Cervical Cancer Screening Approach Using AI
Desak Gde Diah Dharma Santhi, M. Carmel Sobia, M. Jayalakshmi · Applied Artificial Intelligence · 2023
The powerful tool of artificial intelligence is computer vision. Computer vision is an arena of computer science. It enables computers to identify and process images. The driving mechanism behind the development of computer vision is that the processing of data we generate today are used to train and improve computer vision. Cervical cancer is a widespread disease among women all over the world. Cervical cancer occurs in the cervix present in between the uterus and vagina in the female reproductive system. Abnormal changes in the cells lead to cancer. Identifying the abnormal cells in the image is necessary to diagnose cancer. A study of the standard cell is essential to distinguish it from affected cells. This chapter concentrates on discriminating normal cervical cells as superficial, intermediate, and parabasal cells, such that this work would help precisely differentiate normal cells from abnormal cells. Here, the segmentation of the nucleus and cytoplasm from the single cervical cell is done using the threshold and cluster method on two types of datasets. Thirty-one features are extracted from the nucleus and cytoplasm for classification, and only three features are selected based on the statistical analysis. A feed-forward backpropagation neural network is used to classify normal cervical cells, and classification accuracy is 100%.