Artificial Intelligence (AI) holds great promise for transforming patient care, yet its integration often neglects the lived experiences of patients. This paper explores how AI can be responsibly applied to support patient-centered care across the patient journey. Drawing on twelve expert interviews with researchers and practitioners from the fields of AI and healthcare, we identify both critical potentials and challenges of AI deployment across care phases – as from symptom awareness to aftercare. Findings reveal three foundational pillars – Technology, Humanity, and Environment (THE) – that frame responsible AI use. Results emphasize the importance of explainability, human-in-the-loop systems, inclusive design, and institutional alignment. We develop a conceptual framework that maps context-specific AI applications along the patient journey and derive guiding principles for their responsible use. The study contributes by complementing socio-technical perspectives with relational and ethical dimensions, and offers practical orientation for aligning AI deployment with clinical realities and infrastructural constraints in healthcare.