Pneumonia Diagnosis Using Chest X-ray Images and Machine Learning
Sara Lee Kit Yee, Wong Jee Keen Raymond
Abstract
In this paper, a pneumonia diagnosis system was developed using convolutional neural network (CNN) based feature extraction. InceptionV3 CNN was used to perform feature extraction from chest X-ray images. The extracted feature was used to train three classification algorithm models to predict the cases of pneumonia from a Kaggle dataset. The three models are K-Nearest Neighbor, Neural Network, and Support Vector Machines. Performance evaluation and confusion matrix were presented to represent the sensitivity, accuracy, precision, and specificity of each of the models. Results show that the Neural Network model achieved the highest sensitivity of 84.1%, followed by Support vector machines (83.5%) and K-Nearest Neighbor Algorithm (83.3%). Among all the classification models, Support vector machines model achieved the highest AUC of 93.1%.