X-ray phase contrast imaging permits to reach nanometric resolution in tomographic imaging with several orders of magnitude higher sensitivity than using the attenuation . The main drawback is that it needs an additional reconstruction step, known as phase retrieval, to yield quantitative images. This reconstruction problem is a non-linear inverse problem. We have previously developed algorithms based on linear approximations to solve this problem . The non-linear problem remains difficult to treat, however. Therefore, the goal of this internship is to use deep learning for phase retrieval from X-ray phase contrast images. In particular, we will investigate various deep learning architectures  and compare them to model-based approaches. Training data sets will be generated using in-house software. Deep learning methods will be implemented using the TensorFlow library in Python. The developed methods will be compared to previously developed algorithms on data from the European Synchrotron Radiation Facility (ESRF).
 R. Mokso, P. Cloetens, E. Maire, W. Ludwig, and J.-Y. Buffière, “Nanoscale zoom tomography with hard x rays using Kirkpatrick-Baez optics,” Appl. Phys. Lett., vol. 90, no. 14, p. 144104, Apr. 2007.
 M. Langer et al., “Priors for X-ray in-line phase tomography of heterogeneous objects,” Philos. Trans. R. Soc. A, vol. 372, p. 20130129, 2014.
 Y. LeCun, Y. Bengio, and G. Hinton, “Deep learning,” Nature, vol. 521, no. 7553, pp. 436–444, May 2015.