X-rays, MRIs, and other medical images are distributed for the purposes of research and diagnosis. Often these images are distributed with critical personal information, but is it necessary to share that information? The goal of this presentation is to expand on the work of previous researchers in describing an automatable solution to hide personal information on medical images, specifically MRI images of the brain. This algorithm applies the Haar wavelet transform to color or grayscale medical images to locate text objects, embed the details within the image, and make the text illegible to unauthorized users. Application of this algorithm can produce an automated system for processing and distributing medical images which better protects the privacy of patients and relieves researchers and others from liability.
