Optical coherence tomography is an emerging technique for performing noninvasiveoptical biopsy of tissue with micron level resolution via low coherence interferometry. Recently, as the demand is growing for high throughput and deep imaging, OCT seems to become a good candidate to accomplish these tasks. The coherence gate with heterodyning detection achieve high sensitivity in imaging scattering tissue and the Fourier domain OCT acquires millimeter of depth simultaneously with micron level axial resolution. In addition to these great advantages, in this dissertation, I will present approaches that makes use of the complex phase from OCT, which enables the split of image formation process into a combination of hardware and computational components. By synergistically leveraging the benefits of both hardware and computational techniques, we can achieve a higher volumetric imaging throughput and a deeper penetration in OCT, with experimental data demonstrated in both live and ex vivo imaging. I will first provide analysis on the theoretical framework for hardware and computational image formation in OCT. Then I will demonstrate how this hybrid approach can enable high throughput and deep imaging with experimental data.