įurthermore, CXR can be used as a tool for diagnosing a number of diseases and complications, such as thoracic diseases, fractures, tooth decay, infections, osteoporosis, enlarged hearts, blocked blood vessels, etc. Deep neural network-based models have been successful in learning the discriminative features in image-based disease classification tasks such as tuberculosis detection and lung disease classification in radiographs and lung nodule classification in CT scans. It was first used to detect and classify COVID-19 and viral pneumonia. We used the CXR dataset compiled by Rahman, which is freely available for research purposes. Most of the systems for lung disease classification are either stand-alone or cloud-based. However, there is a demand to utilize the power of edge computing for disease detection. Much work has been done in the recent past. The detection of lung diseases has gained a lot of popularity due to the prevailing COVID-19 spread. Computed tomography (CT) scan and CXR imaging are two very commonly used diagnostic techniques used for the detection of lung diseases. It has applications in disease diagnosis, classification, and prognosis. Machine learning is very useful in healthcare informatics. Two feature-based ML classification techniques, namely support vector machine with local binary pattern (SVM + LBP) and decision tree with histogram of oriented gradients (DT + HOG) yield an accuracy of 87.98% and 86.87%, respectively. Further, a DL model, trained from scratch, gives an accuracy of 92.43%. The accuracies of other TL models, namely SqueezeNet, VGG19, ResNet50, and MobileNetV2 are 97.33%, 91.66%, 90.33%, and 76.00%, respectively. The proposed technique gives an average accuracy of 98.66%. Comparative analysis has been given in various metrics such as precision, recall, accuracy, AUC-ROC score, etc. The model is deployed in an edge environment using Amazon IoT Core to automate the task of disease detection in CXR images with three categories, namely pneumonia, COVID-19, and normal. A hybrid Inception-ResNet-v2 transfer learning model coupled with data augmentation and image enhancement gives the best accuracy. A combination of the synthetic minority over-sampling technique (SMOTE) and weighted class balancing is used to alleviate the effects of class imbalance. In the present study, a range of machine learning (ML), deep learning (DL), and transfer learning (TL) approaches have been evaluated to classify diseases in an openly available CXR image dataset. Furthermore, there remains an acute shortage of trained radiologists worldwide. However, even for expert radiologists, it is a challenge to accurately diagnose diseases from CXR samples. Chest X-ray (CXR) imaging is one of the most widely used and economical tests to diagnose a wide range of diseases.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |