Neutrosophy is a branch of philosophy introduced by Florentin Smarandache. Neutrosophic set (NS) is the derivative of neutrosophy; it is a powerful tool to handle uncertainty. Here we applied neutrosophic set to gray scale image domain for image analysis. Several authors contributed in neutrosophic image analysis and image processing. We propose a novel approach on representation of grayscale images in bipolar neutrosophic domain (BNS). The reduction of noise in images is one of the challenging task in every field. While we transform a grayscale image into bipolar neutrosophic domain, the indeterminacy degree of both positive and negative memberships are reduced significantly. Indeed, we extract some useful information from indeterminacy domain; it leads to perform image analysis and processing in noisy images in a better manner. We discuss the representation of medical images in bipolar neutrosophic domain with examples.