The functionality of a CADx scheme comprises not only the investigation of nodules and microcalcifications, but also a detailed analysis able to characterize all signals present in mammographic image. Some secondary signals are can also provide important diagnostic information, such as: breast tissue asymmetry, nipple retraction and axillary lymph node. They can provide more detail on the breast characteristics, aiding the case interpretation. Thus, as part of a CADx scheme under development, this work aims the development of a module that characterizes automatically asymmetries, nipple retraction and axillary lymph nodes in digital mammographic images, through computer vision and artificial intelligence techniques. The purpose is that such a module could help the CADx user in attempting to other signals and information – besides nodules and/or clustered microcalcifications – which better describe the breasts under investigation, aiming to improve the diagnosis accuracy.