A medical image compression and decompression technique is presented which exploits the special characteristics of medical images so as to increase the achievable compression ratio over existing generic image compression techniques. In general, the presented technique categorizes medical images based on the type of images. Medical images within the same category will typically have a very high level of similarity to each other. For each category, a type of standard image is computed which represents the typical characteristics of images within a category. For each image being compressed, only the difference between the image and the standard image is compressed. Due to the high level of similarity between images in the same category, the aforementioned difference is typically small and therefore a high compression ratio can be achieved.