Background: In recent years, the NUMEN project has highlighted that the systematic study of double charge exchange (DCE) nuclear reactions could provide precious information on the nuclear matrix elements of neutrinoless double β decay. To achieve this goal, a multichannel approach was adopted, where a plethora of different reaction channels is measured under the same experimental conditions and analyzed in a coherent framework. In this context, the initial (ISI) and final (FSI) state interactions are fundamental building blocks to investigate all the relevant direct nuclear reactions. To date, such building blocks are not sufficiently established for many of the projectile-target pairs of interest for NUMEN experiments. This is also the case of the O18+Ti48 system, which is relevant for the double β decay of the Ca48 nucleus into the Ti48 one. Purpose: This work aims at deducing the ISI to be used in the multichannel study of the O18+Ti48 collision at 275 MeV incident energy. To this purpose, the optical potential and the strength of the couplings between elastic and inelastic scattering channels are determined and characterized by comparison with new experimental data. Methods: Cross section angular distributions were measured for the elastic and low-lying inelastic scattering channels in a wide range of momentum transfer. Experimental data were compared to theoretical calculations performed in optical model (OM), distorted-wave Born approximation (DWBA), coupled channels (CC), and coupled channels equivalent polarization potential (CCEP) approaches. For all of them, the São Paulo double-folding potential was adopted as the optical potential. The comparison between the predictions of the OM/DWBA and CC frameworks allowed to evaluate the strength of the couplings to the inelastic channels which were explicitly taken into account. Within the CC formalism, different coupling schemes were compared to assess the contributions of the states included in the model space. A further test of our approach was carried out by analyzing the experimental data of the same system at 54 MeV incident energy. Results: The achieved energy resolution allowed to resolve the elastic scattering and the excitation of the 21+ state of the Ti48 nucleus. A broad structure, associated to the superposition of the inelastic transitions to the 21+ state of O18 and to three excited states of Ti48, was also observed. The OM and DWBA calculations are not able to reproduce the experimental elastic and inelastic angular distributions in the explored range of momentum transfer. A significant improvement in the description of the data is found by using the CC approach. In particular, when the 21+ and 31- collective states of projectile and target and their simultaneous excitations are included in the coupling scheme, a satisfactory agreement is achieved. The CCEP predictions describe reasonably well the elastic scattering data. Conclusions: Channel coupling effects have to be taken into account for a good description of the elastic and inelastic experimental angular distributions in the measured range of momentum transfer. The inclusion of such couplings in the ISI is envisaged for the analysis of all the direct nuclear reactions induced by the O18+Ti48 collision in this incident energy region.