Phase-sensitive coherent imaging exploits changes in the phases of backsca ered light to observe tiny alterations of sca ering structures or variations of the refractive index. But moving sca erers or a fluctuating refractive index decorrelate the phases and speckle patterns in the images. It is generally believed that once the speckle pa ern has changed, the phases are scrambled and any meaningful phase di erence to the original pattern is removed. As a consequence, di usion and tissue motion below the resolution handicap phase-sensitive imaging of biological specimen. Here, we show that, surprisingly, a phase comparison between decorrelated speckle pa erns is still possible by utilizing a series of images acquired during decorrelation. The resulting evaluation scheme is mathematically equivalent to methods for astronomic imaging through the turbulent sky by speckle interferometry. We thus adopt the idea of speckle interferometry to phase-sensitive imaging in biological tissues and demonstrate its e icacy for simulated data and for imaging of photoreceptor activity with phase-sensitive optical coherence tomography. The described methods can be applied to any imaging modality that uses phase values for interferometry.