PhD thesis defended, on December 19th, 2013 at the University of Montpellier (France), in cotutelle doctoral program with the University of Béjaia (Algeria). This program is funded by the European Commission. Title: Experimental studies and modeling of contrast agent distribution dynamics in NMR imaging Specialty: physics
Download the Thesis in pdf format
Download from https://tel.archives-ouvertes.fr
Non destructive studies of physiological processes in agronomic products require increasingly higher spatial and temporal resolutions. Nuclear Magnetic Resonance (NMR) imaging is a completely non-invasive technique providing access to several types of variables (tissue architecture, spatial variability of the composition, external and internal flow during fruit growth) more difficult to quantify with conventional destructive methods. One of major challenge lies in the ability to spatially localize the physiological and the morphological changes in the agricultural products. The main objective of the research work in this thesis is to carry out a methodology in order to calculate and analyze quantitative NMR imaging applied to agronomy. The implementation, optimization and validation of the FLASH imaging sequence is performed in combination with innovative biocompatible contrast agents efficient in terms of relativity which allow to map in vivo relaxation parameters and then to explore water transport in an agronomic model : the tomato during its growth. Nanoparticles of contrast agent have been used as markers to localize the water flow in the fruit. The choice of the FLASH imaging sequence is motivated by the necessity to achieve sufficient high temporal resolution for monitoring the dynamics of physiological changes related to the water transport. The validation of the T1 calculation method performed on a phantom shows a good agreement compared to T1 measured by relaxometry. An experimental investigation relating to the evaluation of the B0 and B1 magnetic fields inhomogeneities was proposed. A systematic procedure for the estimation of the signal to noise ratio on the parametric images is also proposed which ensures a carefull determination of the intrinsic parameters of living tissues (M0 and T1) and their uncertainties. This step in the analysis ensures reliable interpretation of NMR images and permits image segmentation in order to precisely localize the tissues where there is a high cellular activity. Finally, the time dependance and the compartmental modeling allowed us to quantify the kinetic parameters associated with the water transport in the fruit. Keywords : Quantitative NMR imaging, intrinsic parameters, segmentation, compartmental modeling, contrast agents, plant tissues Download the pdf version