COVID-19 Lung Research Articles

COVID-19 lung infection segmentation from CT imaging using statistics and edge-region-based active contour

COVID-19 lung infection segmentation from CT imaging using statistics and edge-region-based active contour

Spatiotemporal cellular dynamics of germinal center reaction in COVID-19 lung draining lymph node based on imaging-based spatial transcriptomics

Although lymph node structures may be compromised in severe SARS-CoV-2 infection, the extent and parameters of recovery in convalescing patients remain unclear. Therefore, this study aimed to elucidate the nuances of lymphoid structural recovery and their implications for immunological memory in non-human primates infected with SARS-CoV-2. To do so, we utilized imaging-based spatial transcriptomics to delineate immune cell composition and tissue architecture formation in the lung draining lymph nodes during primary infection, convalescence, and reinfection from COVID-19. We noted the establishment of a germinal center with memory B cell differentiation within lymphoid follicles during convalescence accompanied by contrasting transcriptome patterns indicative of the acquisition of follicular helper T cells versus the loss of regulatory T cells. Additionally, repopulation of germinal center-like B cells was observed in the medullary niche with accumulating plasma cells along with enhanced transcriptional expression of B cell activating factor receptor over the course of reinfection. The spatial transcriptome atlas produced herein enhances our understanding of germinal center formation with immune cell dynamics during COVID-19 convalescence and lymphoid structural recovery with transcriptome dynamics following reinfection. These findings have the potential to inform the optimization of vaccine strategies and the development of precise therapeutic interventions in the spatial context.

Advancing Medical Image Analysis: The Role of Adaptive Optimization Techniques in Enhancing COVID-19 Detection, Lung Infection, and Tumor Segmentation

Artificial intelligence (AI) holds significant potential to revolutionize healthcare by improving clinical practices and patient outcomes. This research explores the integration of AI in healthcare, focusing on methodologies such as machine learning, natural language processing, and computer vision, which enable the extraction of valuable insights from complex medical imaging and clinical data. Through a comprehensive literature review, the study highlights AI’s practical applications in diagnostics, treatment planning, and predicting patient outcomes. Additionally, ethical issues, data privacy, and legal frameworks are examined, emphasizing the importance of responsible AI usage in healthcare. The findings demonstrate AI’s ability to enhance diagnostic accuracy, streamline administrative tasks, and optimize resource allocation, leading to personalized treatments and more efficient healthcare management. However, challenges remain, including data quality, algorithm transparency, and ethical concerns, which must be addressed to ensure safe and effective AI deployment. Continued research, collaboration between healthcare professionals and AI experts, and the development of robust regulatory frameworks are essential for maximizing AI’s benefits while minimizing risks. This research underscores the transformative potential of AI in healthcare and stresses the need for a multidisciplinary approach to address the ethical and regulatory complexities involved in its widespread adoption

LungXpertAI: A deep multi-task learning model for chest CT scan analysis and COVID-19 detection

Addressing the urgent need for accurate COVID-19 identification and lung infection segmentation in CT scans, our study introduces LungXpertAI, a novel Multi-Task Learning architecture. Previous studies on multi-task structures have been limited in providing detailed insights into their architectures. Addressing these details is crucial for enhancing the performance of multi-task structures. Our proposed multi-task structure, based on the U-Net architecture, incorporates a shared encoder, a specialized segmentation decoder, and a multi-layer perceptron for classification. In the proposed approach, image processing algorithms, including histogram equalization, median filtering, and morphological operations have been employed in the pre-processing stage of input images. Also, by combining these algorithms in pairs, an attempt has been made to enhance the performance of tasks. Initial preprocessing and the integration of the Convolutional Block Attention Module (CBAM) contribute to effective feature extraction. Post-processing with Conditional Random Field (CRF) further refines segmentation outputs. CRF considers spatial dependencies within the segmentation results, contributing to more precise delineation of COVID-19-affected areas in CT scans. Evaluation on using dataset demonstrates the effectiveness of pairwise combinations of image processing algorithms, achieving superior results with a segmentation accuracy of 95.66 % and a classification accuracy of 95.62 %. The incorporation of the CBAM module improves classification accuracy to 96.22 %. Moreover, Integrating CBAM and CRF significantly improves segmentation accuracy to 96.51 %. The application of our proposed method steps to U-Net++ and ResUnet showcases their potential for enhancing multi-task structures. This study establishes new COVID-19 detection standards, promising progress in medical image analysis.

P.515: ICU-acquired muscle weakness in covid-19 lung transplant cases: The imperative for enhanced surveillance and prophylaxis.

P.515: ICU-acquired muscle weakness in covid-19 lung transplant cases: The imperative for enhanced surveillance and prophylaxis.

Enhancing Medical Image Quality Using Fractional Order Denoising Integrated with Transfer Learning

In medical imaging, noise can significantly obscure critical details, complicating diagnosis and treatment. Traditional denoising techniques often struggle to maintain a balance between noise reduction and detail preservation. To address this challenge, we propose an “Efficient Transfer-Learning-Based Fractional Order Image Denoising Approach in Medical Image Analysis (ETLFOD)” method. Our approach uniquely integrates transfer learning with fractional order techniques, leveraging pre-trained models such as DenseNet121 to adapt to the specific needs of medical image denoising. This method enhances denoising performance while preserving essential image details. The ETLFOD model has demonstrated superior performance compared to state-of-the-art (SOTA) techniques. For instance, our DenseNet121 model achieved an accuracy of 98.01%, precision of 98%, and recall of 98%, significantly outperforming traditional denoising methods. Specific results include a 95% accuracy, 98% precision, 99% recall, and 96% F1-score for MRI brain datasets, and an 88% accuracy, 91% precision, 95% recall, and 88% F1-score for COVID-19 lung data. X-ray pneumonia results in the lung CT dataset showed a 92% accuracy, 97% precision, 98% recall, and 93% F1-score. It is important to note that while we report performance metrics in this paper, the primary evaluation of our approach is based on the comparison of original noisy images with the denoised outputs, ensuring a focus on image quality enhancement rather than classification performance.

The Role of Artificial Intelligence in Supporting Radiologists in Detecting Lung Lesions Caused by COVID-19 - A Scoping Review

The COVID-19 pandemic has posed significant challenges to healthcare systems, particularly in the realm of medical imaging and the diagnosis of COVID-19 pneumonia lung lesions. Artificial intelligence (AI) has become essential in assisting radiologists by swiftly analyzing extensive volumes of computed tomography (CT) scan data to identify lung abnormalities. Radiologists, who typically conduct thorough examinations of CT scans, benefit from AI's capability to pre-screen images, flag potential issues, and prioritize urgent cases, thereby enhancing efficiency during times of high demand. AI, especially through deep learning, can recognize subtle patterns in lung images that may be overlooked by human eyes, offering valuable second opinions and improving diagnostic accuracy and consistency. The early detection of COVID-19 lung lesions by AI facilitates prompt treatment, helping to prevent disease progression and improve patient outcomes. This review article aims to consolidate current knowledge on the topic by examining existing literature and discussing the advantages and disadvantages of employing CT imaging and AI tools for diagnosing COVID-19.

FMD-UNet: fine-grained feature squeeze and multiscale cascade dilated semantic aggregation dual-decoder UNet for COVID-19 lung infection segmentation from CT images

With the advancement of computer-aided diagnosis, the automatic segmentation of COVID-19 infection areas holds great promise for assisting in the timely diagnosis and recovery of patients in clinical practice. Currently, methods relying on U-Net face challenges in effectively utilizing fine-grained semantic information from input images and bridging the semantic gap between the encoder and decoder. To address these issues, we propose an FMD-UNet dual-decoder U-Net network for COVID-19 infection segmentation, which integrates a Fine-grained Feature Squeezing (FGFS) decoder and a Multi-scale Dilated Semantic Aggregation (MDSA) decoder. The FGFS decoder produces fine feature maps through the compression of fine-grained features and a weighted attention mechanism, guiding the model to capture detailed semantic information. The MDSA decoder consists of three hierarchical MDSA modules designed for different stages of input information. These modules progressively fuse different scales of dilated convolutions to process the shallow and deep semantic information from the encoder, and use the extracted feature information to bridge the semantic gaps at various stages, this design captures extensive contextual information while decoding and predicting segmentation, thereby suppressing the increase in model parameters. To better validate the robustness and generalizability of the FMD-UNet, we conducted comprehensive performance evaluations and ablation experiments on three public datasets, and achieved leading Dice Similarity Coefficient (DSC) scores of 84.76, 78.56 and 61.99% in COVID-19 infection segmentation, respectively. Compared to previous methods, the FMD-UNet has fewer parameters and shorter inference time, which also demonstrates its competitiveness.

A Histological Analysis and Detection of Complement Regulatory Protein CD55 in SARS-CoV-2 Infected Lungs.

A complement imbalance in lung alveolar tissue can play a deteriorating role in COVID-19, leading to acute respiratory distress syndrome (ARDS). CD55 is a transmembrane glycoprotein that inhibits the activation of the complement system at the intermediate cascade level, blocking the activity of the C3 convertase. In our study, lung specimens from COVID-19 and ARDS-positive COVID+/ARDS+ patients were compared with COVID-19 and ARDS-negative COVID-/ARDS- as well as COVID-/ARDS+ patients. Histochemical staining and immunolabeling of CD55 protein were performed. The COVID-/ARDS- specimen showed higher expression and homogeneous distribution of glycosaminoglycans as well as compactly arranged elastic and collagen fibers of the alveolar walls in comparison to ARDS-affected lungs. In addition, COVID-/ARDS- lung tissues revealed stronger and homogenously distributed CD55 expression on the alveolar walls in comparison to the disrupted COVID-/ARDS+ lung tissues. Even though the collapse of the alveolar linings and the accumulation of cellular components in the alveolar spaces were characteristic of COVID+/ARDS+ lung tissues, evaluating CD55 expression could be relevant to understand its relation to the disease. Furthermore, targeting CD55 upregulation as a potential therapy could be an option for post-infectious complications of COVID-19 and other inflammatory lung diseases in the future.

Surfactant derived protein type B: a new biomarker linked to respiratory failure and lung damage in mild to moderate SARS-CoV-2 pneumonia

BackgroundThe COVID-19 pandemic has led to significant concern due to its impact on human health, particularly through pneumonia-induced lung damage. Surfactant proteins A and D (SP-A and SP-D) are implicated in COVID-19 lung damage, but the role of surfactant protein B (SP-B) remains unclear.MethodsWe conducted a single-center prospective observational study involving 73 hospitalised COVID-19 pneumonia patients. SP-B levels were measured within 48 h of admission, alongside SP-A and SP-D in a subset. Clinical data were collected, and follow-up visits were conducted after 6 months.ResultsAt hospitalisation, circulating immature SP-B levels measured in 73 patients (26.31 AU [14.27–41.31]), correlated significantly with lung involvement (r=0.447, p<0.001) and oxygen support requirement (p=0.005). SP-B levels did not predict mechanical ventilation or ICU admission. SP-B decreased significantly (p<0.001) from 25.53 AU [14.36–41.46] at the acute hospitalisation to 12.73 AU [9.12–20.23] at 6 months follow up, unlike SP-A and SP-D that did not change significantly. Immature SP-B (but not SP-A and SP-D) confirmed to be significantly associated with the need of oxygen support (n=26, 58%) during the hospitalisation (p<0.05).ConclusionImmature SP-B emerges as a potential biomarker for COVID-19 pneumonia severity and prognosis. Its dynamic changes suggest utility in monitoring disease progression and long-term outcomes, despite limitations in predicting hard endpoints. Larger studies are needed to validate these findings and understand the underlying mechanisms of surfactant protein dysregulation in COVID-19 pathogenesis.

Mathematical Modeling of Impacts of Patient Differences on Renin-Angiotensin System and Applications to COVID-19 Lung Fibrosis Outcomes.

Patient-specific premorbidity, age, and sex are significant heterogeneous factors that influence the severe manifestation of lung diseases, including COVID-19 fibrosis. The renin-angiotensin system (RAS) plays a prominent role in regulating the effects of these factors. Recent evidence suggests patient-specific alteration of RAS homeostasis concentrations with premorbidity and the expression level of angiotensin converting enzyme 2 (ACE2) during COVID-19. However, conflicting evidence suggests decreases, increases, or no changes in RAS after SARS-CoV-2 infection. In addition, detailed mechanisms connecting the patient-specific conditions before infection to infection-induced RAS alteration are still unknown. Here, a multiscale model is developed to quantify the systemic contribution of heterogeneous factors of RAS during COVID-19. Three submodels are connected-an ABM COVID-19 in-host lung tissue model, a RAS model, and a fibrosis model to investigate the effects of patient-group-specific factors in the systemic alteration of RAS and collagen deposition in the lung. The model results indicate cell death due to inflammatory response as a major contributor to the reduction of ACE and ACE2. In contrast, there are no significant changes in ACE2 dynamics due to viral-bound internalization of ACE2. Reduction of ACE and ACE2 downregulates the homeostasis concentration of RAS, including angiotensin II (ANGII), in the lung tissue. At the same time, the decrease in ACE2 increases systemic ANGII and results in severe lung injury and fibrosis. The model explains possible mechanisms for conflicting evidence of RAS alterations in previously published studies, and simulated results are consistent with reported RAS peptide values for SARS-CoV-2-negative and SARS-CoV-2-positive patients. We observed decreased RAS peptides for all virtual patient groups with aging in both sexes. In contrast, large variations in the magnitude of reduction were observed between male and female virtual patients in the older and middle-aged groups. We also predicted that feedback of ANGII·AT1R to renin could restore ANGI homeostasis concentration but fails to restore homeostasis values of RAS peptides downstream of ANGI. In addition, the results show that ACE2 variations with age and sex significantly alter the concentrations of RAS peptides and lead to collagen deposition with slight variations depending on age and sex. This model may find further applications in patient-specific calibrations of tissue models for acute and chronic lung diseases to develop personalized treatments.

Novel Deep CNNs Explore Regions, Boundaries, and Residual Learning for COVID-19 Infection Analysis in Lung CT.

COVID-19 poses a global health crisis, necessitating precise diagnostic methods for timely containment. However, accurately delineating COVID-19-affected regions in lung CT scans is challenging due to contrast variations and significant texture diversity. In this regard, this study introduces a novel two-stage classification and segmentation CNN approach for COVID-19 lung radiological pattern analysis. A novel Residual-BRNet is developed to integrate boundary and regional operations with residual learning, capturing key COVID-19 radiological homogeneous regions, texture variations, and structural contrast patterns in the classification stage. Subsequently, infectious CT images undergo lesion segmentation using the newly proposed RESeg segmentation CNN in the second stage. The RESeg leverages both average and max-pooling implementations to simultaneously learn region homogeneity and boundary-related patterns. Furthermore, novel pixel attention (PA) blocks are integrated into RESeg to effectively address mildly COVID-19-infected regions. The evaluation of the proposed Residual-BRNet CNN in the classification stage demonstrates promising performance metrics, achieving an accuracy of 97.97%, F1-score of 98.01%, sensitivity of 98.42%, and MCC of 96.81%. Meanwhile, PA-RESeg in the segmentation phase achieves an optimal segmentation performance with an IoU score of 98.43% and a dice similarity score of 95.96% of the lesion region. The framework's effectiveness in detecting and segmenting COVID-19 lesions highlights its potential for clinical applications.

COVID-19 pandemic and lung cancer rates in Kazakhstan: a comparative study of pre-pandemic and pandemic periods

Aim of the study. To assess the impact of the COVID-19 pandemic on the incidence of lung cancer by component analysis.Materials and methods. The research material was the data of the Ministry of Health of the Republic of Kazakhstan concerning LC (Form 7). To analyze the dynamics of incidence based on the number of cases from 2011 to 2020, a component method was used.Results and discussion. Analyzing 2011 and 2020, we found a tendency to decrease the incidence of lung cancer (p=0.000). However, comparing 2019 and 2020, we found a sharp decline in the incidence. The overall decrease was −2,220/0000 and depended on changes in the age structure of the population (∑ΔA=+0,340/0000), the risk of getting illness (∑ΔR=−2,520/0000) and the combined effect of the age structure and the risk of getting illness (∑ΔAR=−0,040/0000). The decrease was mainly due to the impact of the risk of getting illness. According to the calculations of the component analysis, 3,856 patients were expected in 2020, but instead the number of patients decreased to 3,375 and this is characterized by a decrease in the number of cases as a result of changes in the risk of getting illness.Conclusions. The COVID-19 pandemic has significantly reduced the incidence rate of lung cancer in Kazakhstan. Additional analysis and research are needed to understand the full impact and take appropriate measures to combat cancer in a pandemic.

Communication between alveolar macrophages and fibroblasts via the TNFSF12-TNFRSF12A pathway promotes pulmonary fibrosis in severe COVID-19 patients

BackgroundSevere COVID-19 infection has been associated with the development of pulmonary fibrosis, a condition that significantly affects patient prognosis. Understanding the underlying cellular communication mechanisms contributing to this fibrotic process is crucial.ObjectiveIn this study, we aimed to investigate the role of the TNFSF12-TNFRSF12A pathway in mediating communication between alveolar macrophages and fibroblasts, and its implications for the development of pulmonary fibrosis in severe COVID-19 patients.MethodsWe conducted single-cell RNA sequencing (scRNA-seq) analysis using lung tissue samples from severe COVID-19 patients and healthy controls. The data was processed, analyzed, and cell types were annotated. We focused on the communication between alveolar macrophages and fibroblasts and identified key signaling pathways. In vitro experiments were performed to validate our findings, including the impact of TNFRSF12A silencing on fibrosis reversal.ResultsOur analysis revealed that in severe COVID-19 patients, alveolar macrophages communicate with fibroblasts primarily through the TNFSF12-TNFRSF12A pathway. This communication pathway promotes fibroblast proliferation and expression of fibrotic factors. Importantly, silencing TNFRSF12A effectively reversed the pro-proliferative and pro-fibrotic effects of alveolar macrophages.ConclusionThe TNFSF12-TNFRSF12A pathway plays a central role in alveolar macrophage-fibroblast communication and contributes to pulmonary fibrosis in severe COVID-19 patients. Silencing TNFRSF12A represents a potential therapeutic strategy for mitigating fibrosis in severe COVID-19 lung disease.

An in-depth investigation of serum Krebs von den Lungen-6 and other biomarkers in COVID-19 severity and mortality.

Krebs von den Lungen-6 (KL-6) is a glycoprotein mainly expressed by type II pneumocytes and recently known as a lung injury biomarker. However, the number of studies is still limited, especially in Indonesian COVID-19 populations. Therefore, we aim to provide correlation, sensitivity, and specificity analyses of KL-6 and other biomarkers in Indonesian COVID-19 severity and mortality. We conducted a cross-sectional study involving adult COVID-19 patients at Universitas Airlangga Hospital, Surabaya, East Java, Indonesia, between March 26, 2021, and August 25, 2021. KL-6 and other biomarker levels were compared according to severity (severe versus non-severe) and mortality (non-survivor versus survivor). We also included the receiver operating characteristic analysis to define the optimal cut-off, sensitivity, and specificity of KL-6 to determine COVID-19 severity and mortality. We enrolled 78 COVID- 19 patients (23 non-survivors), including 39 non-severe and 39 severe patients. There was no significant difference in serum KL-6 levels, neither in severity nor mortality groups. KL-6 had the strongest positive correlations with ferritin in severe patients (r=0.313) and non-survivors (r=0.467). We observed that the best sensitivity was KL-6 combined with platelet-to- lymphocyte ratio (PLR) (0.818) in severe patients and with neutrophil-to-lymphocyte ratio (NLR)/PLR/ferritin/C-reactive protein (0.867) in non-survivors. In contrast, the best specificity was found when KL-6 was combined with NLR/D-dimer (0.750) in severe patients and with D-dimer (0.889) in non-survivors. Serum KL-6 is a useful auxiliary laboratory evaluation index for COVID-19 lung injury to depict its severity and mortality.

Deciphering Alveolo-Capillary Gas Transfer Disturbances in Patients Recovering from COVID-19 Lung Disease.

Impaired lung gas exchange is commonly seen in patients with pulmonary involvement related to SARS-CoV-2 acute infection or post-acute COVID-19 syndrome (PACS). The primary aim of our study was to assess lung gas transfer, measuring the pulmonary diffusion capacity for nitric oxide (DLNO) and carbon monoxide (DLCO) in all COVID-19 patients. Our secondary aim was to decipher the respective roles of perturbed lung membrane conductance (DM) and reduced pulmonary capillary volume (VC) in patients with impaired lung gas exchange. From May to October 2020, we measured DLNO-DLCO in 118 patients during their post-COVID-19 period (4.6 months after infection) to decipher alveolo-capillary gas transfer disturbances. DLNO-DLCO measurement was also performed in 28 healthy non-smokers as controls. Patients were classified into three groups according to the severity (mild, moderate, and severe) of acute COVID-19 infection. Patients with mild COVID-19 had normal lung volumes and airways expiratory flows but impaired pulmonary gas exchange, as shown by the significant decreases in DLNO, DLCO, DM, and VC as compared with controls. VC was significantly impaired and the DLNO/DLCO ratio was increased in patients with moderate (n = 4, 11%) and severe COVID-19 (n = 23, 49%). Abnormal membrane conductance was also seen in all three groups of post-COVID-19 patients. These findings suggest a persistent alveolo-capillary gas transfer defect, implying not only reduced membrane conductance but also abnormal pulmonary vascular capacitance in all PACS patients, even those with a milder form of COVID-19 infection.

IL-13 decreases susceptibility to airway epithelial SARS-CoV-2 infection but increases disease severity in vivo.

Treatments available to prevent progression of virus-induced lung diseases, including coronavirus disease 2019 (COVID-19) are of limited benefit once respiratory failure occurs. The efficacy of approved and emerging cytokine signaling-modulating antibodies is variable and is affected by disease course and patient-specific inflammation patterns. Therefore, understanding the role of inflammation on the viral infectious cycle is critical for effective use of cytokine-modulating agents. We investigated the role of the type 2 cytokine IL-13 on SARS-CoV-2 binding/entry, replication, and host response in primary HAE cells in vitro and in a model of mouse-adapted SARS-CoV-2 infection in vivo. IL-13 protected airway epithelial cells from SARS-CoV-2 infection in vitro by decreasing the abundance of ACE2-expressing ciliated cells rather than by neutralization in the airway surface liquid or by interferon-mediated antiviral effects. In contrast, IL-13 worsened disease severity in mice; the effects were mediated by eicosanoid signaling and were abolished in mice deficient in the phospholipase A2 enzyme PLA2G2D. We conclude that IL-13-induced inflammation differentially affects multiple steps of COVID-19 pathogenesis. IL-13-induced inflammation may be protective against initial SARS-CoV-2 airway epithelial infection; however, it enhances disease progression in vivo. Blockade of IL-13 and/or eicosanoid signaling may be protective against progression to severe respiratory virus-induced lung disease.

Lung Transplantation in the United States for COVID-19 Related Lung Disease During the Pandemic.

Lung transplantation (LTx) is a potential intervention for end-stage COVID-19 lung disease. Current literature is sparse regarding the outcomes of LTx for COVID-19 related acute respiratory distress syndrome (ARDS) and pulmonary fibrosis (PF). This study aims to characterize outcomes and patterns of LTx for COVID-19 related lung disease throughout the pandemic. Patients who underwent LTx during the pandemic for COVID-19 related lung disease were retrospectively identified using the UNOS registry. Demographics, as well as outcomes measures and nationwide patterns of care were collected and analyzed. A total of 510 adult cases of LTx for COVID-19 (259 ARDS, 251 PF) were compared to 4,031 without COVID-19 (3,994 PF, 37 ARDS). Patients who received LTx for COVID-19 ARDS did not differ in 2-year survival when compared to those with COVID-19 PF (81.9% vs 77.2%, p = 0.4428). Compared to non-COVID-19 etiologies, COVID-19 ARDS patients had higher rates of stroke (2.3% vs 0%, p = 0.0005), lower rates of graft failure (12.8% vs 36.1%, p = 0.0003) and post-transplant ECMO (29.6% vs 41.7%, p = 0.0002), and improved 2-year survival following LTx (81.9% vs 61.7%, p = 0.0064). No difference in 2-year survival following LTx was observed between patients with COVID-19 and non-COVID-19 PF (77.2% vs 71.8%, p = 0.34). Rates of LTx spiked with variant emergence and declined with rounds of vaccination. Our results are consistent with early reports of survival outcomes following LTx for COVID-19 ARDS and PF while providing an increased layer of granularity. LTx may be considered as a safe and effective intervention for COVID-19 lung disease.

Augmenting Radiological Diagnostics with AI for Tuberculosis and COVID-19 Disease Detection: Deep Learning Detection of Chest Radiographs.

In this research, we introduce a network that can identify pneumonia, COVID-19, and tuberculosis using X-ray images of patients' chests. The study emphasizes tuberculosis, COVID-19, and healthy lung conditions, discussing how advanced neural networks, like VGG16 and ResNet50, can improve the detection of lung issues from images. To prepare the images for the model's input requirements, we enhanced them through data augmentation techniques for training purposes. We evaluated the model's performance by analyzing the precision, recall, and F1 scores across training, validation, and testing datasets. The results show that the ResNet50 model outperformed VGG16 with accuracy and resilience. It displayed superior ROC AUC values in both validation and test scenarios. Particularly impressive were ResNet50's precision and recall rates, nearing 0.99 for all conditions in the test set. On the hand, VGG16 also performed well during testing-detecting tuberculosis with a precision of 0.99 and a recall of 0.93. Our study highlights the performance of our deep learning method by showcasing the effectiveness of ResNet50 over traditional approaches like VGG16. This progress utilizes methods to enhance classification accuracy by augmenting data and balancing them. This positions our approach as an advancement in using state-of-the-art deep learning applications in imaging. By enhancing the accuracy and reliability of diagnosing ailments such as COVID-19 and tuberculosis, our models have the potential to transform care and treatment strategies, highlighting their role in clinical diagnostics.

The use of Efficientnet_b0 network to identify COVID-19 in chest X-ray images

A newly discovered coronavirus called COVID-19 poses the greatest threat to mankind in the twenty-first century. Mortality has dramatically increased in all cities and countries due to the virus's current rate of spread. A speedy and precise diagnosis is also necessary in order to treat the illness. This study identified three groups for chest X-ray images: Covid, normal, and pneumonia. This study's objective is to present a framework for categorizing chest X-ray images into three groups of pneumonia, normal, and Covid scenarios. To do this, chest X-ray images from the Kaggle database which have been utilized in previous studies were obtained. It is suggested to use an Efficientnet_b0 model to identify characteristics in raw data hierarchically. An unedited X-ray image of the chest is enhanced for more reasonable assumptions in order to apply the proposed method in real-world situations. With an overall accuracy of 93.75%, the proposed network correctly identified the chest X-ray images to the classes of Covid, viral pneumonia, and normal on the test set. 90% accuracy rate for the test dataset was attained for the viral pneumonitis group. On the test dataset, the Normal class accuracy was 94.7%, while the Covid class accuracy was 96%. The findings indicate that the network is robust. In addition, when compared to the most advanced techniques of identifying pneumonia, the concluded findings from the suggested model are highly encouraging. Since the recommended network is successful at doing so utilizing chest X-ray imaging, radiologists can diagnose COVID-19 and other lung infectious infections promptly and correctly.