Research On Optimization And Auxiliary Diagnosis Of Medical Breast Cancer Imaging

Breast cancer is one of the most common malignant tumors in the world,and it is a serious threat to women’s health.Early detection and correct diagnosis of breast cancer is essential for the later treatment.Medical imaging is an important diagnostic method in the field of modern medicine.In this thesis,we carry out a research aimed at the optimization of DBT image quality and assisted diagnosis of ultrasound image segmentation of the functional tissue.X-ray limited angle digital tomography(Breast Tomosynthesis Digital,DBT)is by far the highest diagnostic rate of breast cancer imaging method.Due to the limited reception angle of the receiver,the image reconstruction layer has a phenomenon of discontinuity in the vicinity of the perspective,which is called truncation artifact.In this paper,two algorithms are proposed to suppress the truncation artifacts.Method one is digital image processing.It is used to compensate the gray level of the reconstructed data in order to remove the artifacts.Method two is based on the SART reconstruction algorithm,which constrains the special items in the reconstruction process and eliminates artifacts in the reconstruction algorithm.Three groups of data were used to verify the two methods,both of which eliminate or suppress the artifacts while maintaining breast tissue structure after processing,thus verifying the effectiveness of these two algorithms.Some of the Z direction data is missing in the X-ray imaging process of the DBT system.And because the data is not complete,the imaging results have some severe artifacts,which reduces the resolution and shows the contrast with the actual tissue to be reconstructed.The three-dimensional ultrasound image has a high resolution in the Z direction and a clear edge,which can be used as the supplement and auxiliary methods of the X-ray image reconstruction.By adding the three-dimensional ultrasound,the edge information of the ultrasound image is added to the constraint of the reconstruction process,which retains the advantages of DBT imaging in the X-Y plane and improves the image quality in the Z direction.By optimizing the algorithm of reconstruction process and carrying out effective registration and fusion of the composite pattern,it can make the reconstruction results convey internal information for DBT more accurately,which is very important in the DBT clinical application.Ultrasound images can be used to distinguish the image segmentation of the functional tissue,which is very important for the clinical diagnosis of breast cancer.However,manual segmentation of the complete three-dimensional ultrasound body is difficult,and there are internal observation differences between the different radiologists.In this paper,we present an automatic segmentation algorithm to segment all major tissue types of the 3D breast ultrasound body,which assists the correction in breast cancer images and applies pulse echo and sound velocity imaging method to assist diagnosis and analysis.The experimental results show that the proposed method can not only correctly distinguish between fat and non-fat tissue,but also better demonstrate the classification of cysts.Compared with manual segmentation for multiple times,the automatic segmentation shows a good continuity,which helps to locate the distorted tissue,simplify the working process of the image segmentation as well as accelerating the diagnosis of 3D ultrasound image.In addition,the proposed segmentation algorithm can provide breast imaging report and potential important information of the data system,including breast density,tumor size,shape and edge,ultrasonic echo pattern and tumor location characteristics.This method has a good application value in the assisted diagnosis of breast cancer.