Early Diagnosis of Emphysema using Convolutional Neural Networks

Abstract

This study’s goal was to create and assess a deep learning model for emphysema identification in chest Radiographs images using the NIH Chest Radiographs dataset. For the management and treatment of emphysema, a chronic lung illness that can cause severe morbidity and death, early identification and diagnosis are essential. Chest Radiographs are often used to diagnose emphysema and deep learning models have been successful in improving the accuracy and efficiency of this process. A study used a convolutional neural network (CNN) to categorize chest Radiographs images as either normal or indicative of emphysema. A portion of the dataset was utilized for training the model, while a different test set was used for assessment. The dataset underwent preprocessing to guarantee consistency and accuracy in the labelling of emphysema. The study’s findings showed that the CNN model, with an overall accuracy of 83%, was very accurate in identifying emphysema in chest Radiographs images. These findings demonstrate how deep learning models may enhance the precision and effectiveness of emphysema diagnosis.

Topics & Concepts