案例信息
标题：医学图像噪声处理效果评价的设计与实践
【专业学位类别】公共卫生
【专业领域/方向】转化医学
【作者】赵庆亮
【单位】厦门大学公共卫生学院
【关键词】光学相干断层扫描血管成像技术(OCTA)；图像处理；定量分析
【适用课程】分子影像学基础
【案例摘要】
定量数据分析是医学图像处理的一个重要的部分。在基于小波与傅里叶分析结合的去噪算法被应用到OCTA活体成像图像，去除运动伪影中时，基于观察的定性分析主观性太强，因此需要设计一个合理的定量数据评价环节来评估算法的去噪效果。
本研究案例根据医学血管图像的两个主要的应用：提取血管形态结构数据分析、直接分析，分别计算图像的血管参数和图像相似度参数，对所提出的算法性能进行全面的评估。定量的结果显示，在被比较的四种算法中，MW-FFTA得到了最好的效果。
通过实现该研究的关键环节，使学生掌握从医学图像中合理地挖掘有用信息的能力，学习方案设计与实施；培养学生在分子影像领域不断创新的能力和素养，为未来精准医学时代的到来做好准备。
Title:Design and practice of medical image noise processing effect evaluation
Professional Degree Category: Master of Public Health
Specialized fields: Translational Medicine
Authors: Qingliang Zhao
Affiliation: School of Public Health, Xiamen University
Key words: Optical coherence Tomography angiography (OCTA); Image processing; Quantitative analysis
Abstract
Quantitative data analysis is an important part of medical image processing. When the denoising algorithm based on wavelet and Fourier analysis is applied to OCTA live imaging images to remove motion artifacts, the observation-based qualitative analysis is too subjective, so it is necessary to design a reasonable quantitative data evaluation link to evaluate the denoising effect of the algorithm.
This case is based on two main applications of medical vascular images: extraction of blood vessel morphology and structure data analysis, direct analysis, calculation of blood vessel parameters and image similarity parameters respectively, and a comprehensive evaluation of the performance of the proposed algorithm. Quantitative results show that MW-FFTA has the best effect among the four algorithms compared.
Through the realization of the key link of the research, students can master the ability of reasonably mining useful information from medical images, and learn the design and implementation of the program; Cultivate students' ability and accomplishment of continuous innovation in the field of molecular imaging, and prepare for the arrival of the era of precision medicine in the future.
【案例正文】
【案例说明书】
