ІНФОРМАЦІЙНА ТЕХНОЛОГІЯ ВІЗУАЛЬНОГО АНАЛІЗУ РЕНТГЕНІВСЬКИХ ЗОБРАЖЕНЬ ДЛЯ ІНТЕРПРЕТАЦІЇ РЕЗУЛЬТАТІВ ДІАГНОСТУВАННЯ ПНЕВМОНІЇ

Abstract

To date, pneumonia is one of the most common and severe lung diseases in the world. Early diagnosis of pneumonia is a crucial factor in its successful treatment. Over the last decade, automated analysis of chest X-rays has been recognized as an effective tool for diagnosing lung diseases. However, the problem of implementing and configuring methods that explain the results of digital diagnosis remains acute. Convolutional neural networks now show state-of-the-art results in the identification of diseases on X-ray. Therefore, to address the urgent issue in digital diagnosis, we propose information technology for visual analysis of X-ray images to explain the results of diagnosing pneumonia. The technology comprises a classification model based on a convolutional neural network to remove mild features of early viral pneumonia and a modified method of different localization to interpret the classification results. The method of interpretation is to apply weighted gradients to class activation maps. It distinguishes lung masks in the X-ray image and imposes thermal maps with a color gradient from blue to bright red. The red color corresponds to the most probable location of the pneumonia features in the radiograph. Such a modification provides excellent localization of abnormal areas on radiographs, removing the mild target features of early pneumonia. It should be noted that our model based on the convolutional network surpassed other classifiers in precision (98.5%) but slightly conceded in classification accuracy (96.1%) and recall (93.6%). Also, it shows relatively low false positive and false negative rates, with 1.4% and 6.4%, respectively. Overall, according to computational experiments, the proposed information technology can be an effective tool for instant diagnosis in the first suspicion of pneumonia.