Fast angiographic OCT imaging using sparse representations over learned dictionaries

Recent developments in optical coherence tomography (OCT) have enabled wide-field and high-resolution angiographic imaging. However, generation of vascular contrast inherently requires long imaging times. In this work, we demonstrate a reconstruction technique based on sparse and redundant representations over trained dictionaries that can be used to reduce acquisition times and accurately reconstruct angiographic OCT projection images with full vascular detail from a smaller number of B-scans than conventionally required. Our technique for fast angiographic imaging through reconstruction (FAIR), shows excellent reconstruction quality while using only half of the number of B-scans and graceful quality degradation with further undersampling.