摘要
目的提出一种乳腺疾病无损检测新技术及其分析方法。方法基于近红外脉冲光的乳腺疾病诊断系统软件部分以VC为开发工具,主要实现图像同步采集、图像配准、分析处理和数据管理等功能,其中分析处理部分包含感兴趣点分析和区域分析两种方法,点分析主要包含功率谱曲线、相关性曲线和相关系数,区域分析则是计算区域内所有点相关系数形成由相关系数构成的伪彩图像。文中对一例典型的乳腺癌患者做了详细的诊断分析。结果乳腺癌患者病变部位的相关系数一般小于0.6,其功率谱呈现很多次频峰,互相关曲线上出现很多毛刺。而正常人的相关系数基本都在0.8以上,其功率谱和互相关曲线较光滑。结论该系统具备非侵入、无辐射、廉价的特性,通过对典型病例的光强分析展示了系统的临床使用价值。
Objective To propose a new non-invasive detection technology for breast disease and the analysis method. Methods The software of the galactophore disease diagnosis system based on near-infrared pulsed light is developed with Visual C on the platform of windows. The software function is synchronized image acquisition, image registration, analysis of light signal and management of data in this system. The analysis function contains analysis of the interest points and the interest regions. Point analysis consists of the power spectrum curve,correlation e.urve and coorrclation coefficient. Region analysis utilizes the correlation eoefficient of all points ira the interest region to form a pseudo-color image for diagnosis. And a typical breast cancer case is diagnosed and analyzed by this system. Results Commonly,the correlative coefficient of locus in this case is under 0.6, and many wave crests appear in the power spectrum with many burrs in the correlation curve. However,the correlation coefficient of a heahhy person is above 0.8, and the power spectrum and orrelation urve is smooth. Conclusions This system is non-invasive,nonradiative and cheap. The analysis on the typical breast cancer case indicates the clinical application values.
出处
《北京生物医学工程》
2013年第3期248-254,共7页
Beijing Biomedical Engineering
基金
湖北科技学院科研基金(2011C088)资助
关键词
乳腺疾病
近红外光
点分析
区域分析
breast disease
near-infrared light
point analysis
region analysis