Classification of Cervical Cancer using Ultrasound Images

This research introduces on methods and result of a study aiming at developing healthy tool to estimate and categorize ultrasound images of cervical cancer. Using the ultrasound images acquired from the radiological sciences laboratories in Kanyakumari Medical Mission C.S.I.Hospital, Neyyoor. This study aims to categorize cervical ultrasound images based on four well known mined features such as the relative size of detected region and region of interest, first four moments of intensities of detected region and region of interest, localization radius of nuclei and area of the abnormality. In this study, the cervical ultrasound images are categorized into three types:1) normal 2) pre cancer 3) malignant. The final system will take as input an ultrasound image of the cervix containing the epithelium layer and provide classification using the new automated approach, to assist the radiologist in cervical cancer diagnosis.