Dual lookup table algorithm: An enhanced method of displaying 16-bit gray-scale images on 8-bit RGB graphic systems

Abstract

Most digital radiologic images have an extended contrast range of 9 to 13 bits, and are stored in memory and disk as 16-bit integers. Consequently, it is difficult to view such images on computers with 8-bit red-green-blue (RGB) graphic systems. Two approaches have traditionally been used: (1) perform a one-time conversion of the 16-bit image data to 8-bit gray-scale data, and then adjust the brightness and contrast of the image by manipulating the color palette (palette animation); and (2) use a software lookup table to interactively convert the 16-bit image data to 8-bit gray-scale values with different window width and window level parameters. The first method can adjust image appearance in real time, but some image features may not be visible because of the lack of access to the full contrast range of the image and any region of interest measurements may be inaccurate. The second method allows “windowing” and “leveling” through the full contrast range of the image, but there is a delay after each adjustment that some users may find objectionable. We describe a method that combines palette animation and the software lookup table conversion method that optimizes the changes in image contrast and brightness on computers with standard 8-bit RGB graphic hardware—the dual lookup table algorithm. This algorithm links changes in the window/level control to changes in image contrast and brightness via palette animation. The purpose of the algorithm is to use palette animation to mimic changes in image appearance performed by the software lookup table method after the window width and window level parameters have changed. The algorithm combines the advantages of both methods: rapid manipulation of image brightness and contrast by palette animation, and the ability to window and level on the full 16-bit image data using the software lookup table. This algorithm may be useful for applications that display 16-bit radiologic images on computers with standard 8-bit RGB graphic systems.