Bayesian-based optimized deep learning model to detect COVID-19 patients using chest X-ray image data

Abstract

Coronavirus Disease 2019 (COVID-19) is extremely infectious and rapidly spreading around the globe. As a result, rapid and precise identification of COVID-19 patients is critical. Deep Learning has shown promising performance in a variety of domains and emerged as a key technology in Artificial Intelligence. Recent advances in visual recognition are based on image classification and artefacts detection within these images. The purpose of this study is to classify chest X-ray images of COVID-19 artefacts in changed real-world situations. A novel Bayesian optimization-based convolutional neural network (CNN) model is proposed for the recognition of chest X-ray images. The proposed model has two main components. The first one utilizes CNN to extract and learn deep features. The second component is a Bayesian-based optimizer that is used to tune the CNN hyperparameters according to an objective function. The used large-scale and balanced dataset comprises 10,848 images (i.e., 3616 COVID-19, 3616 normal cases, and 3616 Pneumonia). In the first ablation investigation, we compared Bayesian optimization to three distinct ablation scenarios. We used convergence charts and accuracy to compare the three scenarios. We noticed that the Bayesian search-derived optimal architecture achieved 96% accuracy. To assist qualitative researchers, address their research questions in a methodologically sound manner, a comparison of research method and theme analysis methods was provided. The suggested model is shown to be more trustworthy and accurate in real world.

Multi‐texture features and optimized DeepNet for COVID‐19 detection using chest x‐ray images

“…The various comparative methods exploited for the COVID‐19 assessment in comparison with the developed technique are Deep transfer learning, 25 multimodal covid network‐III (MMCOVID‐NET‐III), 41 Bayesian optimization‐based CNN, 42 Deep CNN, 26 Gayathri et al, 43 Auxiliary GAN, 27 and Deep LSTM, 8 randomly initialized CNN (RND‐CNN). 44 …”

COVID-DSNet: A novel deep convolutional neural network for detection of coronavirus (SARS-CoV-2) cases from CT and Chest X-Ray images

“…], and Covid-19)) Accuracy of 93.30 % for 3-classes Kumar et al [77] Chest X-Ray SARS-Net 3-Way (Normal, non-Covid-19 [e.g., viral, bacterial, etc. ], and Covid-19) Dataset Accuracy of 97.60 % for 3-classes Li et al [79] CT COVNet 3-Way (Non-pneumonia, community-acquired pneumonia, and Covid-19) Dataset Sensitivity of 90.00 % for 3-classes Chandra et al [95] Chest X-Ray Majority vote based classifier ensemble 3-Way (Normal, pneumonia, and Covid-19) Dataset Accuracy of 93.41 % for 3-classes Gayathri et al [96] Chest X-Ray Pre-trained model (InceptionResnetV2 + Xception) 2-Way (non-Covid-19, Covid-19) Dataset Accuracy of 95.78 % for 2-classes Loey et al [97] Chest X-Ray CNN Model 3-Way (Normal, pneumonia, and Covid-19) Dataset Accuracy of 96.00 % for 2-classes Li et al [98] CT The modified CheXNet 2-way (non-Covid-19, Covid-19) Dataset Accuracy of 87.00 % for 2-classes Proposed Method I CT + Chest X-Ray COVID-DSNet 4-Way (normal, bacterial pneumonia, viral pneumonia, Covid-19) Chest X-Ray Dataset Accuracy of 88.34 % for 4-classes 3-Way (normal, viral pneumonia, Covid-19) Chest X-Ray Dataset Accuracy of 92.83 % for 3-classes 2-Way (bacterial pneumonia, Covid-19) Chest X-Ray Dataset Accuracy of 99.45 % for 2-classes 3-Way (non-Covid-19, common pneumonia, Covid-19) CT Dataset Accuracy 97.60 % for 3-classes 2-Way (non-Covid-19, Covid-19) CT Dataset Accuracy 100 % for 2-classes Proposed Method II CT + Chest X-Ray COVID-DSNet + LSTM 4-Way (normal, bacterial pneumonia, viral pneumonia, Covid-19) Chest X-Ray Dataset Accuracy of 80.40 % for 4-classes 3-Way (non-Covid-19, common pneumonia, Covid-19) CT Dataset Accuracy of 96.33 % for 3-classes Proposed Method III …”

BO-ALLCNN: Bayesian-Based Optimized CNN for Acute Lymphoblastic Leukemia Detection in Microscopic Blood Smear Images

“…The selected sample is then evaluated using the objective function, and the cycle is continued until the objective function reaches its minimum or the least objective is identified within the given run time. In this work, the objective function is evaluated for a maximum of 30 times as commonly recommended [ 60 ]; each evaluation is executed in single optimization iteration. A stopping criterion is applied if the observed objective passes a threshold value.…”