摘要
实现医学超声图像的自动化定量分析。设计并实现一个基于人机交互的医学超声图像辅助分析系统。首先由医生凭借临床经验在超声图像中划出目标区域,系统生成目标子图像,然后使用直方图均衡和迭代阈值法对目标子图像进行灰度调整和分割处理,最后根据位图标记结果在原图中提取目标轮廓,并自动生成反映目标大小的长、宽等数据。通过采用一定的图像处理技术,成功地实现了对超声图像中目标区域的分割和自动化定量分析。本系统能弥补当前超声图文管理系统在图像分割和图像分析方面的不足,具有较高的实用性。
To realise automatic quantitative analysis of the medical ultrasonic image. A computer-aided analysis system of medical ultrasonic image which is based on human-computer interaction has been designed and implemented. First, an object region is marked out in ultrasonic image according to doctor＇ s clinical experience for system to generate sub-image. Then histogram equalisation is used to adjust the grey range of the object sub-image, and iterative threshold segmentation is used to Segment the sub-image. At last, the object contour is extracted on original image according to the result of the bitmap markedness, and the size data including length and width of the object is given by the computer automatically. Segmentation and automatic quantitative analysis of the object region in medical ultrasonic image has been realised successfully by using some image processing technologies. This system can make up the deficiency of current ultrasonic image and text management system in such aspects as image segmentation and analysis. It has quite high practicability.
出处
《计算机应用与软件》
CSCD
2009年第5期221-224,231,共5页
Computer Applications and Software
关键词
医学超声图像
直方图均衡
迭代阈值分割
位图标记
自动化定量分析
Medical uhrasonic image Histogram equalisation Iterative threshold Bitmap markedness Automatic quantitative analysis
