Various special sessions have been organized throughout ISBI 2019 on special topics of interest to the ISBI community.
PET imaging in the era of multi-modality and big data
PET is a unique molecular imaging tool for in vivo investigation of human biological function. This modality is routinely used for diagnosis, staging and prognosis of cancer with 2 million PET scans that are acquired worldwide ever year in a clinical setting. Beyond oncology, PET also plays an important role for exploring brain function in normal and pathological conditions as well as in drug development.
Latest developments include hybrid PET-MR as one of most important technical achievements in clinical imaging of recent years, but it has not yet made the anticipated impact. Hybrid PET-MR scanners struggle to find a truly unique ‘killer’ application despite the huge research investment.
This session will critically discuss current use of PET in its multi-modal and big data aspects. In particular it will explore the use of PET-MR machines, including image reconstruction and analysis as well as applications in pharmacological and clinical studies. We will present an overview of the most recent methodology using large datasets and discuss the possible added value of multi-modal scanners.
Gaia Rizzo, Invicro London (UK)
Andrew Reader, King’s College London
Roger N Gunn, Invicro London (UK)/Imperial College London
Mattia Veronese, King’s College London
Paolo Zanotti Fregonara, Houston Methodist Research Institute
Is Imaging genetics the frontier for precision medicine?
This special session aims at providing an overview on the state-of-the-art of Imaging Genetics. This is a field that embraces the neuroimaging and genetics communities and aims at blending the respective know-how and methodologies within a unified framework pursuing a holistic view of the human being. Concretely, Imaging Genetics refers to the use of anatomical or physiological imaging technologies as phenotypic assays to evaluate genetic variation. Genetic information and neuroimaging data (structural and functional) are integrated within a unified model enabling the assessment of the link between genes and brain structure and function in health and disease, paving the way to multi-modal multi-scale precision medicine. This special session is directed at researchers who wish to develop their knowledge and skills on state‐of‐the‐art developments in the field of neuroimaging genetics.
Marco Lorenzi, INRIA Sophia Antiopolis
Jonas Richiardi, Lausanne University Hospital, Switzerland
André Altmann, UCL
Levi Waldron CUNY Graduate School of Public Health
Johannes Köster, Institute of Human Genetics, University Hospital Essen, University of Duisburg-Essen
Spline Models in Biomedical Imaging
Splines are unifying mathematical objects that allow making the link between the continuous world of theory and the digital world of computers. In the field of biomedical imaging, a whole class of problems is approached with digital algorithms but benefit from a continuous-world representation, which can be efficiently achieved relying on splines. Spline models are advantageous not only computationally, but also conceptually as they allow drawing a number of fundamental connections between disciplines. The design and use of spline representations are traditionally studied by the computer graphics and applied mathematics communities, although being extremely relevant for biomedical image processing. With this special session, we aim at bringing researchers in these fields together with those in biomedical imaging, featuring talks including but not limited to the development of spline and subdivision tools for image segmentation, the exploration of the connection between splines and stochastic processes, and the formulation of novel spline methods for super-resolution microscopy.
Lucia Romani, University of Milano, Italy
Alireza Entezari, University of Florida, FL, USA
Julien Fageot, Ecole Polytechnique Fédérale de Lausanne, Switzerland
Brigitte Forster, University of Passau, Germany
Geometry-based Methods in Biomedical Image Analysis: Junior Researchers
In a number of applications and problems in biomedical imaging, researchers have realized that the data are non-Euclidean in nature, where standard Euclidean operations do not directly apply. Mechanisms to exploit the geometric structure of the data leads to better and richer representations, more accurate algorithms and improved statistical power for various analyses. Riemannian and topological principles can be applied to a broad spectrum of problems in biomedical image analysis, from image registration to region segmentation to object tracking. Recently, a number of research results have been presented describing how the geometry of the data plays a critical role in informing the design of deep learning architectures as well. This session aims to bring together and showcase new results from junior and mid-career investigators in this area. The topic is of broad interest to ISBI attendees since most, if not all, problems related to geometric analysis of biomedical images are squarely within the scope of ISBI.
Pediatric Brain Imaging
The understanding of brain development and the early detection of a number of neurological and psychiatric pathologies have largely benefited from modern advances in MR pediatric acquisitions in the past few years. Indeed, a variety of MR modalities is now available in the context of pediatric neuroimaging and this has opened the possibility to observe brain maturation and development early on, providing crucial information on the dynamics of brain development from birth to adulthood. Nevertheless, pediatric MRI data has brought new challenges in terms of acquisition, preprocessing, analysis, and modelling. This session will present several crucial aspects of pediatric neuroimaging presented by leading investigators in the field. They will cover key topics such as processing of complex neonatal diffusion data, modeling developmental trajectories, relating pediatric neuroimaging data to phenotypic and behavioral information, and general methodological issues in pediatric neuroimaging. The session is aimed at an audience familiar with neuroimaging data analysis interested in the issues specific to pediatric brain imaging.
Jessica Dubois, Inserm, NeuroSpin, Robert-Debré Hospital, Paris, France
Global Health: Imaging in Developing Countries
The impressive advances harnessed by the bio- and medical imaging communities in the last decades have reached all parts of the world having a profound impact on the quality of life of millions of people. While most of the academic and scientific work on these areas have concentrated in first world countries, developing countries have a long and rich history of academic and clinical institutions, which are addressing a variety of healthcare challenges, both general and specific to their regions. Although the academic communities within these countries have been growing and have an increased presence in international conferences and meetings, there is still little interaction between them and those of the first world. A closer relationship between both communities could be of great benefit. On one hand, research communities from developing countries could have access to cuttingedge technology and findings. On the other hand, first world institutions could get a better understanding of these countries’ problematics by learning from local researchers.
Following similar initiatives held in other conferences, this session seeks to provide a space for researchers from both communities to interact by sharing their experiences in medical imaging research for and within developing countries. As clinicians and radiologists are crucial to identify clinical needs and to enable the development of proper data sharing infrastructure, we expect to include at least one clinician in the program. By bringing researchers and clinicians closer, we aim to encourage the creation of a true international academic network in biomedical imaging addressing the problems of developing countries.
Eduardo Romero, Universidad Nacional de Colombia, Bogotá, Colombia
Alison Noble, University of Oxford, UK
Leo Joskowisz, The Hebrew University of Jerusalem, Israel (tentative)
Bhaskar Raj Pant, Hospital for Advanced Medicine & Surgery (HAMS), Nepal