Computer-Aided Diagnosis System for Cervical Cancer using Feature Analysis

Abstract

Early detection through examination is very important for a medical treatment of cervical cancer, because a survival rate is improved markedly by curing at an initial stage. Ministry of Health, Labor and Welfare of Japan recommends it to receive cervical examination for a woman 20 years or older, but becomes around 20% of examination rate at the present. For the number of patients of cervical cancer tending to increase year by year, it is a problem that the burden of cytotechnologists increases, and the possibility of misdiagnosis becomes high. Therefore the study of a computer-aided diagnosis system using an image processing attracts attention. We make the algorithm to diagnose the malignancy of HSIL and LSIL, in this study. In addition, we build GUI so that it is easy for cytotechnologists to use the proposed diagnosis support system.