An Efficient Approach for Diagnosis System of Cervical Cancer Using Image Fusion to Prevent Earlier Death in Women

Background: Cervical cancer is the second most common cancer in women around the world. It can be cured if the severity level of the cancer is in the category of mild and moderate. The surgery is required for severe case and earlier death can be prevented if timely treatment is stipulated. Methods: In this paper, a computer aided fully automatic methodology for cervical cancer detection using cervigram is proposed. The methodology has used image fusion technique to enhance the contrast level of the cervical image. Then, features are extracted and classified by Support Vector Machine (SVM) classifier to diagnose the cervical image into either normal or malignant. Results: The performance of the proposed cervical cancer diagnosing system is analyzed in terms of sensitivity, specificity and accuracy.