量子衍生超声医学图像去斑方法研究

项目名称： 量子衍生超声医学图像去斑方法研究

项目编号： No.61201423

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 电子学与信息系统

项目作者： 付晓薇

作者单位： 武汉科技大学

项目金额： 24万元

中文摘要： 量子衍生信息处理技术是量子信息学同经典计算机技术相结合的产物，是在经典计算机硬件基础上，充分发挥量子信息技术优势的一种途径。医学超声具有无损无痛、价廉、实时性强等优点，被广泛地应用在临床诊断中。由于超声信号在体内的散射，图像存在大量复杂的斑点噪声，严重影响着图像质量，增加了临床诊断与治疗的难度。因此，超声图像去斑是医学图像处理的一个重要课题。目前，基于统计模型的小波图像去斑方法是医学超声图像去斑的主要方法，但仍存在去斑平滑与图像细节保持之间的矛盾、鲁棒性较差、计算复杂度大等缺点。本项目将借鉴并利用量子力学的基本理论，充分发挥量子特性优势，提出一种新的量子衍生医学超声图像去斑方法并将其应用在数字超声成像系统中。该方法不依赖于量子级物理设备，将实现量子力学理论与图像处理技术的相互渗透与有机结合，为解决超声图像处理提供一条新的有效途径。

中文关键词： 量子信号处理；图像去斑；复小波变换；参数估计；

英文摘要： Quantum-inspired information processing technology is derived from quantum informatics with classic computer techniques, which takes good advantage of the quantum information based on classic computer hardware. Medical ultrasound is widely used in the clinical diagnosis with the characteristics of the painless, low price, good real-time performance and etc. With the scattering of ultrasonic signals in the body, there are lots of complex speckle noises in the medical ultrasound image. It influences the image quality seriously and increases the difficulty of clinical diagnosis and treatment. Therefore, the despeckling of ultrasound image is an important issue in medical image processing. At present, the main despeckling methods are based on the image statistical model in wavelet domain. However, they still have the contradiction between despeckling and preserving details of medical ultrasound image, and have the shortcomings of poor robustness, the great computational complexity. With the principle of quantum mechanics and the advantages of quantum properties, a novel despeckling method of medical ultrasound image is proposed, which will be applied to digital ultrasound imaging system. The method doesn't depend on the quantum level of physical equipment. It promotes the realization of the combination and mutual pe

英文关键词： Quantum Signal Processing；Image Despeckling；Complex Wavelet Transform；Parameter Estimation；