Research About Web-based Medical Image Three-dimensional Visualization And Processing

Based on current clinical requirements of medical imaging information, the developmental trends of internet plus medical treatment and our lab’s long-term accumulated work in the field of medical image 3D visualization, this paper carried out design and key technology research of web-based medical image 3D visualization and patient medical information 3D visualization system, which can be run under cloud computing environment. We developed 3D quantitative diagnosis component of lung/liver/brain tumor through international cooperation and intracerebral hemorrhage localization 3D visualization algorithm and software component independently. The main contents and innovations of this paper are as follows:(1) This paper have studied and designed the prototype of web-based medical image 3D visualization system, which possesses the new technical characteristics of medical image 3D visualization such as implementing medical image storage, processing, analysis and visualization computing in the clouds(server side) while performing 3D image display operations on the client side. This system consists of a graphical user interface, the client/server communication layer and the server application layer. The graphical user interface is implemented using HTML5 which has many worthwhile benefits such as thin client, zero footprint, cross platform. The client/server communication layer is based on WebSocket protocol, which can enhance user experience through full-duplex communication. The server application layer adopts GPU accelerated technique to optimize 3D visualization algorithms so as to respond quickly to client requests.(2) This paper designed a 2D / 3D image processing integration middleware, which can integrate 3D visualization features developed in this paper with clinical PACS 2D visualization system(AxVS) seamlessly. This middleware is developed using Java and Python programming language. And it can make Ehcache collaborate closely with PostgreSQL database, which can reduce the burden of image database and enhance user experience.(3) By migrating from web-based medical image 3D visualization to patient medical information visualization index system VIS, we conducted exploratory research on web-based patient medical information 3D visualization technique, which performs 3DS visualization human body model parsing and rendering in the clouds(server side) and transfers each frame to the client browser for display. The graphical user interface supports a series of 3D operations, such as rotate, zoom, pan and picking.(4) This paper has developed the 3D quantitative diagnosis component of lung/liver/brain tumor through international cooperation and intracerebral hemorrhage localization algorithm and software component independently. By integrating this two components with our lab’s medical image 3D visualization components MedViewCtrl, we conducted preliminary clinical testing and evaluation. The 3D quantitative diagnosis component of lung/liver/brain tumor is aiming at lung/liver/brain tumor 3D quantitative diagnosis. The user firstly selects the lesion ROI, then calls the relevant 3D segmentation algorithm, and finally reconstructs tumor model through 3D visualization algorithm. This can help doctors make a diagnosis by providing a series of quantitative indicators(such as tumor size, density, etc.). Intracerebral hemorrhage localization algorithm and software component helps physicians to achieve localization of cerebral hemorrhage. The user finds the interesting aspect of the cerebral hemorrhage with the help of MPR, and then calculates the cranial radius, frontal distance, temporal distance and height through measurement widgets provided by this component, which help doctor rapidly rescue patients with cerebral hemorrhage.