Automatic Abdominal Aortic Aneurysm Segmentation On Medical Image Sequence

Abdominal aortic aneurysm is caused by the dilation and weakness of the aorta.The segmentation of abdominal aortic aneurysm in the medical image plays an important role in clinical diagnosis and treatment of abdominal aortic aneurysm. It is because the inner and outer contour could be used to measure the volume of abdominal aortic aneurysm, which helps to assess the risk of abdominal aortic aneurysm. In research of the abdominal aortic aneurysm,the segmentation result is also need for stress analysis or the interaction of each component.Automatic and precise segmentation of the inner and outer contour of abdominal aortic aneurysm contributes the quantification analysis and improves the efficient of the treatment of this disease. Segmentation of the object with fuzzy contour is the main issue to be concerned in segmentation of medical images. The outer contour of abdominal aortic aneurysm is the typical fuzzy contour for the low contrast between the thrombus and the surrounding tissues.Manually segmentation of the outer contour is not a good solution. In this paper, the property of abdominal aortic aneurysm is fully used to get the precise segmentation of abdominal aortic aneurysm, which improving the efficiency of abdominal aortic aneurysm medical image sequence processing.1. According to the characters of the abdominal aortic aneurysm medical image sequence,the paper get the inner and outer contour respectively. Region restricting Chan-Vase model is used to get the inner contour. LBF based binary level set method restricted by narrow band is used to get the primary outer contour. Then the boundary detection method based on context information and narrow band restriction is used to optimize the primary outer contour, and the precise outer contour get after that.2. With the continuity character and context information of medical image sequence, this paper use the dynamic processing window and the inner and outer contour segmentation technique in single CTA slice to achieve the continuous segmentation result in the CTA sequence. B-Spring method is used to reduce the accumulation of error during the continuous segmentation processing. The contrast experiment shows the segmentation method proposedby this paper can get the precise outer contour of abdominal aortic aneurysm in automatically segmentation of the CTA sequence. The branches of the abdominal aorta and the component of abdominal aortic aneurysm is segmented based on outer contour segmentation result.3. An abdominal aortic aneurysm medical image sequence continuous segmentation system based on Matlab GUI is designed and implemented in this paper. The input of this system is medical image sequence with abdominal aortic aneurysm. These images loaded by the system can be shown by customized window wide and window center. All the component can be segmented in this system, including the lumen, thrombosis, and calcification. The segmentation result can be shown in the system and can be exported for further research.