Matthieu Guerquin-Kern
Data
Since September 2011, I am lecturer at
ENSEA, a french
graduate school in electrical engineering and computer science.
You can contact me these ways:
Research
I have been working on wavelet-based methods for the solution of inverse problems with application to MRI.
Here is a summary of this project in collaboration with Prof. Pruessmann at ETHZ.
Comparison of reconstructions in MRI |
Deconvolution of a 3-D fluorescence microscopy dataset |
|
|
|
Classical linear |
Wavelet-regularized |
Left: measured data, Right: deconvolution |
In the following, some news and documents I share on that topic:
- August 2012: I converted my thesis manuscript in a website, available here.
- July 2012: My Ph.D. public defense took place on July, 6th.
- April 2012: My Ph.D. private defense took place on April, 26th. The jury members were: Prof. Pierre Vandergheynst (president), Prof. Michael Unser (thesis advisor), Prof. Jalal Fadili, Prof Klaas Pruessmann, and Prof. Dimitri Van de Ville.
- October 2011: Paper on realistic analytical MRI simulations published in IEEE Transactions on Medical Imaging. The code is available here.
- April 2011: Paper on wavelet-based MRI published in IEEE Transactions on Medical Imaging.
- February 2011: As part of a work on analytical phantoms for parallel MRI, I made PDF [2215 B], SVG [2169 B] and SVGZ [794 B] versions of the famous Shepp and Logan phantom. These vectorial formats seem appropriate for analytical phantoms. Notes: parameters are based upon the default in Matlab.
- January 2011: Conference paper accepted for the ICASSP 2011, on an acceleration strategy for reconstruction algorithms tailored to parallel MRI and bioluminescence tomography.
- February 2010: Conference abstract presented at the joint annual meeting of the ISMRM/ESMRMB, on a new wavelet-based algorithm for MRI reconstruction that is optimized for speed and performance.
- January 2010: Conference paper accepted for ISBI on analytical generation of multichannel MR data for the 2-D Shepp-Logan brain phantom. The software is accessible on this page. Here are the presentation slides [PDF: 922 KB]
- April 2009: Matlab/Octave demo [M-FILE: 12 KB] that shows the interest of non-linear regularization in the case of 1-D MRI. The a priori imposed by the l1 regularization term allows perfect extrapolation of the missing k-space samples in the case of noiseless measurements.
- April 2009: Short reminder [PDF: 272 KB] on IRLS solutions (also known as half-quadratic minimization, see this paper for instance) to non-linear reconstruction.
- March 2009: Conference paper presented at the ISBI 2009: Wavelet-Regularized Reconstruction for Rapid MRI.
- July 2008: Introductive presentation to MRI physics [PDF: 3.2 MB].
- July 2008: Personal recap [PDF: 376 KB] of the NMR detection principle.
- December 2007: Short review [PDF: 1.4 MB] of MRI basics and pMRI technniques.
- November 2007: Matlab scripts [ZIP: 20 KB] for the illustration of pMRI approaches.
- November 2007: Matlab scripts [ZIP: 8 KB] for the simulation of a NMR experiment according to the empirical Bloch equations.
List of publications.
Student projects
Here are some project that I propose(d) for bachelor or master students (some links may be accessible only from the EPFL):
Teaching
I am teaching assistant for the bachelor courses 'Signaux et systemes' of Profs. Unser and Vandergheynst. I share here some additional demo and reminders for the course.
- Reminder on discrete filters [PDF: 291 KB, in french]: transfert function in z, stability, causality, examples. Last update: July 2010.
- Matlab demo [ZIP: 444 KB] on the different Fourier transforms CTFT, DTFT, DFT and their relations.
- Short Matlab demo [M-FILE: 4 KB] on the interest of linear phase filtering.
- Matlab demo [ZIP: 96 KB] on the important role of phase in the Fourier transform of images.
Miscellaneous
How to?
ImageJ Plugins
© 2012 EPFL • • 27.08.2012