School of Computing, Queen's University, Kingston, Ontario, Canada.
Ultrasound Med Biol. 2011 Aug;37(8):1314-26. doi: 10.1016/j.ultrasmedbio.2011.05.004. Epub 2011 Jun 16.
This article presents a novel approach that incorporates an ultrasound slice-thickness profile into a filtered, weighted-least-square framework to improve the reconstruction accuracy of a real-time freehand calibration system. An important part of the system is a slice-thickness calibration device that aids in the extraction of the slice thickness across a wide range of imaging depths. Extensive experiments were conducted on a 10,000-image dataset to evaluate the effects of the framework on the calibration accuracy. The results showed that three-dimensional (3-D) reconstruction errors were significantly reduced in every experiment (p < 0.001). Real-time testing showed that the proposed method worked effectively with a small number of input images, suggesting great potential for intraoperative use where only a limited number of data may be available. This new framework can enable efficient quality control of calibration accuracy in real-time operating-room use.
本文提出了一种新方法,即将超声切片厚度轮廓纳入滤波、加权最小二乘框架中,以提高实时自由手校准系统的重建精度。该系统的一个重要部分是切片厚度校准设备,它有助于在广泛的成像深度范围内提取切片厚度。在一个包含 10000 张图像的数据集上进行了广泛的实验,以评估该框架对校准精度的影响。结果表明,在每个实验中,三维(3-D)重建误差都显著降低(p<0.001)。实时测试表明,该方法在输入图像数量较少的情况下也能有效工作,这表明在只有有限数量的数据可用的术中情况下具有很大的应用潜力。这个新框架可以在实时手术室使用中实现对校准精度的高效质量控制。