S8 © 2016 IMIA and Schattauer GmbH

Medical Imaging Informatics: Towards a Personalized Computational Patient N. Ayache Inria (French Institute for Research in Computiational Scienes), Sophia Antipolis, France

In 2002, my preface to the IMIA Yearbook Summary was entitled “From Digital Anatomy to Medical Imaging Informatics has become a fast evolving Virtual Scalpels and Image-Guided Ther- discipline at the crossing of Informatics, Computational Sciences, apy”. It was announcing a revolution in and Medicine that is profoundly changing medical practices, for medicine brought by the extensive use of the patients’ benefit. medical image computing to better assist the diagnosis and therapy of the patient. Keywords Today, the promised revolution is here: algorithms in order to reproduce more pre- Medical Imaging Informatics, personalized computational Medical images are omnipresent at the cisely the observed structures and functions patient, computational anatomy, computational medicine hospital, and Medical Imaging Informatics leading to the personalized computational

is required more than ever to exploit their model of this particular patient. Yearb Med Inform 2016;Suppl1:S8-9 flood of information. All around the world, The personalized computational mod- http://dx.doi.org/10.15265/IYS-2016-s002 medical image computing is used to extract el of the patient is then used to provide Published online May 20, 2016 the clinically relevant information from quantitative and objective measurements medical images, and to present this infor- on the patient’s condition to better assess mation in a way that is clinically useful to the diagnosis. It is also used to predict a the physician. This is mainly done through pathological evolution resulting in a bet- the construction of a computational and ter assessment of the prognosis. Finally, personalized model of the patient. the computational model of the patient is Building a computational model of the extensively used to plan and simulate the human body requires dedicated algorithms effect of a therapy, in order to optimize its that take into account a thorough knowledge actual delivery. These three steps - comput- of the human anatomy and physiology. Huge er-aided diagnosis, prognosis, and therapy progress has been made during the last de- - announce the fast development of the cades to describe and simulate the structure computational medicine at the service of and functions of organs thanks to advanced the physician. mathematical, biological, physical, and The tremendous progress of Medical chemical models of the living tissues at Imaging Informatics also accompanies the various scales from the nanoscopic (molec- evolution of normative and reactive medi- ular) to microscopic (cellular), mesoscopic cine towards a more personalized, precise, (tissue), and macroscopic (organic) scales. preventive, and predictive medicine. This Computational models of the human progress relies on numerous algorithmic body rely on a set of parameters that allow, advances in medical image analysis and for instance, to specify the structure and inverse problem solving. It also relies on function of organs. Generic models are continuous advances in the modeling of based on average parameters estimated human anatomy and physiology. It benefits over a population. Confronted to in vivo from the improvement of medical image anatomical and functional images and sig- acquisition techniques, and from the in- nals of a singular patient, those parameters troduction of new imaging modalities at are adjusted by efficient personalization various scales. It is supported by the regular

IMIA Yearbook of Medical Informatics 2016 S9 Medical Imaging Informatics: Towards a Personalized Computational Patient

increase of computer performances in terms proposed solutions. The third vertex rep- Medical Imaging Informatics has become of both computing power and memory resents the industrial partnership required a fast evolving discipline at the crossing storage. Finally it is boosted by the develop- to transform the research prototypes into of Informatics, Computational Sciences, ment of large databases of medical images clinically approved products. These three and Medicine that is profoundly changing made available through the Internet and worlds (academic, clinical and industrial) medical practices, for the patients’ benefit! by the introduction of large-scale machine do not have rigid frontiers; actually their learning methods. interactions are very stimulating and rich In 2014, I had the privilege to be elected in terms of innovations. References to the Chair of “Informatics and Computa- A young and active international com- 1. Nicholas Ayache. Des images médicales au pa- tional Sciences” at the Collège de France. munity of researchers supports the research tient numérique [From medical imaging to the The inaugural lecture entitled “From in Medical Imaging Informatics. These digital patient], Leçons inaugurales du Collège de Medical Images to the Digital Patient”, the researchers share a common vision and have France. Collège de France / Fayard, March 2015 course on “The personalized digital patient: developed a scientific society in 2004 called (in French). http://books.openedition.org/cdf/4085 Images, Medicine and Informatics”, and the MICCAI (for Medical Image Computing 2. Nikos Paragios, Jim Duncan, and Nicholas Ayache. Handbook of Biomedical Imaging: Methodologies associated invited lectures of 26 prominent and Computer Assisted Intervention). Born and Clinical Research. Springer, 2014 colleagues are available on the web site of in 1998 with a few hundreds of researchers, 3. Website of the MICCAI Society: http://www. the Collège de France1. its conference currently gathers each year miccai.org/ These lectures reveal that the research in on a different continent about 1,300 par- Medical Imaging Informatics is supported ticipants from 46 countries. About 40% of by a virtuous triangle. The first vertex the participants are PhD students, which is represents a multidisciplinary academic a good sign of the vitality of this research Correspondence to: research where informatics is associated community. Most of these young doctors Nicholas Ayache with other sciences (mathematics, biology, find a position after their thesis, either in Inria physics, chemistry, anatomy, physiolo- the academic, clinical, or industrial envi- Asclepios Project Team 2004 Route des Lucioles gy, etc.). The second vertex represents a ronment. An increasing number of them 06902 Sophia Antipolis clinical partnership, essential to address participate to the creation or development France relevant medical problems and validate the of start-up companies. E-mail: [email protected]

1 Courses and invited seminars available both in French and English languages at http://www.college-de-france.fr/site/en- nicholas-ayache/

IMIA Yearbook of Medical Informatics 2016