The exponential growth of data in public databases, such as GenBank, underscores the need to consolidate information from multiple sources and formats. The diverse strategies adopted by each database create heterogeneities in schemas, file formats, and identifiers, making direct and efficient integration challenging. The absence of a common semantic model for querying various independent biological databases has driven the community to develop biomedical ontologies. In this context, Knowledge Graphs emerge as a viable strategy to represent information in a semantically rich and organized manner, playing an essential role in overcoming integration and semantic interoperability challenges, enabling applications ranging from protein function identification to clinical applications. This work proposes the conception of a semantic representation for the MIMIC-IV database, aimed at creating its Knowledge Graph. This initiative seeks to integrate MIMIC-IV with other existing biomedical databases, facilitating discoveries and significant insights in the biomedical field. The project also addresses an existing gap in the literature regarding MIMIC-IV by providing a detailed analysis of its semantic representation and its application in biomedical research.