Grau Vicente, Becher Harald, Noble J Alison
Wolfson Medical Vision Laboratory, Department of Engineering Science, University of Oxford, Parks Road, OX1 3PJ Oxford, United Kingdom.
Med Image Comput Comput Assist Interv. 2006;9(Pt 1):612-9. doi: 10.1007/11866565_75.
Real time three-dimensional echocardiography opens the possibility of interactive, fast three-dimensional analysis of cardiac anatomy and function. However, at the present time these capabilities cannot be fully exploited due to the low image quality associated to this modality. We propose to increase image quality and information content by combining images acquired from different echocardiographic windows. In this paper, we present an algorithm to register these datasets. Phase-based measures have been proposed as a suitable alternative to intensity-based ones for ultrasound image analysis. The proposed algorithm uses a new cost function, based on local orientation and phase differences, to align the datasets. Visual observation of results, and preliminary numerical analysis, show the robustness and accuracy of this method.
实时三维超声心动图开启了对心脏解剖结构和功能进行交互式快速三维分析的可能性。然而,目前由于这种成像方式相关的图像质量较低,这些功能无法得到充分利用。我们建议通过组合从不同超声心动图窗口获取的图像来提高图像质量和信息含量。在本文中,我们提出了一种用于配准这些数据集的算法。基于相位的度量方法已被提议作为基于强度的方法的合适替代方案用于超声图像分析。所提出的算法使用一种基于局部方向和相位差的新代价函数来对齐数据集。结果的视觉观察和初步数值分析表明了该方法的稳健性和准确性。