Pap-Smear image generation using math model Pap-Smear image generation using generative adversarial networks Comparative analysis for image generation The subject of automatic detection of the presence of cervical cancer by evaluating histopathological Pap-Smear images with computerized diagnosis / detection software is an active field of study. The reason for this is that the objects (cell nucleus, cytoplasm, white blood cell, bacilli, and speckle) in the visuals overlap and change the geometric structure and pattern of each other, they are dispersed in the image with different density, and the noise patterns are different. In addition, the difficulty, and costs of creating a tagged large dataset prevented the emergence of a common dataset in this area. The mentioned difficulties negatively affect the achievements in current classification studies and trigger the need for new approaches. In this paper, a threestep approach based on building large Pap-Smear datasets using Generative Adversarial Networks (GANs) is proposed. Accordingly, in the first step, geometric shape, and pattern models of each object structure in Pap-Smear images are created. In the second stage, synthetic Pap-Smear images (Ground True) with the desired number and distribution of objects are produced using the produced parametric models. In the third stage, the performances of existing GANs (Pix2Pix, CycleGAN, DiscoGAN and AttentionGAN) to produce GT are evaluated and a solution-oriented new current GAN architecture (Pix2PixSSIM) is proposed. Experimental studies show that a large Pap-Smear data set can be produced in a very short time with the proposed GAN architecture. In this way, it is seen that deep networks with high classification success can be trained.Figure A. The flowchart of the proposed method Purpose: In this study, a new GAN model for histopathological Pap-Smear images generation is suggested. To illustrate the advantage of the proposed GAN model (Pix2PixSSIM), a comprehensive experimental study has been carried out. Theory and Methods: Pix2PixSSIM designed as generative adversarial networks model for histopathological Pap-Smear images generation. In addition, the proposed model compared with the existing GANs (i.e., Pix2Pix, CycleGAN, DiscoGAN and AttentionGAN).
Results:In experimental studies, Pix2PixSSIM, which is designed as generative adversarial networks model for histopathological Pap-Smear images generation, has shown high accuracy than other methods.
Conclusion:It is seen that the performance of the proposed GAN architecture to produce patterns similar to real Pap-Smear visuals gives successful results (MSI=23.649, PSNR=37.476) when compared to existing approaches (MathModel and classical image synthesis methods).
