EMIM 2017 Educational Sessions

Wednesday 5 April from 08:30h to 11:30h

Don't miss to sign-in during your registration!

ES 01 | Basics in RADIOMICS

Organized by Philippe Lambin - Head of the Head of Radiotherapy, University Hospital Maastricht


PROGRAMME overview

08:30h-09:15h   Radiomics : the Basics. Philippe Lambin

09:15h-10:00h   Radiomics : the Process. Ralph Leijenaar

10:00h-10:30h   Coffee Break

10:30h-11:00h   Radiomics with MR : HOW? Stefan Klein

11:00h-11:30h   Challenges & Opportunities; Case & Discussion. All


Medical imaging has been the cornerstone for the management of cancer patients for decades. Imaging data such as CT, MRI or PET are routinely acquired for every cancer patient in the process of diagnosis, treatment planning, image-guided interventions and response assessment. Technological advances in non-invasive imaging have enabled the field to move towards the identification of quantitative non-invasive and actionable imaging biomarkers. Imaging biomarkers could play a crucial role in tumor profiling including the identification of intra-tumor heterogeneity. The development of new therapies requires the identification and development of associated biomarkers to select early responders. There are now more than 80 companion diagnostics on the market and the FDA requires the development of associated biomarkers prior to approval.

The use of image analysis in a quantitative way is now considered as one of the most promising techniques to identify imaging biomarkers. Like genomics aims at identifying genes and gene mutation to characterize tumor or normal tissue, Radiomics looks at the phenotypic expression of genes, which results in particular imaging features or signatures able to characterize tumor and normal tissue.  RADIOMICS is a platform for the high throughput extraction of large amounts (1000+) of quantitative image features such as tumor image intensity, (multi-scale) texture, shape and size extracted from standard medical images (e.g., CT, MR, PET) using (semi)automatic software to be combined in signatures that functions as quantitative imaging biomarkers. These imaging biomarkers will guide personalized cancer treatment.

Our starting point is an overview of the history of Radiomics we then discuss the success stories of Radiomics but also the pitfalls.  Next, we will review the process from data acquisition, access to the DICOM objects, feature extraction, towards machine learning-based analysis and validation. Then we will discuss the opportunities and challenges when using MRI for Radiomics, and in particular the exciting prospect of moving towards quantitative MRI techniques.

In the final part of the tutorial, we discuss the current challenges and directions of research in the field; in particular, the necessity of dealing with large annotated data sets and the crucial aspect of rigorous methodology. Some case will be discussed (e.g. using the Radiomics Quality Score).

ES 02 | Basics of molecular IMAGING PROBES development

Organized by Bert Windhorst, Amsterdam & Twan Lammers, Aachen

PROGRAMME overview

08:30h-09:00h   Chemistry of Optical Probes. Kai Licha

09:00h-09:30h   Chemistry of MR probes. Robert Muller

09:30h-10:00h   Chemistry of PET probes. Bert Windhorst

10:00h-10:30h   Coffee Break

10:30h-11:00h   Complementary imaging probes: Cherenkov, optical/PET, MR/PET. Jan Grimm

11:00h-11:30h   Challenges & opportunities, discussion lead by Fabian Kiessling


Imaging probes are essential components of molecular imaging techniques. In particular for optical, MR and PET molecular imaging they provide the basis on which these techniques can accelerate.

In the first three lectures the basics of chemistry and development of optical, MR and PET probes will be presented. By no means can these contributions be comprehensive. Therefore, in the lectures there will be focus on the essentials of the particular chemistry, the most important aspects in the development of molecular imaging probes and special attention will be on combined use of optical, MR and PET probes.

Next, examples of applications of complementary imaging probes will be presented, optical-PET as well as MR-PET combinations. Either as labels in one probe or by combination of individual probes and based on several compound classes such as particles, biologicals or peptides.

Finally the session will be closed with interactive discussion where challenges and opportunities will be addressed and the audience will actively participate via web-based interaction.




ES 03 | Imaging METABOLISM

Organized by Uwe Himmelreich, Leuven- Head of Biomedical MRI unit &

Jordi Llop, San Sebastian - Head of Radiochemistry & Nuclear Imaging unit


PROGRAMME overview

08:30-09:15h      PET and SPECT tracers to follow metabolic pathways. Jaczek Koziorowski

09:15-09:45h      Metabolic imaging in diabetes. Martin Gotthardt

09:45-10:15h      Coffee Break

10:15-10:45h      MR spectroscopy. Arend Heerschap

10:45-11:15h      MRS of hyperpolarized compounds. Kevin Brindle

11:15-11:30h      Discussion


Molecular Imaging aims to visualize biological processes on a molecular level. This includes the breakdown of molecules (catabolism) and the synthesis of molecules by an organism. Metabolism can be broadly defined as the sum of biochemical processes in living organisms that either produce or consume energy. It is well known that metabolic perturbations accompany common human diseases, and because of this, accurate knowledge on metabolic alterations can provide important information about the health status of an organism and about the success of therapy.

In vivo imaging techniques have emerged as excellent tools to investigate metabolic pathways in a non-invasive way, showing application in different areas of research ranging from the investigation of basic metabolic pathways to underlying metabolic alterations in disease, which enables the early diagnosis of certain pathologies or the evaluation of the response to treatments. This includes not only metabolic diseases like diabetes but also tumor metabolism, neurological diseases, cardiac diseases and others.

In this session, the contribution of proton (1H) Magnetic Resonance Spectroscopy (MRS), MR spectroscopic imaging (MRSI), heteronuclear MRS (including hyperpolarized compounds), Positron Emission Tomography (PET), and Single Photon Emission Computerised Tomography (SPECT) to the investigation of metabolism will be discussed, including practical applications both in the preclinical and clinical settings. The focus will be on applications in metabolic diseases, namely diabetes, and tumor metabolism.

The four educational talks will provide a brief introduction to fundamentals in metabolic imaging. We will cover the description and application of the most widely used PET and SPECT tracers to the investigation of metabolic pathways in different organs (brain, bone, heart) and diseases (infection, inflammation, cancer). For historical relevance, special emphasis will be made on the application of nuclear imaging to the investigation of glucose and fatty acid metabolism. A second talk will focus on the application of metabolic imaging in diabetes, with special consideration on beta cell imaging and metabolic imaging. An introduction to MR spectroscopy will provide an overview of different techniques and parameters. While 1H MRS data acquisition and analysis will form the core of this talk, other techniques like 13C or 31P MRS will also be covered.  A forth talk will focus on the basics and applications of MRS of hyperpolarized compounds to follow metabolic pathways.

Current developments, challenges and opportunities of imaging and quantifying metabolic processes by in vivo imaging methods will be discussed at the end of this educational session.



Organized by Michael Unser, Lausanne - Director of the Biomedical Imaging Group at the EPFL


PROGRAMME overview

08:30h-09:15h   Imaging as an inverse problem

09:15h-10:00h   Classical reconstruction algorithms

10:00h-10:30h   Break

10:30h-11:15h   Modern algorithms: the sparsity revolution

11:15h-11:30h   Challenges & Opportunities; Discussion


A fundamental component of the imaging pipeline is the reconstruction algorithm. In this educational session, we review the physical and mathematical principles that underlie the design of such algorithms. We argue that the concepts are fairly universal and applicable to a majority of (bio)medical imaging modalities, including magnetic resonance imaging and fMRI, x-ray computer tomography, and positron-emission tomography (PET). Interestingly, the paradigm remains valid for modern cellular/molecular imaging with confocal/super-resolution fluorescence microscopy, which is highly relevant to molecular imaging as well. In fact, we believe that the huge potential for cross-fertilization and mutual re-enforcement between imaging modalities has not been fully exploited yet.

                The prerequisite to image reconstruction is an accurate physical description of the image-formation process:  the so-called forward model, which is assumed to be linear. Numerically, this translates into the specification of a system matrix, while the reconstruction of images conceptually boils down to a stable inversion of this matrix. The difficulty is essentially twofold:

(i) the system matrix is usually much too large to be stored/inverted directly, and

(ii) the problem is inherently ill-posed due to the presence of noise and/or bad conditioning of the system.

                Our starting point is an overview of the modalities in relation to their forward model. We then discuss the classical linear reconstruction methods that typically involve some form of backpropagation (CT or PET) and/or the fast Fourier transform (in the case of MRI). We present stabilized variants of these methods that rely on (Tikhonov) regularization or the injection of prior statistical knowledge under the Gaussian hypothesis.  Next, we review modern iterative schemes that can handle challenging acquisition setups such as parallel MRI, non-Cartesian sampling grids, and/or missing views. In particular, we discuss sparsity-promoting methods that are supported by the theory of compressed sensing. We show how to implement such schemes efficiently using simple combinations of linear solvers and thresholding operations. The main advantage of these recent algorithms is that they improve the quality of the image reconstruction. Alternatively, they allow a substantial reduction of the radiation dose and/or acquisition time without noticeable degradation in quality. This behavior is illustrated practically.

                In the final part of the tutorial, we discuss the current challenges and directions of research in the field; in particular, the necessity of dealing with large data sets in multiple dimensions: 2D or 3D space combined with time (in the case of dynamic imaging) and/or multispectral/multimodal information.

To top