The chest X-ray is among the most widely used diagnostic imaging for diagnosing many lung and bone-related diseases. Recent advances in deep learning have shown many good performances in disease identification from chest X-rays. But stability and class imbalance are yet to be addressed. In this study, we proposed a CX-Ultranet (Chest X-ray Ultranet) to classify and identify thirteen thoracic lung diseases from chest X-rays by utilizing a multiclass cross-entropy loss function on a compound scaling framework using EfficientNet as a baseline. The CX-Ultra net achieves 88% average prediction accuracy on NIH Chest X-ray Dataset. It takes ≈ 30% less time than pre-existing state-of-the-art models. The proposed CX-Ultra net gives higher average accuracy and efficiently handles the class imbalance issue. The training time in terms of Floating-Point Operations Per Second is significantly less, thus setting a new threshold in disease diagnosis from chest X-rays.
CITATION STYLE
Kabiraj, A., Meena, T., Reddy, P. B., & Roy, S. (2022). Detection and Classification of Lung Disease Using Deep Learning Architecture from X-ray Images. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 13598 LNCS, pp. 444–455). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-20713-6_34
Mendeley helps you to discover research relevant for your work.