Multimodal AI (MMAI) for next-generation healthcare: data domains, algorithms, challenges, and future perspectives

Multimodal Artificial Intelligence (MMAI) is reshaping the landscape of next-generation healthcare by integrating diverse data sources—ranging from medical imaging and electronic health records (EHRs) to wearable sensor data and genomic sequencing. This convergence enables more accurate diagnostics, personalized treatment strategies, and real-time patient monitoring, ultimately transforming healthcare from reactive to predictive and preventive. Additionally, MMAI can lead to improved operational efficiency by enabling automated reporting and streamlining clinical workflows, helping to reduce clinician burnout and accelerate diagnostic turnaround times. Despite significant advancements, several challenges hinder the widespread adoption of MMAI, including data fragmentation, interoperability issues, computational demands, and the need for explainable AI (XAI) in clinical decision-making. This opinion paper explores four key aspects driving the future of MMAI in healthcare: (1) the evolution of multimodal data; (2) advancements in AI models and fusion strategies for extracting insights from heterogeneous data streams; (3) major challenges such as synchronization across modalities, interpretability, and regulatory constraints; and (4) emerging future directions, including the role of digital twins, automated clinical reporting, and precision medicine.