Research On Medical Ultrasound Image Reconstruction For 3D Bioprinting

3D printing technology can quickly produce tissue/organ models for diagnosis,preoperative simulation or surgical implants.However,the STL data for 3D printing based on segmentation and reconstruction of medical image data affects the precision of the printed model,resulting in inaccurate diagnosis and simulation.Using the volume rendering technique,it is not necessary to extract the isosurface from the original medical image data or to extract the contour information.The original data can be directly rendered to the three-dimensional image,and the detail information of twodimensional tomographic image is preserved,so the reconstructed image has high fidelity.In the process of reconstructing the scan data into STL model,the accuracy of segmentation can be greatly improved by basing on the results obtained from the volume rendering,which can improve the accuracy of the reconstructed solid model.In this paper,we study the surface rendering and volume rendering algorithms in 3D reconstruction algorithms,realize the suppression of speckle noise in ultrasound images,and construct a 3D reconstruction system based on VTK.The system realizes the denoising and 3D reconstruction of the 2D medical ultrasound images of the DICOM format,and completes the interactive operation of the reconstructed three-dimensional images,such as rotation,scaling and slicing.On the basis of above,the reconstruction of medical 3D image based on surface rendering algorithm is used to generation STL models.The comparison between the STL model and the volume rendering algorithm shows that the volume rendering images can preserve more image details and can provide reference for the construction of 3D printing model,in order to improve the accuracy of the reconstructed model.The development and research status of speckle suppression algorithm and 3D reconstruction algorithm of ultrasonic image are briefly reviewed.Combined with volume rendering technology,analyzes the feasibility of volume rendering technology to improve the 3D reconstruction and precision of printing in guidance segmentation,and a three-dimensional reconstruction system design scheme of ultrasonic image is determined.The statistical model of the speckle noise of the ultrasonic image is obtained based on the analyses of the causes and probability distribution to improve the weight function of non-local mean filter algorithm.The improved algorithms were used to filter the medical ultrasound images and the results show that the improved algorithm has better filtering effect than other algorithms,and is suitable for ultrasonic image speckle suppression.The algorithm of surface rendering and volume rendering in 3D reconstruction algorithm is studied,the research and simulation of marching cubes algorithm and ray casting algorithm are finished.The reconstruction effect of marching cubes algorithm,3D printed model and ray casting algorithm are compared and discussed.The results show that reconstruction effect of marching cubes algorithm and 3D printed model losses more details due to the segmentation algorithm or human subjective factors.The ray casting algorithm can retain more details,which can be helpful for medical diagnosis and biological 3D printing.A 3D reconstruction system of ultrasonic image is designed base on visualization toolkit,Visual Studio 2013 and MFC framework.The system realizes the reading and preprocessing of 2D ultrasonic sequence images in DICOM format,3D reconstruction and STL model generation based on marching cubes algorithm and ray casting algorithm,and can rotate,zoom and slice the reconstructed threedimensional image observation and other interactive operations.In summary,the three-dimensional model obtained by volume rendering can retain more detailed information,and is suitable to be used as a reference standard to guide the reconstruction of STL models and the printed 3D models of tissues or organs.