Personalized Caring: Integrating EEG/Visual Analysis With ChatGPT for MCI Assistance

This paper presents an AI-driven assistance scheme for individuals with mild cognitive impairment (MCI). The scheme combines electroencephalogram (EEG) sensing to monitor brain activity with a visual sensing module that captures environmental images. The analyzed EEG and visual data are fused to trigger personalized, context-aware conversations through cloud-based AI services such as ChatGPT. Text-to-speech and voice recognition technologies ensure smooth communication, while voice style transfer simulates familiar accents to improve comfort for MCI patients. The effectiveness of the proposal is validated by a prototype and simulations that incorporate EEG/visual data analysis. and personalized AI chatting.