Explainable and Trustworthy Artificial Intelligence in Precision Medicine：Multimodal Data Analysis and Clinical Decision Support

Artificial intelligence (AI) is transforming medicine by enabling the analysis of medical images, electronic health records, pathological data, physiological signals, genomic profiles, and other multimodal health information. AI-assisted technologies have shown considerable potential for disease detection, risk prediction, patient stratification, prognosis assessment, treatment planning, and clinical decision support. However, translating AI models from experimental settings into routine clinical practice remains challenging. Many systems function as “black boxes,” making it difficult for clinicians to understand how predictions are generated or whether the identified patterns are medically plausible. In addition, models developed on internal datasets may be sensitive to differences in patient populations, institutions, imaging devices, data-acquisition protocols, and clinical practices. Biased training data, inadequate uncertainty estimation, privacy risks, limited external validation, and insufficient reproducibility further hinder clinical implementation. Predictive accuracy alone is therefore insufficient to establish the safety, reliability, and clinical value of medical AI. Emerging approaches, including explainable AI, causal and counterfactual learning, uncertainty estimation, multimodal learning, federated learning, and human–AI collaboration, offer opportunities to address these limitations, although important methodological and translational gaps remain.This Research Topic aims to bring together clinicians, medical researchers, epidemiologists, statisticians, computer scientists, and biomedical engineers to advance explainable and trustworthy AI for precision medicine. It will highlight methodological innovations and clinically meaningful applications in medical data analysis, disease-risk assessment, patient stratification, treatment-response prediction, and clinical decision support. Particular emphasis will be placed on model interpretability, reliability, generalizability, fairness, privacy, reproducibility, and clinical utility. Submissions should address clearly defined medical questions and explain how the proposed methods may contribute to disease diagnosis, prognosis, treatment, monitoring, or prevention. Studies involving external validation, multicenter datasets, prospective evaluation, or integration into clinical workflows are especially welcome.To gather further insights into the development, validation, and clinical translation of trustworthy AI for precision medicine, we welcome articles addressing, but not limited to, the following themes:• Explainable, interpretable, causal, and counterfactual AI for clinically meaningful reasoning;• Robust and generalizable medical AI, including uncertainty estimation, fairness, and bias mitigation;• AI-assisted analysis of medical images and multimodal clinical data;• Individualized disease-risk prediction, patient stratification, and treatment-response assessment;• Privacy-preserving and federated learning for medical applications;• Clinical validation, human–AI collaboration, and integration into clinical workflows;• Ethical, regulatory, and reproducibility considerations in medical AI;• AI applications in disease diagnosis, prognosis, treatment, monitoring, and prevention. Topic editors: Shanshan Huang, Weimiao YU, Yongping Lu Source: https://www.frontiersin.org/research-topics/85624/explainable-and-trustworthy-artificial-intelligence-in-precision-medicinemultimodal-data-analysis-and-clinical-decision-support