I conduct mixed-methods research at the intersection of digital health, human-computer interaction (HCI), and computer science, specializing in intelligent systems for individuals with cognitive impairments, including mild cognitive impairment (MCI) and dementia. My work evaluates how older adults with and without cognitive challenges engage with text-based digital dementia information and Generative AI (GenAI) systems, using webcam-based eye-tracking, facial expression analysis, and affective computing to measure cognitive load, emotional responses, trust, accessibility, readability, tone, hallucinations, and stigma. I develop and assess LLM-based chatbots with Retrieval-Augmented Generation (RAG), fine-tuned on curated dementia datasets, and co-trained models with people living with MCI/dementia (n=14) to enhance accuracy, reliability, and trustworthiness compared to mainstream systems. Leveraging quantitative modeling, qualitative insights, and multimodal methods, I design inclusive, explainable AI solutions that reduce bias, improve transparency, and promote well-being across diverse populations.