In the big data era, with the enrichment of data collection and description measures, a wide array of medical data in various formats are much easier to collect. It is significant to discover the knowledge hidden in mass data by comprehensive understanding and learning to realize the medical data intelligence, which can help us in areas such as intelligent decisions and predictive services. However, the high-dimensional, heterogeneous, real-time, and low-quality characteristics of the collected medical data pose great challenges to the design of knowledge discovery methods. If we can effectively perform multi-modal feature learning on massive high-dimensional, heterogeneous, real-time, and low-quality medical big data to discover the hidden knowledge and rules, the potential values and insights can be identified. Thus, it will provide a comprehensive understanding and a favourable decision-making framework based on massive data to realize the real medical data intelligence.

With the development of modern medicine and the subsequent use of data mining, artificial intelligence, and other information technologies, auxiliary diagnosis research is gradually becoming an important part of the field of medicine and informatics. At present, it mainly focuses on two aspects: first, analyse and process the laboratory test data, quickly extract key indicators from a large amount of data through reasoning, analysis, comparison, induction, summary, and demonstration, and draw cognitive conclusions about the patient's physical state and illness. Second, through the analysis and understanding of text, image, video and other multimedia forms of diagnostic data, mining and distinguishing the characteristics of the disease, diagnosis, and evaluation. However, most of the existing studies only consider the single modality of case data and carry out disease feature analysis, extraction, and fusion. Although it can improve the accuracy of the results of different detection methods to a certain extent, the clinical diagnosis of diseases is a complex process. Doctors are difficult to give accurate judgments based on a certain modal result, and it is necessary to comprehensively analyse the high-dimensional, heterogeneous, real-time, and low-quality patient's signs, symptoms, medical detection results and other multimodal information.

Therefore, this Special Issue aims to seek high-quality research papers and reviews from academics and industry-related researchers in the areas of big data, data mining, machine learning, artificial intelligence, and multimedia analysis to present the most recently advanced methods and applications for realizing medical data intelligence.

Potential topics include but are not limited to the following:

• Big data theory and methods for medical data analysis
• Artificial intelligence theory and methods for medical data analysis
• Multimodal analysis of medical data
• Domain adaption and transfer learning for medical data analysis
• Deep learning and reinforcement learning for medical data analysis
• Knowledge graphs for medical data
• Natural language processing for medical data
• Cross-modal index for medical data analysis
• Uncertainty data analysis for medical data
• Data reliability analysis for medical data
• Medical big data analysis and application