计算机辅助的微创手术工具跟踪算法综述-中国数字医学.pdfVIP

数字医学基础研究 Basic Research of Digital Medicine 计算机辅助的微创手术工具跟踪算法综述 ① ＊ 陈兆瑞 赵子健 摘 要 为了深入研究计算机辅助的手术工具跟踪算法，总结了已有的研究成果，归纳了这项技术的先进的主流算法，基于 其各自特点进行了分类总结，并对各自结果进行比较评价。结果表明，学习算法具有较高的准确度；普通视频图像分析算法 是未来工作中有潜力的算法；立体视频分析算法需要依靠硬件设备的发展；有辅助设备的算法则是传统的有效算法。各类算 法都有发展空间，结合目前该领域的发展趋势，基于深度学习结合计算机视觉的算法将是未来发展的趋势。 关键词　手术工具跟踪 学习 视频图像分析 辅助设备 Doi:10.3969/j.issn.1673-7571.2016.09.028 [中图分类号] R318 [文献标识码] A Survey of Computer Aided Minimally Invasive Surgical Tools Tracking Algorithms / CHEN Zhao-rui, ZHAO Zi-jian//China Digital Medicine.-2016 11(9): 83 to 87 Abstract In order to explore the computer aided surgical tools tracking algorithms, we review the former research results, summarize the advanced main algorithms of the technology, classify and conclude these algorithms base on their features, and compare and evaluate their results. The results indicate that learning algorithms achieve high accuracy; ordinary video image analysis algorithm is a potential in future work; stereo video analysis algorithms need to rely on the development of hardware devices; auxiliary equipment algorithms are traditional and efficient algorithms. All kinds of algorithms have development space, combining the actual development trend of this field; we think it will be a possible development trend to the algorithms based on deep learning and computer vision on tracking surgical tools. Keywords surgical tools tracking, learning, video image analysis, auxiliary equipment Fund project The Award Research Fund of Outstanding Young and Middle-aged Scientists of Shandong Province (No. BS2013DX027); Doctor Station Fund of the Ministry of Education (No. 20130131120036); National Natural Science Foundation (No. Corresponding author S