Approximate Expressions for Error Statistics in Point-Based Image Registration

基于点的图像配准中误差统计的近似表达式

• 批准号: 9802982
• 负责人: J. Michael Fitzpatrick
• 金额: $ 23.65万
• 依托单位: Vanderbilt University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1998
• 资助国家: 美国
• 起止时间: 1998-08-15 至 2002-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9802982&HistoricalAwards=false
• 关键词: Approximate; Expressions; Error; Statistics; Point

9802982FitzpatrickThe goal of this research is to develop, via perturbation methods, new expressions for error statistics associated with point-based rigid registration in medical imaging, with emphasis on target registration where the targets are typical "targets" of surgery. A specific motivation is that knowledge of these errors is crucial in the planning and guidance of surgery, especially brain surgery. The numerical methods to estimate these errors are extremely unwieldy and time consuming and if reliable closed form approximations were available they could substantially advance the utility of image guided surgery, among other applications of imaging.***

9802982Fitzpatrick本研究的目标是通过扰动方法开发与医学成像中基于点的刚性配准相关的误差统计的新表达式，重点是目标配准，其中目标是手术的典型“目标”。 一个具体的动机是，这些错误的知识是至关重要的手术，特别是脑外科手术的规划和指导。 估计这些误差的数值方法非常笨拙和耗时，如果可靠的封闭形式近似可用，它们可以大大提高图像引导手术的实用性，以及其他成像应用。