Pneumonia Detection Using Chest X-Ray Images Through Deep Learning Technique

Abstract

Pneumonia is a type of infection disease that cause high deaths from worldwide. In addition, most of the patients that got infected are children and people aged five and above. Besides, a timely diagnosis is essential to treat the disease due to which delay in care contributes a lot to death. In addition, doctors assess chest X-rays for diagnosis of the disease. Furthermore, these X-Rays need the medical staff to be trained extensively for proper reading. Moreover, this specialist may not be available in a low-resource setting. Despite having a training program, the problem of availability continues to be a challenge. High accuracy in deep learning networks was revealed in recent research through effective pattern detection. It’s been seen that these models are focusing on the features which are highly predictive of disease. The proposed system is used in this paper for development of framework for development of pneumonia detection using chest X-ray images. The suggested system involves the use of convolution neural network for feature extraction and classification. The architecture of the network along with a pseudo-code and details concerning the data pre-processing steps and method. Also, the article discusses different metrics to assess the results obtained through the experiments. Furthermore, it is mentioned about the limitations of the data and other ethical concerns regarding implementation. According to the results obtained, an AI system can be an excellent aid to the medical facilities for quickly screening the patients and reliable results. That’s because of the risks of wrong results but, it is only to be used as an addition.