Author Affiliations
1Beijing Engineering Research Center of Mixed Reality and Advanced Display, School of Optics and Photonics, Beijing Institute of Technology, Beijing 100081, China2Key Laboratory of Photoelectronic Imaging Technology and System (Beijing Institute of Technology), Ministry of Education, Beijing 100081, China3Department of Hand Surgery, Beijing Ji Shui Tan Hospital, Beijing 100035, Chinashow less
Fig. 1. MRW algorithm segmentation procedure of intensity image. First stage shows the process of upper boundary segmentation. Second stage shows the process of bottom boundary segmentation. Then, they are combined to create an intact boundary mask (scale bar: 500 μm).
Fig. 2. Selection of seed points in the segmentation stages for OCT intensity image. (a) Two kinds of seeds are selected for the upper boundary segmentation in the first stage, marked by green and blue lines, respectively. (b) Two kinds of seeds are selected for the bottom boundary segmentation in the second stage, marked by green and blue lines, respectively (scale bar: 500 μm).
Fig. 3. Segmentation procedure of the phase image. We combined phase image with a boundary mask to remove the background, and then made use of the threshold condition to get the blood flow region (scale bar: 500 μm).
Fig. 4. Probability map and corresponding segmentation results using different regularization. (a) Probability map using the regularization . (b) Segmentation result of (a). (c) Probability map using the regularization . (d) Segmentation result of (c) (scale bar: 500 μm).
Fig. 5. Segmentation results of different OCT frames. (a-1)–(d-1) OCT intensity images of frames 1, 65, 131, 192, and segmentation results, respectively. (a-2)–(d-2) OCT phase images of frames 1, 65, 131, 192, and segmentation results, respectively (scale bar: 500 μm).
Fig. 6. 3D reconstruction of the vessel. (a) The upper part of the blood vessel. (b) The bottom part of the blood vessel. The arrows point at the thrombosis position.
Fig. 7. (a) Blood flow area variation along the blood flow axis. (b) Blood vessel radius variation along the blood flow axis.
Group | Frame 1 | Frame 65 | Frame 131 | Frame 192 |
---|
Image Type | Intensity | Phase | Intensity | Phase | Intensity | Phase | Intensity | Phase |
---|
DC (%) | 96.60 | 95.66 | 96.55 | 94.98 | 97.11 | 92.31 | 96.12 | 95.13 | SNC (%) | 98.79 | 92.54 | 97.04 | 91.01 | 98.95 | 90.97 | 97.96 | 92.08 | SPC (%) | 99.46 | 99.93 | 99.53 | 99.96 | 99.42 | 99.61 | 99.39 | 99.88 |
|
Table 1. Evaluation Parameter Comparison Between Segmentation Results and Ground Truth