Magnetic Resonance Imaging (MRI) results in overall quality that usually calls for human intervention in order to correctly identify details present in the image. More recently, interest has arisen in automated processes that can adequately segment medical image structures into substructures with finer detail than other efforts to-date. Relatively few image processing methods exist that are considered accurate enough for automated MRI image processing where the edge-to-pixel ratio is relatively high, largely due to the inherent pixel noise. ANN training, though ideal for non-linear solutions, is generally considered inefficient for most image processing operations based on the limitations of the most commonly known training algorithms and their derivatives. We present a rapid and accurate method for segmentation of a cerebral cortex image using a unique ANN training algorithm that most notably handles the very large sets of associated training patterns (one per pixel) inherent in an image file. This method also operates on intensity image data converted directly from raw RGB MRI images without any pre-processing.