Nuclear Magnetic Resonance (NMR) has become an inescapable technique for spectroscopic identification. Its main advantage comes from the sensitivity of NMR active nuclei embedded in a molecule to their chemical environment. NMR is also used daily in medical imaging. Magnetic Resonance Imaging (MRI) is not only remarkably versatile, but has the precious advantage of being non-invasive; moreover, the range of radiofrequency used implies that MRI deposit a limited amount of energy in tissues under investigation. Nevertheless, compared to other spectroscopic methods, NMR suffers from a relative lack of sensitivity. Indeed as the NMR transitions are low in energy, the difference of populations between the levels involved, known as polarization, is extremely low. The NMR signal, which is directly proportional to this polarization, is thus many orders of magnitude inferior to the theoretical maximum at full polarization. Dynamic Nuclear Polarization (DNP) allows one to circumvent this disadvantage by transferring the high electron spin polarization to the nuclear spins. This transfer happens via microwave irradiation under optimized conditions. The method has been constantly developed since the ‘fifties. A substantial breakthrough was achieved in 2003 by Golman, Ardenkjaer-Larsen and their collaborators. They proposed to dissolve a sample that has been hyperpolarized at low temperature (at about 1 K) in order to inject it into an animal, or, in fine, into a human patient, and to follow its bio-distribution and eventual metabolic conversion by MRI. This technique, called Dissolution-DNP (D-DNP), allows a signal increase on the order of 10'000, opening the way to many new experimental possibilities. Research in Dissolution-DNP was largely oriented toward the optimization of 13C polarization because of its long lifetimes. In the course of this Thesis, an alternative way that takes advantage of the proton (1H) polarization will be explored. Protons have the advantage that they can be polarized to a higher levels in a shorter time compared to 13C. Unfortunately, once ejected from the polarizer, in solution and at room temperature, the high proton magnetization will be short-lived compared to 13C. Along the chapters of this Thesis, different approaches will be proposed to maximize the advantages of 1H polarization, while minimizing its inconveniences. It is possible to use a standard NMR technique, known as Cross-Polarization (CP), to transfer the abundant magnetization of hyperpolarized protons to other nuclei like 13C, using suitable radiofrequency pulse sequences. The advantages of the 1H polarization are exploited inside the polarizer, while the interesting properties of 13C are put to use after dissolution. Still, it is also possible to observe directly the proton signal after dissolution. This can be extremely interesting, especially in the context of drug screening for pharmaceutical research. Two examples of such methods will be described. Finally, the use of hyperpolarized 1H signals after dissolution can be greatly improved if their relaxation rates could be attenuated. A first way of doing this, consisting in removing the paramagnetic species by filtration, will be explored. The use of Long-Lived States (LLS) will also be presented.