Computational Medicine

The ambition attached to this priority is to create explanatory and predictive mathematical models in biology and medicine. These models have to be based on multi-disciplinary contributions, in order to better understand human biology, but also to be able to diagnose, design, implement and optimize new therapeutic means. A few important categories of pathology including cancer, cardiovascular diseases and neurodegenerative and nervous system diseases will be studied by the Institute’s teams. They will design complex models that will enable them to observe and thus inverse the models in a data assimilation process; they will also develop medical and pharmacological technologies from customized biological models in order to predict change or help guide a therapy protocol or surgical intervention. In partnership with biologists, physicians, chemists and physicists, INRIA researchers will address, most particularly, problems relating to imaging, modeling, calculation and simulation associated with these objectives.

A first point is to develop a complete signal processing and image interpretation algorithm, in order to acquire anatomic and functional data with the best possible spatial and temporal resolutions from intracellular imaging, MRIs, magneto encephalographies, etc., by merging biomedical signals from various origins. A quantitative interpretation will allow the acquisition and measurement to be guided more accurately, thus aiding the diagnosis or therapeutic action.

« In partnership with biologists, physicians, chemists and physicists, INRIA researchers will address, most particularly, problems relating to imaging, modeling, calculation and simulation. »

Data modeling and assimilation will focus on processes in things ranging from cells, organs, complex functions, and even entire organisms. They will include both anatomical and physiological representations and heterogeneous data. Solving inverse problems using estimation, learning and optimization techniques is very difficult in terms of algorithms. Several types of models will likely be developed for one specific organ or function, depending on the application.

An anatomy and an algorithmic physiology will be developed. The latter must identify, for a given organ, the normal variabilities between individuals, must distinguish pathological deviations, and be able to detect these deviations in the images and medical data. The former will enable individualized physiological models to be produced, which will explain and predict functional properties; the physiopathologies will allow therapies to be customized.

These algorithmic researches in imaging and modeling will be associated with virtual reality, surgical robotics and augmented reality research. In this field of application, INRIA will pay special attention to questions of ethics and rules of professional conduct.