Computational Techniques for the Support of Breast Tumor Diagnosis on Ultrasound Images Miguel Alemán-Flores 1 , Patricia Alemán-Flores 2 , Luis Álvarez-León 1 , José Manuel Santana-Montesdeoca 2 , Rafael Fuentes-Pavón 2 and Agustín TrujilloPino 1 1 Departamento de Informática y Sistemas Universidad de Las Palmas de Gran Canaria, 35017, Las Palmas, Spain 2 Sección de Ecografía, Servicio de Radiodiagnóstico Hospital Universitario Insular de Gran Canaria, 35016, Las Palmas, Spain Abstract This work presents a new approach in the application of computer vision tech- niques to asses the diagnosis of solid breast tumors on ultrasound images. Most works related to medical image analysis for breast cancer detection refer to mammo- graphies. However, the studies carried out by the radiologists prove the significance of some aspects observed on ultrasound images, which are spiculation, calcifications, ellipsoid shape, taller than wide shape, echogenicity, capsule, angular margins, mi- crolobulations, two or three gentle lobulations, acoustic shadow and ramifications. We have developed a common framework for the analysis of such diagnosis crite- ria, so that they can be measured by the computer and a set of parameters are available for the physicans to decide whether the biopsy is necessary or not. The results produced by the system have been compared to those given by three spe- cialists to adjust the parameters and test their accuracy. The presence or absence of the diagnosis factors coincides for human and computer in most of the nodules. In some cases, the computer indicated the presence of suspicious factors which were not actually worrisome, but the most important aspect is avoiding overlooking the malignant nodules without needing carrying out a biopsy in all of them. We present a set of mathematical methods to extract objective evidence of the presence or ab- sence of the diagnosing criteria. This system is able to extract the relevant features for solid breast nodules with high accuracy and represents a very valuable help in the assessment of radiologists. 1 Introduction There are certain criteria which are used by the physicians to determine whether a solid breast nodule is benign or malignant. This work uses the computational techniques for the study and classification of shapes and textures to help the physicians in the early diagnosis of breast tumors on ultrasound images. Contrary to other works which deal with mammographies, this one is based on the analysis of ultrasound images, in which a series 1