key technical components include Graph Neural Networks for EEG connectivity analysis and a Bayesian decision framework for adaptive control. The project is currently in its second year, with work focused on establishing reliable EEG markers of flow in healthy participants ahead of trials with Parkinson’s populations. The research is self-funded and conducted on a part-time basis.