Research On Three-dimensional Reconstruction Of Aortic Dissection Based On Medical CT Image Sequence

Aortic dissection is one of the most dangerous cardiovascular diseases,its main treatment method is lumen isolation technique,which requires that the clinicians should know clearly about the location of aortic intima crevasse and the stripping range of aortic dissection membrane.However,the three-dimensional model of aortic dissection reconstructed by existing medical image processing software can not display aortic dissection membrane intuitively,and it is difficult to obtain the location and the size of the intima crevasse.Therefore,this thesis designed a three-dimensional reconstruction method of aortic dissection based on medical CT image sequence,by which the three-dimensional visualization of aortic dissection can be realized rapidly and accurately,which is of great significance to improve the success rate of surgery and save patients’ lives.This method mainly includes three parts:image preprocessing,target object segmentation and three-dimensional reconstruction.The main works are as follows:First,according to the characteristics of CT image noise,we chose median filtering method for image smoothing.Then image sharpening with laplacian operator is essential to strengthen the edge of aorta and aortic dissection membrane.Finally,we utilized grayscale linear mapping method to improve the overall brightness and image contrast.Second,active shape model algorithm was applied to the aorta segmentation.Meanwhile,in aortic dissection membrane extraction process,the spatial continuity prior model based on Bayesian theory was proposed in this thesis.This method could accurately identify the pixels of dissection membrane and remove other non-target pixels after using Hessian matrix to extract contour pixels.Third,we utilized volume rendering method based on ray casting algorithm for three-dimensional reconstruction of aortic dissection.Moreover,we set the transparency and colors of aortic dissection three-dimensional model to make aortic dissection membrane more intuitive,the location and the size of the intima crevasse can be esaily obtained.The spatial continuity of CT image sequence is fully used in the segmentation process,which can improve the segmentation accuracy.We selected the Dice coefficient as evaluation index of segmentation precision and set manual segmentation results as gold standard.Confirmed by experiments the precision of the proposed segmentation algorithm is satisfactory.The visualization effect of aortic dissection three-dimensional model can meet the requirements of the clinicians,which is of great significance to improve the success rate of surgery and save patients’ lives.