Deteksi Kanker Serviks Berdasarkan Citra Sel Pap Smear dengan Klasifikasi Naïve Bayes
Abstract
One cause of the high mortality rate in patients with cervical cancer is late medical treatment. It can also be caused by the manual detection of cervical cancer that is ridden by potential procedural or human errors. One of the solutions that can be done is automating the detection of cervical cancer using pap smear cell image. This study uses the Naïve Bayes method to identify normal and abnormal cells and group them into seven more specific classes. The data used is single cell images from pap smear. This research focuses on a classification method and does not perform segmentation and feature extraction process. The results showed 97% accuracy for two-class classification and 9% for the seven-class classification. The low accuracy in classification of seven classes is caused by overlapping on the features used. Naïve Bayes gives satisfactory results on the two-class classification but for the seven-class classification this method has not given good results.
Collections
- UT - Computer Science [2322]