Optical Coherence Tomography (OCT) is instrumental in providing high-resolution images/volumes of biological tissues, revealing micro-structural information. Its significance is particularly pronounced in oncology for the early detection of malignant tumours and optical biopsy. The acquisition of OCT data involves a mechanical scan of the tissue surface by a light beam, delivering interferometric information along the tissue depth. Due to the mechanical scan, the acquisition frequency and spatial resolution are antagonistic (increasing the acquisition frequency compromises spatial resolution, and vice versa). Additionally, a rolling shutter acquisition mode is induced, causing geometric artefacts in the OCT volume. In view of these issues, the contributions in this thesis are threefold: i) propose OCT data acquisition using non-raster trajectories, accompanied by a set of criteria to assess both the scan trajectories and the resulting volumetric data. Experimental results demonstrated that non-raster trajectories were nearly twice as fast as the raster trajectory without a significant loss in quality. ii) propose an adaptive strategy in OCT data acquisition, where data is selectively acquired in informative regions of the field of view, optimising the scan process and potentially reducing the acquisition time. And iii) propose a methodology for estimating motion from geometrically distorted OCT volumes, particularly addressing the challenge of estimating shape and motion from only one OCT volume.