Medical image retrieval by color content

We develop an indexing scheme for medical images. In general, for a given medical image, there is one object which is clinically important amongst the rest. We name this object the dominant object. Our proposed index is composed of three parts: (1) the dominant objects in images are standardized; (2) each image will have its own quadtree constructed which depends on the color composition and the color variation in the image; and (3) an R-tree that supports the retrieval of color images. To demonstrate the effectiveness of the index developed, we use images of skin lesions as the image data. Our initial experiments gives promising results for image retrieval.