The article analyzes the current state of research and existing solutions in the field of recognizing respiratory diseases from functional diagnostic images. The solution of the problem of recognizing respiratory diseases such as cardiomegaly, pleural thickening and pulmonary fibrosis from chest X-rays is presented. The end user of the proposed software is a medical professional - radiologist. The developed software application for recognizing respiratory diseases from X-ray images consists of two interacting parts: user interface and neural network. The information system uses a convolutional neural network to solve the recognition problem. The Python programming language is used to create the convolutional neural network using TensorFlow and Keras libraries. As a dataset for training and testing of the neural network, an open set from the Kaggle website is chosen. The performance analysis of the developed application showed an accuracy of 92% in recognizing respiratory diseases by the trained model. The maximum time of recognizing a chest X-ray image took 1 min and the minimum time 20 s, while a radiologist needs from 10 to 20 min to decipher one X-ray image.