Study Of The Liver CT Image Segmentation And3D Reconstruction Algorithm

Liver cancer is one of the most common tumors in our country. Early detection ofthe pathological changes has profound significance for the timely treatment of the livercancer. As a key part of medical image processing, the results of medical imagesegmentation will have a direct influence on the following image processing. It is theoperation of three-dimensional reconstruction that provides more intuitive informationabout the patients in clinical diagnosis. Therefore, the research which aimed at thealgorithm of image segmentation and three-dimensional reconstruction of the CT liveris of great clinical meanings.The major contribution of the thesis is as follows:1、Based on the basic idea and principle of the region growing algorithm, the thesisproposed an improved region growing and mathematical morphology algorithm. A newmethod of adaptive region growing is put forward, which uses local area gray meaninstead of the global average values, the standard deviation of the local region asthreshold. Compared with the results of traditional region growing method、mathematical morphology algorithm、threshold and fuzzy c-means algorithm, theproposed algorithm is proved to be of great accuracy. The aim of the thesis is to discussthe relation between the changes of liver signal intensity and functionality. The livervolume and average gray value of the patients at certain time were recorded, whichprovided theoretical basis for further research.2、In order to evaluate the effectiveness of the improved region growing andmathematical morphology segmentation method, the volume rendering was adopted toreconstruct the structure of liver in consideration with the feature of the liver. Throughthe operation of MATLAB software, the liver can be reconstructed with satisfactoryresults. What’s more, the reconstruction results by professional software Amira4.1indicate the algorithm is available.The structure of this paper is as follows: chapter one illustrates the background andmeaning of the project, relevant home and abroad research situation and major researchcontents; chapter two introduces the theoretical foundation of medical imagesegmentation; chapter three proposes an improved region growing and mathematicalmorphology algorithm. on the basis of region growing algorithm principle; the livervolume and average gray value of the patients at certain time were recorded, which provided theoretical basis for further research. chapter four introduces the volumerendering to reconstruct the structure of liver, including relevant theoretical analysis;chapter five illustrates the major contribution of the paper and the future work.