Fat quantification for the staging of hepatic steatosis at 11.7 T
Recrutement en cours/passé: 
Recrutement en cours
kevin.tsevekoon@creatis.univ-lyon1.fr helene.ratiney@creatis.insa-lyon.fr benjamin.leporq@creatis.insa-lyon.fr

Scientific background and rationale: Obesity is a major issue in developed countries with a high incidence on life expectancy. Hepatic steatosis is a frequently associated pathology and being able to stage and understand its development can greatly help in reversing it through drugs intake and/or changes in lifestyle as demonstrated in mouse models[1]. Biological techniques for murine hepatic tissue evaluation involve invasive procedures which often lead to the death of the animal. On the other hand, magnetic resonance imaging (MRI) with its numerous available techniques is a proven non-invasive diagnostic tool for soft tissues and in particular the liver [2,3]. Applying these techniques on mice model is challenging due to smaller size andbreathing synchronization. Higher magnetic fields enable attaining high image resolution compatible with smaller sizes but renders staging of steatosis more challenging.

Aim: Initial work have enabled the implementation of a multi gradient echo (MGE) MRI sequence for in vivo liver fat quantification on mouse models at 11.7T (Bruker preclinical MRI). However post-processing of MR images do not yet yield satisfactory quantification results. Compared to previous work found in the literature, working at 11.7T highly constrains the sampling possibilities for the gradient trains and new strategies need to be developed for this field intensity (interleaved or hybridation with MR spectroscopic imaging, echo shift in a spin echo sequence) . This internship will implement new sampling strategies, possibly integrating information from MR spectroscopy acquisition and will look into solving issues pertaining to the MR sequences (reducing breathing artifacts, finding the best sequence parameters). The post-processing algorithm will have to be refined in view of getting robust and reproducible fat quantification data. Experiments will involve phantom and in vivo acquisitionson mouse models. Other biomarkers (fat composition, T2*, susceptibility) that can be extracted from the MGE sequence will also be investigated.

Description of the internship work:1. Investigating and correcting artifacts of the multi gradient echo MRI sequences, and studying new sampling strategies and/or new sequence design2. Data processing for extraction of fat content (PDFF) quantification3. Data processing for fat composition

Skills required: Physics of MRI, signal processing, programming (Matlab)

References[1] : E. Tubbs et. al., Diabetes. Vol.63 p3279 -3294, 0ct. 2014[2] : B. Leporq et. al, NMR in Biomedicine. 2017;30:e3766[3] : A. Nemeth et. al. Journ. Magn. Reson. Imag. Vol49, Issue 6, p1587-1599, Jun. 2019