In prehospital and home care settings, early and accurate detection of acute medical conditions is critical for rapid treatment and improved patient outcomes. Our research group, the Care@Distance, focuses on remote and prehospital digital health, aiming to enhance healthcare delivery through advanced and interoperable technologies. Our work includes clinical decision support systems, artificial intelligence (AI), modern IT solutions and innovative user interactions. Stroke, trauma and falls are among the most common emergencies that require rapid assessment. However, current methods often rely on limited data and subjective evaluations, which may delay critical care. This project aims to address these challenges to enhance diagnostic accuracy and optimize resource use in emergency scenarios. Generation of synthetic data that can act as a complementary data source will also be investigated. The synthetic data may increase the amount of accessible relevant data while maintaining integrity and safety of individuals and patients as the traceability to original data is low. The objective of this project is to develop advanced algorithms for the detection and assessment of acute medical conditions, including stroke, trauma and falls in prehospital and home care settings. These algorithms aim to enhance early diagnosis and decision-making to improve patient outcomes and optimize resource use in emergency scenarios. The project will use a diverse range of data, including tabular data from different sources. Currently, the Swedish Stroke Registry will be used, and other registries such as the Swedish Trauma Registry, Norwegian Trauma Registry and open-source datasets like the Medical Information Mart for Intensive Care (MIMIC) will be used. Additionally, video data will be incorporated as another data source to improve the models’ predictive capabilities. These datasets, which include sensitive information from both healthy individuals and patients, will also be used to generate synthetic data to further develop the models and enhance the robustness and generalizability of the algorithms across diverse patient populations and real-world scenarios. For building realistic scenarios to test developed models and algorithms, several types of synthetic data are needed, including patient health data, sensor data, images etc.