基于区域竞争分割结果的体绘制研究

Volume Rendering Based on Level Set Segmentation Method of Region Competition Model

在开发医学图像处理系统时,采用区域竞争模型分割产生的图像只反映图像的轮廓及位置信息,不包含内部的图像信息,所以只能直接进行面绘制。据此,提出了基于区域竞争模型分割结果的体绘制算法。算法根据分割结果中的轮廓信息,对原始医学图像进行数值转换,使目标图像中既包含内部图像信息又包含轮廓边界信息,最后利用转换后的数据对肝脏进行体绘制。实验结果表明本算法有效可行,而且有很高的实用性。

The segmentation result produced by region competition model （RCM） only reflects its contour and position information without its inside image information in medical image processing system, so only the surface rendering can be achieved. Therefore, a volume rendering algorithm based on the segmentation result of RCM is proposed. The algorithm transforms the data of the original medical image according to the contour information in the segmentation result. The final image contains both image information and contour information. Finally, the transformed data is utilized for volume rendering in CT sequence, with the purpose of obtaining the liver area. The experiment result proves that the algorithm is feasible and practical.