In the past 20 years, two magnetic signals have been observed in cuprates materials that cannot be explained by conventional spin-magnetism, but instead by a complex magnetic texture created by microscopic loop currents in the CuO2 planes. This texture inspired from C. M. Varma’s theory could also be at the origin of the Pseudogap opening in cuprates. There two magnetic elastic responses, located respectively at q=0 and q=1/2 positions in reciprocal space, are very weak and often surperimposed with nuclear scattering, making them a challenge to be observed with diffraction techniques. In this context, my thesis focuses on several way to gain information about their microscopic origin: namely i) probing the domain of existence of the q=1/2 one by Polarized Neutron Diffraction, ii) searching for the signature of the q=0 signal in resonnant X ray diffraction and iii) modeling the effect of the loop current pattern  on the band structure of cuprates.