No abstract

The article analyzes and compares the architectures of generativeadversarialnetworks. These networks are based on convolu-tional neural networks that are widely used for classification problems. Convolutional networks require a lot of training data to achieve the desired accuracy. Generativeadversarialnetworks are used for the synthesis of biomedical images in this work. Biomedi-cal images are widely used in medicine, especially in oncology. For diagnosis in oncology biomedical images are divided into three classes: cytological, histological, and immunohistochemical. Initial samples of biomedical images are very small. Getting trainingimages is a challenging and expensive process. A cytological training datasetwas used for the experiments. The article considers the most common architectures of generative adversarialnetworks suchas Deep Convolutional GAN (DCGAN), Wasserstein GAN (WGAN),Wasserstein GAN with gradient penalty (WGAN-GP), Boundary-seeking GAN (BGAN), Boundary equilibrium GAN (BEGAN). A typical GAN network architecture consists of a generator and discriminator. The generator and discriminator are based on the CNN network architecture.The algorithm of deep learning for image synthesis with the help ofgenerativeadversarialnet-worksis analyzed in the work. During the experiments, the following problems were solved. To increase the initial number of train-ingdata to the datasetapplied a set of affine transformations: mapping, paralleltransfer, shift, scaling, etc. Each of the architectures was trainedfor a certain numberof iterations. The selected architectures were compared by the training timeand image quality based on FID(FreshetInception Distance)metric. The experiments were implemented in Python language.Pytorch was used as a machine learning framework. Based on the used softwarea prototype software module for the synthesis of cytological imageswas developed. Synthesis of cytological images was performed on the basis of DCGAN, WGAN, WGAN-GP, BGAN, BEGAN architectures. Goog-le's online environment called Collaboratory was used for the experimentsusing Nvidia Tesla K80 graphics processor

Mo­dern da­ta­ba­ses of bi­ome­di­cal ima­ges ha­ve be­en in­ves­ti­ga­ted. Bi­ome­di­cal ima­ging has be­en shown to be ex­pen­si­ve and ti­me con­su­ming. A da­ta­ba­se of ima­ges of pre­can­ce­ro­us and can­ce­ro­us bre­asts "BPCI2100" was de­ve­lo­ped. The da­ta­ba­se con­sists of 2,100 ima­ge fi­les and a MySQL da­ta­ba­se of me­di­cal re­se­arch in­for­ma­ti­on (pa­ti­ent in­for­ma­ti­on and ima­ge fe­atu­res). Ge­ne­ra­ti­ve ad­ver­sa­ri­al net­works (GAN) ha­ve be­en fo­und to be an ef­fec­ti­ve me­ans of ima­ge ge­ne­ra­ti­on. The archi­tec­tu­re of the ge­ne­ra­ti­ve ad­ver­sa­ri­al net­work con­sis­ting of a ge­ne­ra­tor and a discri­mi­na­tor has be­en de­ve­lo­ped.The discri­mi­na­tor is a de­ep con­vo­lu­ti­onal neu­ral net­work with co­lor ima­ges of 128×128 pi­xels. This net­work con­sists of six con­vo­lu­ti­onal la­yers with a win­dow si­ze of 5×5 pi­xels. Le­aky Re­LU type ac­ti­va­ti­on functi­on for con­vo­lu­ti­onal la­yers is used. The last la­yer used a sig­mo­id ac­ti­va­ti­on functi­on. The ge­ne­ra­tor is a neu­ral net­work con­sis­ting of a fully con­nec­ted la­yer and se­ven de­con­vo­lu­ti­on la­yers with a 5×5 pi­xel win­dow si­ze. Le­aky Re­LU ac­ti­va­ti­on functi­on is used for all la­yers. The last la­yer uses the hyper­bo­lic tan­gent ac­ti­va­ti­on functi­on. Go­og­le Clo­ud Com­pu­te Instan­ce to­ols ha­ve be­en used to tra­in the the ge­ne­ra­ti­ve ad­ver­sa­ri­al net­work. Ge­ne­ra­ti­on of his­to­lo­gi­cal and cyto­lo­gi­cal ima­ges on the ba­sis of the ge­ne­ra­ti­ve ad­ver­sa­ri­al net­work is con­duc­ted. As a re­sult, the tra­ining sample for clas­si­fi­ers has be­en sig­ni­fi­cantly incre­ased.

No abstract

No abstract