B/S Based Vascular Data Analysis In Cloud Environment

As one common serious threat to human health, cardiovascular diseases have widely been paid attention, and the diagnosis and treatment of which have always been the focuses of medical research. Vascular data analysis system assesses cardiovascular diseases from the perspective of images analysis, and provides diagnostic evidence to a doctor. Traditional vascular data analysis system is high computational consuming and requires additional high-performance devices to help, which may lead to increased cost of the system. Cloud computing describes a new supplement, consumption and delivery model for IT services based on the Internet, which has high performance, expandability, information sharing, storage and other features to solve the problems caused by the traditional mode of realization. Cloud computing balances the pressure of vessel data analysis by clouds and provides the vessel analysis services from clouds, this pattern not only makes full use of advantages of the cloud platforms, but also reduces both burden on client-side and cost of whole system.Based on the above, a B/S based vascular data analysis system in cloud environment was proposed and implemented. The overall framework of vascular data analysis system was implemented firstly. The system adopted B/S architecture, the server-side used JSP/Servlet technology and the client-side used HTML5. Through the analysis of the data flow and the division of system function module, the medical image analysis, MPR,3D visualization, vessel analysis function and so on were realized on the server side. The image rotation, window/level adjustment, CT value acquirement and other functions were realized on the client side. Through the client and server data interaction, the whole system was in possession of some fundamental functions such as data manipulation, image processing and visualization.Base on the system framework above, a further study on the vessel segmentation and vascular quantitative analysis was carried out. In the vessel segmentation, boneless, Hessian matrix multiscale enhancement, connectivity judgment and closed operation were studied to extract the vessels; In vascular quantitative analysis, the vessel skeleton was extracted by means of topological thinning firstly; By analyzing the characteristics of the endpoints, nodes, branchs in the vessel skeleton, the number of endpoints, nodes and branches were calculated; The vessel diameter was calculated by distance transform method; The tortuosity was acquired by the rate of change through the vessel tangent vectors. And vascular health status was discussed by the parameters above.The basic image processing part of the above-described system has been implemented in hospital with high efficiency and good display effect, and it can meet the daily requirements of physicians; Vessel analysis function is tested with brain vascular CTA image data. The results of analysis and display are correct.