The internal distribution of refractive indices (RIs) of a living cell is much more complex than usually admitted in multi-shell models. The reconstruction of RI maps from single phase images has rarely been achieved for several reasons: (i) we still have very little knowledge of the impact of internal macromolecular complexes on the local RI and (ii) phase changes produced by light propagation through the sample are mixed with diffraction effects by internal cell bodies. We propose the implementation a 2D wavelet-based contour chain detection method to distinguish internal boundaries thanks to their greatest optical path difference gradients. These contour chains correspond to the highest image phase contrast and follow the local RI inhomogeneities linked to the intracellular structural intricacy. Their statistics and spatial distribution are morphological indicators for distinguishing cells of different origins and to follow their transformation in pathologic situations. We use this method to compare non adherent blood cells from primary and laboratory culture origins, in healthy and pathological situations (chronic myelogenous leukaemia). In a second part of this presentation, we concentrate on the temporal dynamics of the phase contour chains and we discuss the spectral decomposition of their dynamics in both health and disease.Read less <