2518-1092Research result. Information technologies2518-109210.18413/2518-1092-2020-5-1-0-31983INFORMATION SYSTEM AND TECHNOLOGIESINCREASING ACCURACY CLASSIFICATION OF X-RAY IMAGES USING TRAINING OF COMPOSITE NEURAL NETWORKINCREASING ACCURACY CLASSIFICATION OF X-RAY IMAGES USING TRAINING OF COMPOSITE NEURAL NETWORKVygoniailoVictor RomanovichVygoniailoVictor Romanovich1078978@bsu.edu.ruMikhelevVladimir MikhailovichMikhelevVladimir Mikhailovichvm.mikhelev@gmail.com20205100

This article is devoted to solving the problem of classifying chest x-ray images by using the retraining of a pre-trained convolutional neural network trained on small data sets. A trained binary classifier is used to detect the presence or absence of lower respiratory tract pathology. The paper presents the results of a computational experiment and shows an improvement in accuracy in solving the classification problem. The study aims to identify a qualitative improvement in the accuracy index when using a composite neural network.

This article is devoted to solving the problem of classifying chest x-ray images by using the retraining of a pre-trained convolutional neural network trained on small data sets. A trained binary classifier is used to detect the presence or absence of lower respiratory tract pathology. The paper presents the results of a computational experiment and shows an improvement in accuracy in solving the classification problem. The study aims to identify a qualitative improvement in the accuracy index when using a composite neural network.

image classificationchest x-rayscomposite neural networksneural networksconvolutional neural networkskerastensorflowimage classificationchest x-rayscomposite neural networksneural networksconvolutional neural networkskerastensorflow