Workshop on Spectral Analysis in Medical Imaging (SAMI)

We gladly announce the workshop for Spectral Analysis in Medical Imaging (SAMI), which is held in conjunction with the conference on Medical Image Computing and Computer Assisted Interventions (MICCAI) in Munich, Germany. This workshop aims at presenting leading methods and applications for spectral analysis in medical imaging. At the core of these techniques is the spectral decomposition, also known as eigenvalue decomposition, which finds a wide range of applications in image processing and analysis. The sequence of eigenvalues (the spectrum) and eigenfunctions have properties often exploited for image segmentation, mesh matching, classification or reduction of complex data. In this workshop, we will cover theoretical contributions to spectral analysis, as well as exciting applications of spectral methods in medical image processing and analysis. Such applications include but are not limited to:  

  • Spectral Shape Analysis

  • Spectral Image and Mesh Segmentation & Matching

  • Spectral Classification

  • Spectral Dimensionality Reduction

  • Spectral Graph Methods
  • Any other topic related to Spectral Decompositions & Dimensionality Reduction

This workshop will provide an invaluable opportunity for researchers to present recent work on spectral analysis in medical imaging. Interactions among fellow researchers will foster a deeper understanding of the underlying concepts and will provide insights on the benefits of incorporating spectral analysis into medical imaging applications. Researchers with no or little prior experience in spectral analysis can especially profit from the introductory talks that will cover basic background and give an overview of the field.

The workshop takes place on Monday, October 5, 2015 (8:30-12:00AM) at Forum 12, Holiday Inn. (map below)

** Please Submit Final Article to the MIDAS journal: **

Preliminary Program

08:30 - 09:20 Introduction to spectral methods in medical imaging (M. Reuter, C. Wachinger,  H. Lombaert)
09:20 - 10:00 Oral session I

Tensor Spectral Matching of Diffusion Weighted Images

Eliza Orasanu, Andrew Melbourne, Marco Lorenzi, Marc Modat, Herve Lombaert, Zach Eaton-Rosen, Nicola J. Robertson, Giles Kendall, Neil Marlow and Sebastien Ourselin

The graph windowed Fourier transform: a tool to quantify the gyrification of the cerebral cortex
Hamed Rabiei, Frédéric Richard, Muriel Roth, Jean-Luc Anton, Olivier Coulon and Julien Lefèvre
10:00 - 10:30  Coffee Break
10:30 - 11:20 Invited talk: Daniel Cremers     
11:20 - 12:00 Oral session II

The Bag of Words Approach for the Classification of Head and Neck Cancers Using Diffuse Reflectance Spectroscopy
Ahmed Karam Eldaly, Yannick Benezeth, Virginie Flaus, Christian Duvillard, Alexis Bozorg Grayeli and Franck Marzani

Transitively Consistent and Unbiased Multi-Image Registration using Numerically Stable Transformation Synchronisation
Florian Bernard, Johan Thunberg, Andreas Husch, Luis Salamanca, Peter Gemmar, Frank Hertel and Jorge Goncalves


Martin Reuter

Assistant Professor of Radiology

Assistant Professor of Neurology

Harvard Medical School, Athinoula Martinos Center for Biomedical Imaging

Charlestown, USA

Christian Wachinger

Massachusetts Institute of Technology, Computer Science and Artificial Intelligence Laboratory

Cambridge, USA

Hervé Lombaert

Microsoft Research - Inria Joint Centre, Inria Asclepios Team

Sophia Antipolis, France

Advisory Panel

David Germanaud, NeuroSpin

Leo Grady, HeartFlow

Ender Konukoglu, Harvard Medical School

Julien Lefèvre, Université Aix-Marseille

Jean-François Mangin, NeuroSpin

Diana Mateus, Technische Universität München

Washington Mio, Florida State University

Marc Niethammer, University of North Carolina at Chapel Hill

Robert Pless, Washington University

Kilian Pohl, SRI

Yonggang Shi, University of Southern California

Martin Styner, University of North Carolina at Chapel Hill

Important Dates

June 29th: Paper submission deadline

July 15th: Acceptance notification

August 1st MICCAI Early bird registration deadline

September 15th: Camera-Ready paper deadline for proceedings
October 5th: Workshop day (8:30-12:00am)

Paper Submission

Submissions will be accepted in the LNCS format, with 8 to 10 pages, single column (Latex package here). Accepted and revised papers will be published as post-proceedings at the MIDAS journal. The review is single-blind, so please list the authors with their affiliations on the submission. Reviewers will be chosen among experts in the medical image computing community. The most original contributions will be selected for presentations.

** Submission website is closed: **
** Submit final version to the MIDAS journal: **

Martin Reuter <mreuter at>

Christian Wachinger <wachinger at>

Hervé Lombaert <herve.lombaert at>