Project

Back to overview

Advanced high-field MR imaging and quantitative image analysis for segmentation of the thalamic nuclei

Applicant Bach Cuadra Meritxell
Number 157040
Funding scheme Project funding (Div. I-III)
Research institution Centre d'Imagerie BioMédicale (CIBM) CHUV
Institution of higher education University of Lausanne - LA
Main discipline Information Technology
Start/End 01.04.2015 - 30.09.2018
Approved amount 244'841.00
Show all

All Disciplines (5)

Discipline
Information Technology
Other disciplines of Physics
Neurophysiology and Brain Research
Biomedical Engineering
Electrical Engineering

Keywords (8)

High-field Magnetic Resonance Imaging; Diffusion weighted MR imaging; Targeting procedures; Validation; Multi-modal imaging; Variational segmentation; Radiosurgery; Human Thalamus Nuclei

Lay Summary (French)

Lead
Les développements récents de l’imagerie par résonance magnétique (IRM) et des méthodes de traitement d'images permettent la visualisation et l'analyse quantitative du cerveau humain in vivo. En dépit de ces avancées technologiques, dans certaines procédures cliniques, comme la neurochirurgie stéréotaxique des troubles du mouvement, la planification repose toujours sur l'IRM classique qui ne permet pas une visualisation directe de la cible, en l’occurrence les noyaux thalamiques. Aujourd'hui, grâce aux champs magnétiques élevés, il est possible d'obtenir un meilleur contraste des tissus et une bonne résolution spatiale. Ainsi, un besoin apparait de réunir les meilleures méthodes d'imagerie et d'analyse quantitative pour la délinéation automatique de noyaux thalamiques, ce qui pourrait être utilisé non seulement dans la neurochirurgie fonctionnelle, mais aussi dans les études en neurosciences, vue l'importance capitale du thalamus dans de nombreux troubles neurologiques.
Lay summary

La principale difficulté de notre projet de recherche est liée à la complexité architecturale présente à l’intérieur du thalamus: structure composée des petits noyaux (de seulement quelques millimètres) connectés à un grand nombre des régions corticales et, en même temps, à d'autres noyaux de la matière grise et des zones du cerveau telles que le cervelet.

 L'objectif principale de notre projet est le développent d'une méthode précise et robuste de segmentation pour la délinéation des différents noyaux thalamiques. L'originalité de notre projet consiste en l'utilisation d’imagerie par résonance magnétique (IRM) de diffusion à 3 Tesla permettant d’estimer, avec une grande résolution angulaire et spatiale, l’architecture des fibres cérébrales. Le deuxième défi de notre recherche consiste à améliorer la visualisation directe des différents noyaux du thalamus par IRM à haut champ (7 Tesla). Nous allons évaluer le potentiel et les limitations de différentes méthodes pour réduire le niveau courant d'artefacts présents dans les images morphologiques du thalamus à 7T. La contribution principale de notre projet consiste en la combinaison des caractéristiques locales de diffusion avec l'information de haut contraste d'intensité à haute résolution. Nous pourrons ainsi comparer les informations provenant de ces différentes modalités et les intégrer dans un cadre méthodologique robuste pour la segmentation des différents noyaux thalamiques afin d'offrir une meilleure compréhension de la cytoarchitecture à l'intérieur du thalamus.

Même si aujourd'hui l’IRM à 7T n'est pas encore disponible dans la pratique clinique, ce projet montrera son utilité, notamment en matière de planification de la radiochirurgie. Ceci pourrait représenter un point d'inflexion pour les procédures de planification chirurgicales ainsi que pour l'analyse quantitative des noyaux thalamiques dans les études en neurosciences.

 

Direct link to Lay Summary Last update: 29.10.2014

Responsible applicant and co-applicants

Employees

Project partner

Publications

Publication
Robust thalamic nuclei segmentation method based on local diffusion magnetic resonance properties.
Battistella Giovanni, Najdenovska Elena, Maeder Philippe, Ghazaleh Naghmeh, Daducci Alessandro, Thiran Jean-Philippe, Jacquemont Sébastien, Tuleasca Constantin, Levivier Marc, Bach Cuadra Meritxell, Fornari Eleonora (2017), Robust thalamic nuclei segmentation method based on local diffusion magnetic resonance properties., in Brain structure & function, 222(5), 2203-2216.
Tracking discrete off-resonance markers with three spokes (trackDOTS) for compensation of head motion and B 0 perturbations: Accuracy and performance in anatomical imaging TrackDOTS-A New Approach for Head Motion and Field Monitoring
Jorge João, Gretsch Frédéric, Gallichan Daniel, Marques José P. (2017), Tracking discrete off-resonance markers with three spokes (trackDOTS) for compensation of head motion and B 0 perturbations: Accuracy and performance in anatomical imaging TrackDOTS-A New Approach for Head Motion and Field Monitoring, in Magnetic Resonance in Medicine, 0.
In-vivo probabilistic atlas of human thalamic nuclei based on diffusion weighted magnetic resonance imaging
NajdenovskaElena, Alemán-GómezYasser, BattistellaGiovanni, DescoteauxMaxime, HagmannPatric, JacquemontSebastien, MaederPhilippe, ThiranJean-Philippe, FornariEleonora, Bach CuadraMeritxell, In-vivo probabilistic atlas of human thalamic nuclei based on diffusion weighted magnetic resonance imaging, in Scientific Data, 1.

Datasets

Dataset Direct Visualization of the Ventral Intermediate Thalamic Nucleus at 7T: Comparison with Current Indirect Clinical Targeting and Existing Automated Segmentation Technique

Author Najdenovska, Elena; Tuleasca, Constantin; Joao, Jorge; Marques, José; Bach Cuadra, Meritxell
Persistent Identifier (PID) https://doi.org/10.5281/zenodo.1438358
Repository Zenodo
Abstract
Data associated to the submitted paper to Scientific Reports - Nature##################################"Direct Visualization of the Ventral Intermediate Thalamic Nucleus at 7T: Comparison with Current Indirect Clinical Targeting and Existing Automated Segmentation Technique"##################################E. Najdenovska*, C. Tuleasca*, J. Jorge, P. Maeder, J.P. Marques, T. Roine, D. Gallichan, J.-P. Thiran, M. Levivier, and M. Bach Cuadra*Equally contributed authorsCopyright (c) - All rights reserved. University of Lausanne. 2018.To reproduce the analyses presented in the referred study, in this repository you could find the MR images acquired from nine young healthy subjects (YS1-YS5), four elderly healthy subject (ES1-ES4) and two drug-resistant tremor patients treated treated with Vim radiosurgery by Gamma Knife (P1 and P2).The provided dataset includes the following NifTI files:- MPRRAGE @3T- DWI @3T (together with the corresponding bvals and bvecs)- MP2RAGE @7T- SWI @7T- binary masks of the manual delineation of both left and right Vim respectively that were done on the SWI (as NifTI files as well).Additionally, for the young cohort (YS1-YS5) we include as well the images used for building the quadrilateral of Guiot:- T2-w @3T- T2 CISS @3TFor the patients (P1 and P2), a follow-up MPRAGE (acquired at 3T) with Gadolinium enhancement is also provided.--------------

Dataset In-vivo probabilistic atlas of human thalamic nuclei based on diffusion weighted magnetic resonance imaging

Author Najdenovska, Elena; Alemán-Gómez, Yasser; Bach Cuadra, Meritxell
Persistent Identifier (PID) https://doi.org/10.5281/zenodo.1405484
Repository Zenodo
Abstract
This is the dataset related to the paper "In-vivo probabilistic atlas of human thalamic nuclei based on diffusion weighted magnetic resonance imaging" submitted to Scientific Data - Nature, by E. Najdenovska*, Y. Aléman-Gómez*, G. Battistella, M. Descoteaux, P. Hagmann, S. Jacquemont, P. Maeder, J.-P. Thiran, E. Fornari and M. Bach Cuadra (*Equally contributed authors).We provide NifTI-1 files representing a digital atlas of seven thalamic subparts per hemisphere. More precisely, the files include the spatial probabilistic atlas maps for each thalamic subpart (Thalamus_Nuclei-HCP-4DSPAMs.nii.gz) and the maximum likelihood atlas (Thalamus_Nuclei-HCP-MaxProb.nii.gz) in MNI space. The region corresponding to each labeled thalamic part respectively is given in the look-up table Thalamic_Nuclei-ColorLUT.txt. The NIFTI files can be visualised with the main available tools such as tkmedit, freeview or 3D-Slicer.We also provide a step by step pseudo code for creating the atlas.This work is currently in restricted access for revision purposes. The work will be publicly available under CC BY license after acceptance of the paper.

Collaboration

Group / person Country
Types of collaboration
CIBM-MR CHUV core - Lausanne University Hospital Switzerland (Europe)
- Research Infrastructure
Prof. Jean- Marie Régis, Aix-Marseille University France (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
- Research Infrastructure
Prof Jean-Philippe Thiran, EPFL Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
J.P. Rebelo Ferreira Marques, Donders Institute Netherlands (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
Prof Dimitri Van De Ville, EPFL Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication

Scientific events

Active participation

Title Type of contribution Title of article or contribution Date Place Persons involved
NeuroCoG Seminar Series: Advanced Methods for brain exploration Individual talk Magnetic resonance image analysis of thalamic nuclei at high and ultra-high fields 26.10.2018 Grenoble, France Najdenovska Elena; Bach Cuadra Meritxell;
Congres of neurological surgeons Talk given at a conference Ventro-Lateral Motor Thalamus Abnormal Connectivity in Essential Tremor Before and After Stereotactic Radiosurgical Thalamotomy: A Resting-State fMRI Study 06.10.2018 Houston, Texas, USA, United States of America Najdenovska Elena; Bach Cuadra Meritxell; Levivier Marc;
XXIII Congress of the european society for stereotactic and functional neurosurgery Talk given at a conference Ventro-lateral motor thalamus abnormal connectivity in essential tremor before and after thalamotomy: a resting-state fMRI study 26.09.2018 Edinburgh, Scotland, Great Britain and Northern Ireland Najdenovska Elena; Bach Cuadra Meritxell; Levivier Marc;
26th Annual Meeting & Exhibition of the International Society for Magnetic Resonance in Medicine Talk given at a conference Improved susceptibility-weighted imaging of the thalamic nuclei at 7T with enhanced contrast and venous exclusion 16.06.2018 Paris, France Levivier Marc; Bach Cuadra Meritxell; Maeder Philippe; Najdenovska Elena; Forjaco Jorge Joao Pedro;
Swiss congress of Radiology 2018 Talk given at a conference Improved non-invasive susceptibility-weighted MR imaging of the thalamic nuclei at 7 Tesla with enhanced contrast and venous vessel exclusion 10.05.2018 Lausanne, Switzerland Forjaco Jorge Joao Pedro; Bach Cuadra Meritxell; Levivier Marc; Najdenovska Elena; Maeder Philippe;
Center for Psychiatric Neurosciences (CPN) Scientific seminars Individual talk Segmentation of Thalamic Nuclei in Magnetic Resonance Imaging 12.10.2017 Cery, Lausanne, Switzerland Najdenovska Elena; Bach Cuadra Meritxell;
20th International Conference on Medical Image Computing and Computer Assisted Intervention 2017 Talk given at a conference Towards and Automated Segmentation of the Ventro-Intermediate Thalamic Nucleus 10.09.2017 Quebec, Canada Najdenovska Elena; Maeder Philippe; Levivier Marc; Forjaco Jorge Joao Pedro; Bach Cuadra Meritxell;
13th International Stereotactic Radiosurgery Society Congress Talk given at a conference Vim’s Anatomical Landmarks: Indirect Targeting vs. Direct Visualization at 7T MRI 29.05.2017 Montreux, Switzerland Maeder Philippe; Bach Cuadra Meritxell; Najdenovska Elena; Forjaco Jorge Joao Pedro; Levivier Marc;
13th International Stereotactic Radiosurgery Society Congress Talk given at a conference Exploring Local Diffusion MRI Properties for Vim Localization: Evaluation in Clinical Cases 29.05.2017 Montreux, Switzerland Najdenovska Elena; Levivier Marc; Maeder Philippe; Bach Cuadra Meritxell;
Séminaires groupe Visages, INRIA Individual talk Looking for the VIM: Multi-Modal MR Image Processing at High Field for Thalamic Nuclei Segmentation 14.02.2017 Rennes, France Bach Cuadra Meritxell;
2nd Swiss Medical Image Computing Day Talk given at a conference Robust Thalamic Nuclei Segmentation Method Based on Local Diffusion Magnetic Resonance Properties 24.11.2016 Bern, Switzerland Maeder Philippe; Najdenovska Elena; Bach Cuadra Meritxell; Levivier Marc;
22nd Meeting of the Organization for Human Brain Mapping Poster Drug-resistant Parkinson’s disease vs essential tremor: probing tremor network by resting-state fMRI 26.06.2016 Geneva, Switzerland Bach Cuadra Meritxell; Maeder Philippe; Levivier Marc; Najdenovska Elena;
22nd Meeting of the Organization for Human Brain Mapping Poster Localization of the Ventro-Intermediate Thalamic Nucleus Using Local Diffusion Properties 26.06.2016 Geneva, Switzerland Levivier Marc; Bach Cuadra Meritxell; Najdenovska Elena; Maeder Philippe;
24rd Annual Meeting & Exhibition of the International Society for Magnetic Resonance in Medicine Poster Head motion tracking and correction using discrete off-resonance markers (trackDOTS) for high-resolution anatomical imaging at 7T 07.05.2016 Singapoure, Singapore Forjaco Jorge Joao Pedro;
23rd Annual Meeting & Exhibition of the International Society for Magnetic Resonance in Medicine Poster Novel Robust Segmentation of the Thalamic Nuclei – Validation on Healthy Subjects and Patients 30.05.2015 Toronto, Canada Levivier Marc; Bach Cuadra Meritxell; Najdenovska Elena; Maeder Philippe;


Communication with the public

Communication Title Media Place Year
Media relations: print media, online media Des rayons gammas contre les tremblements Horizons Western Switzerland 2018

Awards

Title Year
Magna Cum Laude distinction for the paper "Improved susceptibility-weighted imaging of the thalamic nuclei at 7T with enhanced contrast and venous vessel exclusion" presented at the Joint Annual Meeting ISMRM-ESMRMB 16-21 June 2018 2018

Use-inspired outputs

Associated projects

Number Title Start Funding scheme
182602 Advanced super-resolution reconstruction methods for quantitative magnetic resonance imaging of the developing fetal brain 01.04.2019 Project funding (Div. I-III)
185872 Uncovering the interplay of structure, function, and dynamics of brain connectivity using MRI 01.06.2020 Ambizione

Abstract

In the last years, impressive developments of Magnetic Resonance Imaging (MRI) techniques and advanced image processing methods have been devoted to the visualization and quantitative analysis of the human brain in vivo. The marriage of these techniques could lead to robust and accurate treatment planning, computer assisted diagnosis and intervention tools. Despite these technological advances, in some clinical procedures, like stereotactic neurosurgery for movement disorders, the planning is still based on classical MR imaging that does not allow a direct visualization of the target, the ventro-intermediate thalamic nucleus. Today, thanks to the increased signal available with the most recent high magnetic field clinical and research MR scanners, it is possible to obtain higher tissue contrast and improved spatial resolution. Thus, there is an increasing opportunity to bring together this advanced imaging and computer-aided methods to improve thalamic nuclei clustering that could be used not only in functional neurosurgery targeting but also in neuroscience studies, since the thalamus is of key importance in many neurodevelopmental and neurodegenerative disorders.The major challenge of our research is related to high complex architecture within the thalamus, made of small nuclei of a few mm only connected to many cortical regions and, at the same time, to other grey matter nuclei and brain areas like the cerebellum. Interesting efforts have been made towards the automated segmentation of thalamic nuclei segmentation, mostly based on diffusion weighted MR imaging. However all these methods face major drawbacks such as low spatial resolution (about $2mm$ isotropic) and coarse diffusion features. Moreover, the existing approaches are usually tested in few subjects, they lack of ground-truth, and they simply proceed to visual assessment with an anatomical atlas. Thus, none of these approaches has been tested in clinical setting, such radiosurgery targeting.An alternative to exploring diffusion information, would be to perform the segmentation using high-resolution structural imaging. In this sense, various studies have put efforts in tuning the MR sequence acquisition parameters of T1-weighted and T2*-based imaging both at 3T and 7T to optimize the contrast between different thalamic nuclei while allowing up to 0.35 mm in-plane resolution in a reasonable scanning time. Most of these developments have shown an increased ability to directly visualize some thalamic nuclei. However, these observations are so far done in a qualitative manner on single subject evaluations and a more systematic evaluation using automated methodologies is missing. Yet, despite the richness of the information obtained at ultra high fields, images acquired at long echo times are prone to artifacts such as motion that make the obtained data quality not reproducible. This problem will be even more severe in the context of patients with movement disorders such as those needing stereotactic neurosurgery.The primary aim of this project resides in developing a variational segmentation framework for thalamic nuclei segmentation, providing significantly improved parcellation method as compared to state-of-the-art techniques. The originality of our method resides in the use of the complete orientation distribution function (ODF) from diffusion MR images at 3 Tesla at high spatial resolution (1 mm isotropic). We will promote the use of spherical harmonics for the ODF representation, providing both angular and shape representation of diffusion at each voxel. A second major aim of our research is dedicated to improve direct thalamic nuclei visualization at 7T. We will evaluate the potential and limitations of different methods to reduce the level of artifacts in morphological imaging. Different approaches will be probed to correct for artifacts of physiology and of motion origin. A major contribution of this project will be the combination of local diffusion proprieties and high-resolution high-contrast structural information. It will allow to compare the two information sets from a segmentation perspective and observe its amount of overlap. Bringing the two modalities into a new unified segmentation will offer new insights into the cito- and myelo- architectonic arrangement of its nuclei. Even if today 7T MR is not yet routinely used in clinical practice, this project will show its clinical usefulness, especially applied to radiosurgery planning. We believe this combination of features will achieve improved segmentation of thalamic nuclei and could represent a breakthrough for current therapy planning procedures as well as for the quantitative analysis of thalamic nuclei in neuroscience studies.
-