"Retrieval of Information in Medical Image Databases"

Christian Roux, Ph.D.
Télécom Bretagne – Inserm, France


Abstract

Medical archives are a great asset to develop the medical decision supports of tomorrow. This talk will survey the state-of-the-art of this technical field with a focus on the retrieval of information in medical image databases, present some important results and point towards future research directions. It will be highlighted that, thanks to major advances in information retrieval, and related topics such as security, virtually any medical decision problem can benefit from information stored in medical archives.

During a medical examination, clinicians build a patient record containing all available information about their patient. A promising way to support their decisions is to retrieve similar patient records from a medical archive. Confronted to similar cases, clinicians may confirm or revise their decisions by analogy reasoning. In order to retrieve patient records, two challenges need to be addressed. First, how to characterize complex elements in patient records (images, videos, etc.)? Second, how to combine heterogeneous elements in these records (demographic and clinical data, images, videos, etc.) in order to define clinically-relevant similarity metrics?

Different solutions have been proposed to characterize images in the past. Some rely on a semantic modeling of images (lexical description of lesions, spatial organ modeling, etc.). Others characterize their numeric content, using either application-specific lesion segmentation algorithms or application-independent image features (texture, color or shape features, etc.). Recently, video characterizations have also been proposed. One future challenge will be to characterize and match videos as fast as possible, for real-time decision support, during a video-monitored surgery for instance.

Different approaches have been proposed to define similarity metrics between heterogeneous patient records. Some rely on medical ontologies, others rely on information fusion: early fusion (using data mining techniques) or late fusion (using the theory of evidence). A future challenge will be to take into account the cognitive processes of each clinician in order to adapt the similarity metrics and therefore provide personalized decision supports.

Biography

Christian ROUX (IEEE Fellow, AIMBE Fellow, IAMBE Fellow) received the Agregation degree in physics from the Ecole Normale Superieure, Cachan, France in 1978 and the Ph.D. from the Institut National Polytechnique, Grenoble, France, in 1980. He is with Institut Télécom – Télécom Bretagne since 1982 as an Associate Professor and since 1987 as a Professor. He has been a Visiting Professor with the Medical Image Processing Group, Department of Radiology, University of Pennsylvania (1992-1993), and a Distinguished International Research Fellow with the department of Electrical Engineering, University of Calgary, Canada (1996, 2003). He is the founding Director of the Laboratoire de Traitement de l’Information Medicale, INSERM U650.

His research interests concerns advanced medical information processing, and spatial, temporal and functional information modeling and analysis in medical images, with applications in various medical domains including orthopedics, gastroenterology ophthalmology, cardiology and nuclear medicine.

Dr. ROUX has published around 160 papers, holds nine patents and is the co-founder of three spin-off companies in orthopedic computer-assisted surgery. He was an Associate Editor for the IEEE TRANSACTIONS ON MEDICAL IMAGING and Chairman of its Steering Committee (1993-2003), and is member of the Editorial board of the IEEE TRANSACTIONS ON INFORMATION TECHNOLOGY and of the PROCEEDINGS OF THE IEEE. He is the founding co-chairman of the IEEE EMBS International Summer School held every other year on Berder Island, France, which is now recognized as a worldwide reference. He served as President of the IEEE Engineering in Medicine and Biology Society in 2001. Dr. ROUX played a seminal role in the creation of IEEE ISBI International Symposium on Biomedical Imaging in 2002. In 2004 he became Senior advisor with the French Ministry of Research and in 2006 Program officer with the French National Research Agency in the area of Technologies for Health. Since 2009, Dr. ROUX is Dean for Research with Telecom Bretagne.