Multimodal and Explainable Artificial Intelligence for Precision Healthcare: Integrating Federated Learning, Governance, and Affective Computing

Author: Venus Sosa Iglesias, Nand Janakbhai Modi, Riba Sebastian Purackal, Khushnood Shoukat, Alicem Koyun, Maria Ndatiwelao Nangolo, and Kyungchan Park

Issue: Summer Issue, 2026

Abstract

Healthcare systems face increasing pressure from an aging population, rising rates of chronic disease and comorbidity, workforce shortages, clinician burnout, escalating care costs and fragmented digital infrastructure. Artificial intelligence (AI) has emerged as a transformative enabler to support descriptive, diagnostic, predictive, and prognostic big data analytics for personalized treatment planning, risk stratification, and longitudinal patient monitoring. However, the current AI paradigm is fragmented across clinical domains and is constrained by limited interoperability, insufficient external validation, algorithm opacity, and demographic bias. Governance and regulatory frameworks lag behind technological advancement, impeding AI adoption and eroding stakeholder trust. This narrative review uses a five-pillar framework for AI-enabled precision healthcare composed of (1) multimodal AI that integrates heterogeneous data sources; (2) explainable AI to improve interpretability, clinical accountability, and regulatory transparency; (3) affective computing and human-centered AI to create a therapeutic alliance with patients; (4) privacy-preserving infrastructures including federated learning (FL), differential privacy, and blockchain-enabled auditability to secure interinstitutional collaboration; and (5) adaptive governance systems for equitable, ethical, and sustainable deployment. Peer-reviewed scientific advances published between 2018 and 2026 are examined. The authors in this review argue that responsible and trustworthy AI-enabled precision healthcare should transition from isolated predictive models to complex socio-technical systems. Through the integration of these socio-technical systems into clinical workflows using adaptive ethical governance and privacy-preserving collaborative infrastructures, clinical reasoning and patient-centered care can be improved.