Fully automated detection of lamina cribrosa in optical coherence tomography: Framework and illustrative examples

We present an algorithm for automated detection of lamina cribrosa (LC) using optical coherence tomography scans. To the best of our knowledge, it is the first algorithm of this type, as previous attempts relied heavily on characteristic points marked a priori by a human expert and were hence semi-automated at best. First, we highlight the unwanted, yet unavoidable, influence of image rescaling necessary to provide the detection algorithm with real-world image proportions. Next, we simulate an LC phantom and test the influence of various combinations of the rescaling method and SNRs on the LC detection efficacy. Finally, we present some illustrative results from real-data analysis.