Motion compensation for image-guided coronary intervention

经皮冠状动脉介入治疗（PCI）是心脏病学中的常规手术，旨在使用导丝、球囊和支架进行冠状动脉血管的血运重建。到目前为止，PCI的成功主要依赖于心脏病专家的经验，以心理上补偿冠状动脉运动，并在移动的管腔内安全导航，只有一个放射学视图。术前3D图像与2D在线X射线成像的图像配准将有助于导航引导。然而，这将仅提供冠状血管的静态路线图，而没有在手术期间观察到的心脏和呼吸运动的组合效应。本研究计划的主要目标是提出一种在线心脏和呼吸运动补偿的方法，以在冠状动脉血管的图像引导手术期间提供来自多切片计算机断层扫描（MSCT）的动态3D路线图。该研究计划的科学贡献是多方面的：1）在运动占主导地位的困难情况下的创新2D/3D配准技术，这是医学成像中一个非常具有挑战性的研究领域，因为我们依赖于2D投影来恢复3D姿态; 2）基于模式识别的鲁棒和个性化的运动补偿方案; 3）医学成像中的新颖的可变形2D/3D配准，其允许在经皮手术期间从可用的单个视图匹配患者解剖结构。心脏病和中风每年使加拿大经济在医生服务、医院费用、工资损失和生产力下降方面损失超过222亿美元。因此，具有运动补偿的导航引导系统对于为心脏病专家提供额外的视觉提示具有极大的兴趣。本研究计划探讨改善PCI导航与新的对象识别和图像配准算法。医学图像数据的高级可视化工具对于确保PCI期间的安全导航至关重要，并且可以有效地改善患者的治疗并缩短手术时间。这项研究的影响将是更安全的PCI，从而降低医疗成本。

Percutaneous Coronary Intervention (PCI) is a routine procedure in cardiology aimed at providing revascularization of coronary vessel using a guide wire, a balloon and a stent. As of today, the success of PCI relies mostly on the experience of the cardiologist to mentally compensate for coronary motion and navigate safely inside the moving lumen, with a single radiographic view. Image registration of preoperative 3D image with 2D online X-ray imaging would be helpful for navigation guidance. However, this would provide only a static roadmap of the coronary vessels without the combined effect of cardiac and breathing motion observed during the procedure. The main goal of this research program is to propose a methodology for online cardiac and breathing motion compensation, to provide a dynamic 3D roadmap from Multi Slice Computed Tomography (MSCT) during image guided procedures of coronary vessels. The scientific contributions of this research program are numerous: 1) innovative 2D/3D registration techniques in the difficult case where motion is predominant, a very challenging field of study in medical imaging, since we rely on a 2D projection to recover a 3D pose; 2) robust and personalized motion compensation scheme, based on pattern recognition; 3) novel deformable 2D/3D registration in medical imaging, which allows matching patient anatomy from a single view available during percutaneous procedure. Heart disease and stroke costs the Canadian economy more than $22.2 billion every year in physician services, hospital costs, lost wages and decreased productivity. Hence, a navigation guidance system with motion compensation is of great interest to provide additional visual cues for cardiologists. This research program investigates the improvement of PCI navigation with novel object recognition and image registration algorithms. Advanced visualization tools of medical image data are paramount to ensure safe navigation during PCI and can be effective to improve patient's treatment and to reduce operating time. The impact of this research would to perform safer PCI and hence reduce medical costs.