基于标记的肝脏CT图像智能提取

针对腹部CT图像所含脏器多且复杂的特点，结合解剖学知识，提出了一种新的基于多尺度梯度的医学图像分割方法。结合图像的灰度直方图的梯度变化，对图像进行内外标记，而后利用强制最小技术对多尺度梯度进行修改。在修改后的梯度图像中直接进行全局阈值分割，并结合解剖学知识再进行相应的形态学处理，最终可以较好的将肝脏从腹部图像中提取出来，适合用于肝脏的智能提取。

Intelligent segmentation for liver CT images based on labeling

A novel method of medical image segmentation based on multi-scale gradient is developed in this paper. With the changes of the gradient of the image's histogram, the internal and external markers are marked in the image. These markers are then used to modify the multi-scale gradient using the minima imposition procedure. Using the global threshold segmentation of the modified gradient, a binary image is got. Combining the anatomical knowledge and morphological process, the liver can be extracted from the abdominal CT image successfully. The result demonstrates that the method can be used for intelligent segmentation for liver CT images.