Automatic Classification of Cervix Type

In a developed country like United States, cervical cancer rate has been improved by 70% in the last 40 years due to a good screening programmes implemented in their country. However, it is not happening in those rural areas due to lack of expertise in this field. Due to that, in this study, few algorithms were developed by using image classification methods to correctly classify the cervix types based on cervical images by using segmentation and classification method. There are three different types of cervix, these different types of cervix in our data set are all considered normal (not cancerous, but some are pre-cancerous), but since the transformation zones are not always visible, some of the patients require further testing while some do not. This decision is very important for the healthcare provider and critical for the patient. Identifying the transformation zones is not an easy task for the healthcare providers, therefore, an algorithm-aided decision will significantly improve the quality and efficiency of cervical cancer screening for these patients. In this experiment, the SVM method of classification was used to obtain an accuracy of 76.2%.