
On the 26th of November 2024, the 46th CIBM Breakfast & Science Seminar was given by Antoine Lutti from CHUV-UNIL. The event was chaired by Elda Fischi-Gomez, Research Staff Scientist, CIBM SP CHUV-EPFL Computational Medical Imaging & Machine Learning Section.
Advanced MRI relaxometry methodologies for neuroscience
Abstract
MRI relaxometry data are markers of properties of brain tissue such as myelination or iron concentration. Combined with clinical phenotype, they provide insights into the microscopic processes that underlie brain disease and are used to study a broad range of neurological and psychiatric disorders. The methodologies used for data acquisition and computation are instrumental to enable the detection of small tissue change and to use this data as early markers of disease progression.
In this talk, I will present recent physics-inspired methodological advances that aim to bolster the use of MRI relaxometry data in neuroscience. In particular, I will introduce a framework that optimizes sensitivity in group analyses of brain imaging data with heterogeneous quality and ensures the validity of statistical results. I will also introduce novel strategies that allow the acquisition of data with reduced levels of cardiac-induced noise, leading to ~25% improvements in reproducibility. I will show how the combination of improved data quality and advanced models of the MRI signal allows the computation of MRI markers with increased specificity.
Antoine Lutti,
CHUV-UNIL
About the speaker
My research activities focus on the development of Magnetic Resonance Imaging (MRI) acquisition techniques tailored for neuroscience. My PhD in NMR physics, completed in 2007, involved the development of diffusion measurement techniques for the study of liquid crystals under shear (supervision Prof. PT Callaghan, Victoria University of Wellington, New Zealand).
I transitioned into MRI and brain imaging in 2007, as a senior MRI scientist at the Functional Imaging Laboratory, University College London (UK). It is there that I became an expert developer of MRI acquisition sequences tailored for neuroscience: 3D EPI and multi-echo sequences for enhanced BOLD sensitivity in fMRI, MRI relaxometry and B1-field mapping techniques…. as well as many other features in support of individual neuroscience projects.
I have joined the Laboratory for Research in Neuroimaging (LREN) at Lausanne University Hospital in 2013, as the technical director of a newly established research-dedicated MRI platform. There, the acquisition strategies developed in-house are instrumental to many neuroscience projects such as BrainLaus – the neuroimaging branch of the ‘Colaus’ longitudinal study of the Lausanne population. I am currently a senior lecturer at CHUV-UNIL. In addition to the development of MRI acquisition sequences, my group also focuses on the development of novel MRI markers of brain microstructure and on improving the analysis and computation of relaxometry data.