光学相干层析成像
光学
成像体模
物理
运动(物理)
噪音(视频)
生物医学工程
计算机科学
计算机视觉
材料科学
运动检测
断层摄影术
光学相干断层摄影术
图像质量
图像分辨率
人工智能
帧速率
作者
Jinyu Fan,Caihua Dong,Yi He,Lina Xing,Guohua Shi
标识
DOI:10.1002/jbio.202000469
摘要
In this work, we developed a motion estimation and correction method which real-time obtained the direction and displacement of repetitive micro bulk motion (such as cardiac and respiratory motion) on an SS-OCT system without additional tracking hardware, and reduced the motion noise in optical coherence tomography angiography (OCTA). In the approach, the direction of repetitive micro bulk motion was considered fixed, and proportional relationships between the motion components in three directions were determined; Then we performed one-dimension cross-correlation to obtain depth displacement which was further used to obtain other two motion components, and greatly reduced the computation; The processing speed on a graphic processing unit was 478 pairs of B-Scans per second, and the measurement range was larger than the range of the angiogram-based methods. Lastly, corrupt angiograms were recovered by adaptive scan protocol, and reduced acquisition time in comparison with the previous work. This article is protected by copyright. All rights reserved.
科研通智能强力驱动
Strongly Powered by AbleSci AI