Deep learning-based algorithms provide an efficient and reliable diagnosis for medical imaging. This paper proposes COVID-19 diagnosis based on analysis of Computerised tomography (CT) chest scans. In recent years, deep learningbased analysis of CT chest scans has demonstrated competitive sensitivity for pneumonia prognosis. This paper presents our submission for the 2021 ICASSP Signal Processing Grand Challenge (SPGC). We exploit a 3D Networkbased transfer learning approach to classify volumetric CT scans with a novel pre-processing method to render the volume with salient features. This work uses the pre-trained 3D ResNet50 as the backbone network. The 3D network is trained on a dataset consisting of 3 classes: Community Acquired Pneumonia (CAP), COVID-19 and Normal patient. The final testing results have shown an overall accuracy of 85.56% with the COVID-19 sensitivity attaining 82.86%.
Coronavirus disease 2019, commonly known as COVID-19, is an extremely contagious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Computerised Tomography (CT) scans based diagnosis and progression analysis of COVID-19 has recently received academic interest. Most algorithms include two-stage analysis where a slice level analysis is followed by the patient-level analysis. However, such an analysis requires labels for individual slices in the training data. In this paper, we propose a single-stage 3D approach that does not require slice-wise labels. Our proposed method comprises volumetric data pre-processing and 3D ResNet transfer learning. The pre-processing includes pulmonary segmentation to identify the regions of interest, volume resampling and a novel approach for extracting salient slices. This is followed by proposing a region-of-interest aware 3D ResNet for feature learning. The backbone networks utilised in this study include 3D ResNet-18, 3D ResNet-50 and 3D ResNet-101. Our proposed method employing 3D ResNet-101 has outperformed the existing methods by yielding an overall accuracy of 90%. The sensitivity for correctly predicting COVID-19, Community Acquired Pneumonia (CAP) and Normal class labels in the dataset is 91.8%, 80% and 100%, respectively.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.