A novel adaptive class weight adjustment-based multiclass segmentation error minimization technique for COVID-19 X-ray image analysis

Coronavirus outbreaks during the last couple of years created a huge health disaster for human lives. Diagnosis of COVID-19 infections is, thus, very important for the medical practitioners. For a quick detection, analysis of the COVID-19 chest X-ray images is inevitable. Therefore, there is a strong need for the development of a multiclass segmentation method for the purpose. Earlier techniques used for multiclass segmentation of images are mostly based on entropy measurements. Nonetheless, entropy methods are not efficient when the gray-level distribution of the image is nonuniform. To address this problem, a novel adaptive class weight adjustment-based multiclass segmentation error minimization technique for COVID-19 chest X-ray image analysis is investigated. Theoretical investigations on the first-hand objective functions are presented. The results on both the biclass and multiclass segmentation of medical images are enlightened. The key to our success is the adjustment of the pixel counts of different classes adaptively to reduce the error of segmentation. The COVID-19 chest X-ray images are taken from the Kaggle Radiography database for the experiments. The proposed method is compared with the state-of-the-art methods based on Tsallis, Kapur's, Masi, and Renyi entropy. The well-known segmentation metrics are used for an empirical analysis. Our method achieved a performance increase of around 8.03% in the case of PSNR values, 3.01% for FSIM, and 4.16% for SSIM. The proposed technique would be useful for extracting dots from micro-array images of DNA sequences and multiclass segmentation of the biomedical images such as MRI, CT, and PET.

An Advanced Deep Learning Medical Image Recognition and Diagnosis of Respiratory System Viruses

Respiratory infections are a confusing and time-consuming task that caused recently a pandemic that affected the whole world. One of the pandemics was COVID-19 that has exposed the vulnerability of medical services across the world, particularly in underdeveloped nations. There comes a strong demand for developing new computer-assisted diagnosis tools to present cost-effective and rapid screening in locations wherein enormous traditional testing is impossible. Medical imaging becomes critical for diagnosing disease, X-rays and computed tomography (CT) scan are employed in the deep network which will be helpful in diagnosing diseases. This paper proposes a scanning model based on using a Mel Frequency Cepstral Coefficients (MFCC) features extracted from a respiratory virus CT-Scan image and then filtered by applying Gabor filter (GF). The filtered image is passed to Convolutional Neural Network (CNN) for classifying the image for the presence of a respiratory virus such as Covid, Viral Pneumonia or being a healthy normal image. The proposed system achieved a validation accuracy of 100% with an overall accuracy of 99.44%. © 2022 IEEE.

Novel Square Error Minimization-Based Multilevel Thresholding Method for COVID-19 X-Ray Image Analysis Using Fast Cuckoo Search

Coronavirus outbreaks in 2019 (COVID-19) have been a huge disaster in the fields of health, economics, education, and tourism in the last two years. For diagnosis, a quick interpretation of the COVID-19 chest X-ray image is required. There is also a strong need to find an efficient multiclass segmentation technique for the analysis of COVID-19 X-ray images. Most of the threshold selection techniques are entropy-based. Nevertheless, these techniques suffer from their dependencies on the spatial distribution of grey values. To tackle these issues, a novel non-entropic threshold selection method is proposed, which is the primary key contribution having found a new source of information to the biomedical image processing field. The firsthand Square Error (SE)-based objective function is suggested. The second key contribution is the new optimizer called Fast Cuckoo Search (FCS), which is useful and brings novel ideas into the subject, used to optimize the suggested objective functions for computing the optimal thresholds. To ensure a faster convergence with a quality optimal solution, we include extra exploitation together with a chance factor. The FCS is validated using the well-known classical and CEC 2014 benchmark test functions, which shows a significant improvement over its predecessors—Adaptive Cuckoo Search (ACS) and other state-of-the-art optimizers. Further, the SE minimization-based optimal multilevel thresholding method using the FCS, coined as SE-FCS, is proposed. To experiment, images are considered from the Kaggle Radiography database. We have compared its performances with Tsallis, Kapur’s, and Masi entropy-based techniques using well-known segmentation metrics and achieved a performance increase of 2.95%, 5.51% and 10.50%, respectively. The proposed method shows superiority using Friedman’s mean rank statistical test and ranked first. [ FROM AUTHOR] Copyright of International Journal of Image & Graphics is the property of World Scientific Publishing Company and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full . (Copyright applies to all s.)

Explainable Augmented Intelligence and Deep Transfer Learning for Pediatric Pulmonary Health Evaluation

Biomedical Instrumentation is one of the fastest health emerging innovative technologies with proven contribution towards interdisciplinary medicine, it helps physicians to diagnose complex medical problems and provide treatment to patients precisely and safely. With this technological trend, explainable artificial intelligence, biomedical image processing and augmented intelligence can provide a tool that can help pediatricians, pulmonology and otolaryngology physicians, epidemiologists and pediatric practitioners to interpretably and reliably diagnose chronic and acute respiratory disorders in children, adolescents and infants. Unfortunately, the reliability of digital image processing for pulmonary disease diagnosis often depends on availability of large chest X-ray image datasets. This work presents a reliable interpretable deep transfer learning approach for pediatric pulmonary health evaluation regardless of the scarcity and limited annotated pediatric chest X-ray Image dataset sizes. This approach leverages a combination of computer vision tools and techniques to reduce child morbidity and mortality through predictive and preventive medicine with reduced surveillance risks and affordability in low resource settings. With open datasets, the deep neural networks classified the generated augmented images into 4 classes namely;Normal, Covid-19, Tuberculosis and Pneumonia at an accuracy of 97%, 97%, 70%, and 73% respectively with recall of 100% for Pneumonia and overall accuracy of 79% at only 10 epochs for both regular and transferred learning. © 2022 IEEE.

Analysis of GLRLM Texture Features Derived From Computed Tomography Scans For COVID-19 Diagnosis

Since its discovery in late 2019, COVID-19 has become a major worldwide concern due to its incredibly high degree of contagion, and early diagnosis is crucial to limit this global progression. Computed Tomography (CT) scans of the chest offer a low-cost alternative diagnosis modality to the standard reverse polymerase chain reaction (RT-PCR) test for COVID-19. In this paper, we analyze texture features extracted from chest CT scans using Gray Level Run Length Matrix (GLRLM) techniques for their ability to distinguish between COVID-19 and non-COVID-19 patients. Quantitative texture analysis of CT scans provides a measure of the biological heterogeneity in tissue microenvironment which can be useful in the diagnosis of a wide range of diseases, and we hypothesize that GLRLM texture features may hold significance for diagnosis of COVID-19. 13 GLRLM features were extracted from CT scans of 349 positive COVID-19 cases and 397 negative COVID-19 cases. Holdout validation was used to randomly split 70% of the images for training, and the remaining 30% for testing. A GentleBoost classifier was used to evaluate performance. A significant AUROC of 0.92 along with a high classification accuracy of 85.7% was obtained on the independent test set, indicating that GLRLM texture features extracted from chest CT scans have the potential to be a significant tool in the rapid and accurate diagnosis of COVID-19. © 2021 IEEE.

Cascaded deep learning classifiers for computer-aided diagnosis of COVID-19 and pneumonia diseases in X-ray scans.

Computer-aided diagnosis (CAD) systems are considered a powerful tool for physicians to support identification of the novel Coronavirus Disease 2019 (COVID-19) using medical imaging modalities. Therefore, this article proposes a new framework of cascaded deep learning classifiers to enhance the performance of these CAD systems for highly suspected COVID-19 and pneumonia diseases in X-ray images. Our proposed deep learning framework constitutes two major advancements as follows. First, complicated multi-label classification of X-ray images have been simplified using a series of binary classifiers for each tested case of the health status. That mimics the clinical situation to diagnose potential diseases for a patient. Second, the cascaded architecture of COVID-19 and pneumonia classifiers is flexible to use different fine-tuned deep learning models simultaneously, achieving the best performance of confirming infected cases. This study includes eleven pre-trained convolutional neural network models, such as Visual Geometry Group Network (VGG) and Residual Neural Network (ResNet). They have been successfully tested and evaluated on public X-ray image dataset for normal and three diseased cases. The results of proposed cascaded classifiers showed that VGG16, ResNet50V2, and Dense Neural Network (DenseNet169) models achieved the best detection accuracy of COVID-19, viral (Non-COVID-19) pneumonia, and bacterial pneumonia images, respectively. Furthermore, the performance of our cascaded deep learning classifiers is superior to other multi-label classification methods of COVID-19 and pneumonia diseases in previous studies. Therefore, the proposed deep learning framework presents a good option to be applied in the clinical routine to assist the diagnostic procedures of COVID-19 infection.