Image-Guided Telemedicine System via the Internet

摘要

This paper presents an Internet software application that is readily available and that can greatly affect practice. Any medical facility with Internet access can use this system to visually diagnose breast fine needle aspirates since the program is implemented as a Java applet and accessed remotely via the World Wide Web. The proposed system automatically indexes objects based on shape and groups them into a set of clusters, or prototypes. Queries are first compared to the set of prototypes and then to the objects in the most similar prototypes. In terms of dissimilarity, each query is performed based on shape, or possibly in combination with other features as size, radius, perimeter, area, smoothness, compactness, concavity, and concave points. Experimental results show that the system achieved satisfactory performance.