Under the AI4EU project Virtech developed the MED-AI solution, an early prototype of a Clinical Decision Support System (CDSS) aimed at enhancing the Centre Hospitalier Universitaire de Liège’s (CHUL) transition towards personalized medicine. This web-based platform focuses on the early detection and diagnosis of COVID-19 and pneumonia by analyzing CT and X-ray images using deep convolutional neural networks. The development explored both custom neural network training and transfer learning approaches. With support from AI4EU and CHUL, the MED-AI solution achieved Technology Readiness Level 7, indicating successful testing and validation in an operational environment.
Similar AI-driven initiatives have been explored in the medical field. For instance, a study developed a triage system using machine learning and natural language processing on teleconsultation records, resulting in a certified system now in use at a major European telemedicine provider. This system interacts with patients via a mobile application, generating personalized questions based on initial symptoms to recommend appropriate care. Such remote guidance has proven fundamental during high-demand situations like the COVID-19 outbreak.
Another study proposed an AI-driven solution for automatic quantification and prognosis assessment of COVID-19 pneumonia using chest CT scans. This approach combined automatic CT delineation of lung disease with data-driven identification of biomarkers to predict short-term outcomes, such as the need for mechanical ventilation. The integration of CT-based biomarkers with clinical variables offers potential for optimal patient management, especially given the shortage of intensive care resources during the pandemic.
These examples underscore the potential of AI-driven solutions like MED-AI in improving diagnostic accuracy and patient management in healthcare settings.