A digital patient model as the basis for personalized treatment

The goal of the MED²ICIN lighthouse project is to deliver enhanced disease prevention, diagnostics and treatment with point-and-click speed and simplicity. Seven Fraunhofer institutes are currently implementing and evaluating a prototype patient model for inflammatory bowel disease (IBD) at Frankfurt University Hospital.

Patient data, such as medical histories, MRI scans, laboratory tests and courses of treatment are increasingly captured and stored in digital form. However, these data are often unstructured and inaccessible. Seven Fraunhofer institutes are tackling this issue within the scope of the MED²ICIN lighthouse project. “The prototype digital patient model takes us into new territory in healthcare,” states Dr. Stefan Wesarg, MED²ICIN project coordinator and Head of the Visual Healthcare Technologies Competence Center at Fraunhofer Institute for Computer Graphics Research IGD.

MED²ICIN combines all healthcare information related to a particular patient and compares this with parameters from population studies and with data for specific diseases, e.g., diagnoses, medication and treatments for other individuals. Consideration is also given to clinical guidelines and healthcare economics. The result is a complete digital patient model.

Frankfurt University Hospital is already working with the digital model. Under the leadership of Dr. med. Frank Behrens, its capabilities are, for instance, being used and evaluated for chronic inflammatory bowel disease (IBD). The model is based on data for 170 parameters for more than 600 patients. “AI helps us to understand each patient in all their complexity, and to identify the best treatment for individuals living with chronic diseases,” explains Behrens. The first tests under real-world conditions are very promising. The digital model saves physicians an enormous amount of time, as all medical data are stored centrally. Moreover, it allows data analysis and user-friendly presentation. One proposal is to use artificial intelligence to compare images taken of a patient’s bowel over time in order to identify changes.

Collaboration with Frankfurt University Hospital gives the Fraunhofer institutes direct feedback from medical professionals. As a result, the software solution can be tailored to the needs of the experts who will employ the system in clinical practice. The digital patient model lends itself particularly well to chronic diseases but can be adapted for many use cases. At a later stage, the project will be extended to specialist doctors with their own practices. The plan is also to subsequently grant access to patients themselves, as well as to research organizations and health insurers. To market the solution to these latter two target groups, the researchers at Fraunhofer intend to form partnerships with life-science companies and healthcare IT providers.

“The digital patient model has enormous potential for improved treatment of the individual,” highlights Dr. Wesarg. It will also permit more targeted healthcare spending for society as a whole—an issue of central importance in light of the challenges associated with demographic change. Technology-driven innovations such as those pursued by the Fraunhofer MED²ICIN project will help to reconcile cost efficiency with the best possible patient treatment. The MED²ICIN project goes much further than existing digital healthcare projects, such as electronic patient records or hospital information systems (HISs), as MED²ICIN draws on data from similar cases of the same disease.

MED²ICIN processes data and visualizes them in a modular dashboard. This interface is designed to be intuitive and can be configured to the specific needs of the user. Also, an intricate 3D model of the human body and its organs boasts a degree of detail that far exceeds that of conventional 3D models. The dashboard is a powerful tool for medical practitioners, helping them to draw the right conclusions and make the best treatment decisions.

MED²ICIN is being developed in strict compliance with the European Union’s General Data Protection Regulation. All sensitive data are pseudonymized, preventing information being traced back to an individual patient. Following the first successful tests, the model is now to be developed further. Fraunhofer is also actively looking for an IT partner who could implement the solution at hospitals.