We present a method to utilize 'guide stars' within a sample, as in hardware-based adaptive optics, in order to correct aberrations using computational adaptive optics. Interferometric synthetic aperture microscopy (ISAM), a computed imaging technique to overcome the depth-of-field limitations in optical coherence tomography/microscopy (OCT/OCM), is utilized to correct defocus and reveal local high-scattering regions that can serve as potential guide stars. The correction of pupil aberrations using Zernike polynomials can be applied pre- or post-ISAM in order to improve resolution and signal-to-noise ratio of these guide stars. Aberrations of the effective pupil function can then be determined through subsequent windowing and Fourier transforming of a guide star signal. Computational adaptive optics based on guide stars is demonstrated in a phantom consisting of sub-resolution scatterers, with work underway to extend this method to biological tissues.