A Semantic Segmentation of Nucleus and Cytoplasm in Pap-smear Images using Modified U-Net Architecture

Abstract

Pap-smear images can help early detection of cervical cancer, but the manual interpretation by a pathologist can be time-consuming and prone to human error. Semantic segmentation of the cell nucleus and cytoplasm plays an essential role in Pap smear image analysis for the detection of cervical cancer automatically. This study proposes a modified U-Net architecture by adding batch normalization to each convolution layer. Batch normalization aims to stabilize and accelerate the convergence of the model during training, thus overcoming the vanishing gradient problem. The modified U-Net model achieves high accuracy and low loss during the training process, indicating its ability to learn and recognize patterns in the data. The performance evaluation of the model resulted in 91.4 % accuracy, 79.9 % sensitivity, 87.7 % specificity, 81.7 % F1-score, and 83.7 % precision. The results show that the proposed modification of U-Net architecture with batch normalization improves the segmentation performance for cervical cancer cells in Pap smear images. However, improvement in architecture is still required to increase the ability to overcome overlapping areas between the nucleus, cytoplasm, and background.