Using convolutional neural networks for binary classification of x-ray images

Autor:	S. R. Korzenevskiy, N. A. Komarov, A. A. Komarskiy
Rok vydání:	2021
Předmět:	Data set Binary classification business.industry Computer science ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION X ray image Pattern recognition Artificial intelligence business Convolutional neural network Task (project management) Image (mathematics)
Zdroj:	7TH INTERNATIONAL CONFERENCE ON X-RAY, ELECTROVACUUM AND BIOMEDICAL TECHNIQUE.
ISSN:	0094-243X
Popis:	Within this study, we made a comparison of the efficiency of several convolutional neural networks with different architectures taken from the Tensorflow open-source library. The paper analyzes the parameters of these networks using validation curves. The InceptionV3 model showed the best results for the X-ray image binary classification task. The selected convolutional network with optimal parameters obtained in this study is easy to be retrained using its own data set and applied to solve classification tasks, select and segment objects on X-ray images.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::8ce8aae241a0814258839b76b008c255 https://doi.org/10.1063/5.0052875 Zobrazit plný text záznamu