Artifacts removal in NEVI medical images based on moving frame domain texture analysis

Abstract

This paper presents a procedure for user-assisted artifact removal from medical images, namely photographic images of nevi and melanomas. Specifically, we propose an artifact removal procedure based on image representation in the moving frame domain. This domain has been recently introduced in the literature for denoising purposes. We show that in this domain the artifacts are well distinguished from the image signal and can be removed; besides, interpolation of the removed points is highly effective since it preserves relevant image feature useful for diagnostic purposes.