Joint Iterative Reconstruction and Motion Compensation for Optical Coherence Tomography Angiography

虽然OCTA成像在改善眼科患者护理方面显示出巨大的前景，但其全部潜力仍然有限。迫切需要健壮和准确的运动检测和校正以及改进的OCTA处理算法，以降低噪声水平并提高图像质量和分辨率。此外，能够让世界各地的研究人员制作高质量图像的开源软件寥寥无几。为了满足这些需要，该项目将追求以下目标：a)改进运动补偿，包括使用OCTA信号以提高数据一致性，建立更精确的运动模型，允许旋转和缩放等仿射运动，B)物理校正OCTA信号提取，在OCTA信号提取中使用基于压缩感知的正则化方法，并集成来自目标A的完整三维运动模型，包括内插过程和正确的物理噪声模型。C)精确学习重建，使用附加的深度学习技术来增强来自目标B的物理校正模型，以学习数据-最优稀疏域和最优导航模式用于信号提取。本项目中创建的所有软件将作为开源软件发布。

While OCTA imaging has shown great promise for improving patient care in ophthalmology, its full potential has been limited. There is an urgent need for robust and accurate motion detection and correction and improved OCTA processing algorithms that will decrease noise levels and improve image quality and resolution. Furthermore, there is only few open-source software that enables researchers from around the world to produce high quality images. In order to address these needs, the project will pursue the following objectives:A) Improved motion compensation including the use of the OCTA signal for improved data consistency, more accurate motion models that allow affine motions such as rotation and scaling, and integration of surrogate signals stemming from camera-based eye-tracking or navigator-based motion estimation approaches yielding a full 3-D correction for all acquired A-scan data.B) Physically correct OCTA signal extraction that employs compressive sensing-based regularization approaches in the OCTA signal extraction and integrates the full 3-D motion model from Objective A including the interpolation process in combination with correct physical noise models.C) Precision Learning Reconstruction that augments the physically correct model from Objective B with additional deep learning techniques to learn data-optimal sparse domains and optimal navigator patterns for motion signal extraction.All software created in this project will be published as open source software.