To solve the phase unwrapping problem for phase images in Fourier domain Doppler optical coherence tomography (DOCT), we propose a deep learning-based residual en-decoder network (REDN) method. In our approach, we reformulate the definition for obtaining the true phase as obtaining an integer multiple of 2π at each pixel by semantic segmentation. The proposed REDN architecture can provide recognition performance with pixel-level accuracy. To address the lack of phase images that are noise and wrapping free from DOCT systems for training, we used simulated images synthesized with DOCT phase image background noise features. An evaluation study on simulated images, DOCT phase images of phantom milk flowing in a plastic tube and a mouse artery, was performed. Meanwhile, a comparison study with recently proposed deep learning-based DeepLabV3+ and PhaseNet methods for signal phase unwrapping and traditional modified networking programming (MNP) method was also performed. Both visual inspection and quantitative metrical evaluation based on accuracy, specificity, sensitivity, root-mean-square-error, total-variation, and processing time demonstrate the robustness, effectiveness and superiority of our method. The proposed REDN method will benefit accurate and fast DOCT phase image-based diagnosis and evaluation when the detected phase is wrapped and will enrich the deep learning-based image processing platform for DOCT images.

No abstract

No abstract

The most challenging problem in the stitching test of large flats with a small-aperture interferometer is the accumulation effect of the second-order error. As it is approximately enlarged by the square of the ratio of full aperture size to subaperture size, a very small amount of the second-order error in the reference surface of a transmission flat can be accumulated and gets far from negligible when the subaperture is far smaller than the full aperture. We present here a solution by using two orthogonally arranged wavefront interferometers. One is responsible for a subaperture test and the other for the simultaneous measurement of relative tilts. Because the accumulation effect originates from the lateral shift of the second-order error, only the tilt along the subaperture scanning direction needs to be measured accurately. It is no longer determined by stitching optimization instead to avoid the error accumulation. Piston and tilt perpendicular to the scanning direction are still determined by stitching optimization. The method is experimentally verified and compared to the stitching test with the reference surface error calibrated out, both referenced to the full aperture test result obtained with a 24-inch interferometer.

No abstract