Design And Implementation Of The System For Ophthalmic Imaging Processing,Analysis And Visualization

With the development of information technology and the promotion of clinical needs,modern medical imaging technology has made breakthroughs and completely changed the traditional way of medical diagnosis and treatment.In this era of digital medical imaging,the use of computing technology and software technology for medical image processing,analysis and visualization as well as intelligent diagnosis of diseases has been one of the research hotspots and a major trend.In the field of ophthalmic imaging,some imaging techniques such as optical coherence tomography(OCT),ocular fundus photograph,etc.have been widely applied to provide more valuable information for ophthalmologists in clinical diagnosis,treatment of ophthalmic diseases and curative effect evaluation.The requirement of the system for ophthalmic imaging processing,analysis and visualization is also increasingly urgent which facilitates the researchers and medical staff to deal with image and analyze data quantitative.According to the actual application requirements,we use both medical science and computer technology to design and implement the cross-platform software system specially oriented to ophthalmic images processing,analysis and visualization.The design and implementation of the system mainly include:1,The architecture of the cross-platform system.We adopt the hierarchical and modular design and the three main plugin modules work in cooperation under the plugin-in mechanism.Low coupling and high cohesion are achieved to ensure the expansibility of the system;2,The data structure of the ophthalmology image and the unified data I/O solution.For the representation and visualization of ophthalmologic data,we define and implement several standard data types by encapsulating basic data structures.In order to solve the problem of files input/output/transmission between different ophthalmic equipments,the system provides the analytic methods for image files from ophthalmic devices manufactured by various venders;3,2D/3D Retinal image processing.By using the algorithms of graph-cut,graph-search,etc.,the system provides the functions of filtering and segmentation especially for retina images.The system supports the the segmentation and layering of histological structures in retina images such as retinal 11-layer segmentation,the anterior lamina cribrosa surface segmentation,etc.;4,Automatic detection of some retina diseases.By using the machine learning algorithms,we design and implement the detection and segmentation of macular hole and cystoid edema,and implement that of other ocular diseases including pigmented epithelial detachment,fundus micro-aneurysms,etc.;5,Data statistics and analysis.We implement the functions of automatic measuring of important parameters in the optical nerve head area and tracking of change in lesion area to assess the ocular health accurately;6,Visualization and interaction of ophthalmology data.It supports real-time displaying multiple ophthalmic data in multi-windows and multiple view modes by applying various rendering methods and provides various interactive tools to assist people in easy using.We build the system and operate and test the system in the actual environment,the analysis results show that the system can assist the doctors and researchers to deal with and analyze data,facilitate the work flow,reduce human error and assist in disease diagnosis.The results verify the effectiveness and reliability of the system,and the user's functional and performance requirements have also been achieved.