With the rapid advancement of medical technology and the accelerated digital transformation, the field of health informatics is encountering unprecedented opportunities and challenges. Multimodal data, including electronic health records (EHR), medical images, genomics data, and monitoring data from wearable devices, are accumulating at an unprecedented rate. These data sources are extensive and diverse in form, containing a wealth of health information that provides strong support for precision medicine, disease prediction, and the formulation of personalized treatment plans. However, how to effectively integrate these multimodal data to achieve more accurate and comprehensive health information analysis and decision - making support has become a crucial issue urgently needing to be addressed in the current field of health informatics. In recent years, the rapid development of artificial intelligence and machine learning technologies has brought new hope for the fusion and application of multimodal data. Nevertheless, it also faces numerous challenges such as data quality, privacy protection, and model interpretability.

This Research Topic aims to promote the innovative application of multimodal data in next-generation health informatics, focusing on the following core directions. Technically, it systematically explores the cutting-edge technical routes for multimodal data fusion, optimizing early, mid-stage, and late-stage fusion strategies, developing new architectures, enhancing cross-modal feature alignment capabilities, optimizing the efficiency of data integration, and establishing an evaluation system. In terms of data quality, it conducts in-depth research on typical problems in multimodal medical data and corresponding countermeasures, such as dealing with data noise and compensating for missing modalities, while emphasizing the technical paths for data fusion that ensure privacy and security. Regarding model interpretability, it explores methods to enhance the interpretability of multimodal medical AI, and improves the transparency and credibility of model outputs. In the dimension of ethics and security, it studies privacy protection technologies and strategies for algorithm fairness assessment. For clinical application translation, it focuses on the innovative applications of multimodal data in key scenarios such as early disease warning and diagnosis, and explores effective paths to translate laboratory research results into clinical practice.

We encourage authors to submit original and distinctive research findings, including theoretical research, experimental research, and review research. Moreover, this topic promotes interdisciplinary collaborative research, integrating expertise from multiple fields such as computer science, clinical medicine, and bioinformatics to jointly overcome the key technical bottlenecks in the development of multimodal health informatics and drive the transformation of medical artificial intelligence from theoretical innovation to practical application. Topics of interest include (but are not limited to):

1. Multimodal medical foundation models and general AI;
2. Robust fusion of multimodal data;
3. Disease-specific multimodal diagnostic systems;
4. Automated medical report generation and interpretable AI;
5. Multilingual and global medical AI;
6. Dynamic multimodal health monitoring;
7. Multimodal data privacy and federated learning;
8. Ethics, bias, and fairness in medical AI.

Are you interested in contributing a manuscript to this Research Topic?

Harnessing Multimodal Data for Next-Gen Health Informatics