Breast Tumor Segmentation and Classification Using Ultrasound Images

Abstract: Breast cancer is one of the leading causes of death among women. Early detection is crucial for successful treatment and better patient outcomes. Although they have benefits, mammograms and other traditional methods have disadvantages. Even though it could be challenging to tell benign from malignant tumors, ultrasonography offers an extra technique. This work explores a new method for detecting breast cancer based on ultrasound images. It makes use of machine learning techniques, particularly deep learning, to analyze these images. This method consists of two steps: segmentation and classification. Classification determines whether the tumor is benign or malignant, while segmentation helps to focus the image's suspected tumor to a specific area.

Keywords: Ultrasonic imaging, Deep learning, Machine learning techniques, Segmentation, Classification, Early diagnosis, Successful therapy, Breast tumour