Synthesis of Microscopic Cell Images Obtained from Bone Marrow Aspirate Smears through Generative Adversarial Networks

Abstract

Every year approximately 1.24 million people are diagnosed with blood cancer. While the rate increases each year, the availability of data for each kind of blood cancer remains scarce. It is essential to produce enough data for each blood cell type obtained from bone marrow aspirate smears to diagnose rare types of cancer. Generating data would help easy and quick diagnosis, which are the most critical factors in cancer. Generative adversarial networks (GAN) are the latest emerging framework for generating synthetic images and time-series data. This paper takes microscopic cell images, preprocesses them, and uses a hybrid GAN architecture to generate synthetic images of the cell types containing fewer data. We prepared a single dataset with expert intervention by combining images from three different sources. The final dataset consists of 12 cell types and has 33,177 microscopic cell images. We use the discriminator architecture of auxiliary classifier GAN (AC-GAN) and combine it with the Wasserstein GAN with gradient penalty model (WGAN-GP). We name our model as WGAN-GP-AC. The discriminator in our proposed model works to identify real and generated images and classify every image with a cell type. We provide experimental results demonstrating that our proposed model performs better than existing individual and hybrid GAN models in generating microscopic cell images. We use the generated synthetic data with classification models, and the results prove that the classification rate increases significantly. Classification models achieved 0.95 precision and 0.96 recall value for synthetic data, which is higher than the original, augmented, or combined datasets.

Automatic generation of artificial images of leukocytes and leukemic cells using generative adversarial networks (syntheticcellgan)

“…The first test of the system was the evaluation of the quality and diversity of the artificial cells. The results of the quantitative evaluation in Table 2 are comparable to the quantitative tests in the papers [36] and [37].…”

“…Very few works have so far been reported in the literature on the generation of microscopic cell images using adversarial networks. As far as the authors know, the closest to our work are [36] and [37]. Four classes of blood leukocytes are synthesized in [36], with no basophils.…”

“…The FID between the real images and artificial images generated by SCG is in the range of 75.766 to 12.918, the lower the better. The value of this metric in [37] is between 76.3 and 67.2 when comparing various approaches for generation of microscopic cell images. It is particularly interesting to observe (see Table 2) the difference between the best FID scores (13.921 and 12.918 for the synthetic images of neutrophils and basophils, respectively) versus the worst highest score (75.766 and 47.34 for synthetic images of lymphocytes and monocytes, respectively).…”

“…Extending the application of GANs to other types of cells is a complex task due to the specific morphological characteristics of each cell class. The work of [37] focuses on bone marrow aspirate smears. It uses a Wasserstein GAN with gradient penalty for the creation of various types of cells with a low resolution of 64 x 64 pixels.…”

“…The work in [37] synthesizes cell images obtained from bone marrow aspirate smears. The target is the five normal leukocyte classes plus erythroblasts, immature granulocytes, monoblasts, myeloblasts, normal promyelocytes and platelets.…”