Mixup-Inf-Net: A data augmentation algorithm for segmentation of new coronary pneumonia infections

In 2020, the global spread of Coronavirus Disease 2019 exposed entire world to a severe health crisis. This has limited fast and accurate screening of suspected cases due to equipment shortages and and harsh testing environments. The current diagnosis of suspected cases has benefited greatly from the use of radiographic brain imaging, also including X-ray and scintigraphy, as a crucial addition to screening tests for new coronary pneumonia disease. However, it is impractical to gather enormous volumes of data quickly, which makes it difficult for depth models to be trained. To solve these problems, we obtained a new dataset by data augmentation Mixup method for the used chest CT slices. It uses lung infection segmentation (Inf-Net [1]) in a deep network and adds a learning framework with semi-supervised to form a Mixup-Inf-Net semi-supervised learning framework model to identify COVID-19 infection area from chest CT slices. The system depends primarily on unlabeled data and merely a minimal amount of annotated data is required;therefore, the unlabeled data generated by Mixup provides good assistance. Our framework can be used to improve improve learning and performance. The SemiSeg dataset and the actual 3D CT images that we produced are used in a variety of tests, and the analysis shows that Mixup-Inf-Net semi-supervised outperforms most SOTA segmentation models learning framework model in this study, which also enhances segmentation performance. © 2023 SPIE.

Residual Feature Attentional Fusion Network for Lightweight Chest CT Image Super-Resolution

The diagnosis of COVID-19 requires chest computed tomography (CT). High-resolution CT images can provide more diagnostic information to help doctors better diagnose the disease, so it is of clinical importance to study super-resolution (SR) algorithms applied to CT images to improve the reso-lution of CT images. However, most of the existing SR algorithms are studied based on natural images, which are not suitable for medical images;and most of these algorithms improve the reconstruction quality by increasing the network depth, which is not suitable for machines with limited resources. To alleviate these issues, we propose a residual feature attentional fusion network for lightweight chest CT image super-resolution (RFAFN). Specifically, we design a contextual feature extraction block (CFEB) that can extract CT image features more efficiently and accurately than ordinary residual blocks. In addition, we propose a feature-weighted cascading strategy (FWCS) based on attentional feature fusion blocks (AFFB) to utilize the high-frequency detail information extracted by CFEB as much as possible via selectively fusing adjacent level feature information. Finally, we suggest a global hierarchical feature fusion strategy (GHFFS), which can utilize the hierarchical features more effectively than dense concatenation by progressively aggregating the feature information at various levels. Numerous experiments show that our method performs better than most of the state-of-the-art (SOTA) methods on the COVID-19 chest CT dataset. In detail, the peak signal-to-noise ratio (PSNR) is 0.11 dB and 0.47 dB higher on CTtest1 and CTtest2 at x3 SR compared to the suboptimal method, but the number of parameters and multi-adds are reduced by 22K and 0.43G, respectively. Our method can better recover chest CT image quality with fewer computational resources and effectively assist in COVID-19.

COVID-19 Detection by Using Handcrafted Features Extracted From Chest CT-Scan Images

Coronavirus illness, which was initially diagnosed in 2019 but has propagated rapidly across the globe, has led to increased fatalities. According to professional physicians who examined chest CT scans, COVID-19 behaves differently than various viral cases of pneumonia. Even though the illness only recently emerged, a number of research investigations have been performed wherein the progression of the disease impacts mostly on the lungs are identified using thoracic CT scans. In this work, automated identification of COVID-19 is used by using machine learning classifier trained on more than 1000+ lung CT Scan images. As a result, immediate diagnosis of COVID-19, which is very much necessary in the opinion of healthcare specialists, is feasible. To improve detection accuracy, the feature extraction method are applied on regions of interests. Feature extraction approaches, including Discrete Wavelet Transform (DWT), Grey Level Cooccurrence Matrix (GLCM), Grey Level Run Length Matrix (GLRLM), and Grey-Level Size Zone Matrix (GLSZM) algorithms are used. Then the classification by using Support Vector Machines (SVM) is used. The classification accuracy is assessed by using precision, specificity, accuracy, sensitivity and F-score measures. Among all feature extraction methods, the GLCM approach has given the optimum classification accuracy of 95.6%. . © 2023 IEEE.

Classification of COVID-19 cases from chest CT volumes using hybrid model of 3D CNN and 3D MLP-Mixer

This paper proposes an automated classification method of COVID-19 chest CT volumes using improved 3D MLP-Mixer. Novel coronavirus disease 2019 (COVID-19) spreads over the world, causing a large number of infected patients and deaths. Sudden increase in the number of COVID-19 patients causes a manpower shortage in medical institutions. Computer-aided diagnosis (CAD) system provides quick and quantitative diagnosis results. CAD system for COVID-19 enables efficient diagnosis workflow and contributes to reduce such manpower shortage. In image-based diagnosis of viral pneumonia cases including COVID-19, both local and global image features are important because viral pneumonia cause many ground glass opacities and consolidations in large areas in the lung. This paper proposes an automated classification method of chest CT volumes for COVID-19 diagnosis assistance. MLP-Mixer is a recent method of image classification using Vision Transformer-like architecture. It performs classification using both local and global image features. To classify 3D CT volumes, we developed a hybrid classification model that consists of both a 3D convolutional neural network (CNN) and a 3D version of the MLP-Mixer. Classification accuracy of the proposed method was evaluated using a dataset that contains 1205 CT volumes and obtained 79.5% of classification accuracy. The accuracy was higher than that of conventional 3D CNN models consists of 3D CNN layers and simple MLP layers. © 2023 SPIE.

A Systematic Literature Review of Deep Learning Algorithms for Segmentation of the COVID-19 Infection

Coronavirus has infected more than 753 million people, ranging in severity from one person to another, where more than six million infected people died worldwide. Computer-aided diagnostic (CAD) with artificial intelligence (AI) showed outstanding performance in effectively diagnosing this virus in real-time. Computed tomography is a complementary diagnostic tool to clarify the damage of COVID-19 in the lungs even before symptoms appear in patients. This paper conducts a systematic literature review of deep learning methods for classifying the segmentation of COVID-19 infection in the lungs. We used the methodology of systematic reviews and meta-analyses (PRISMA) flow method. This research aims to systematically analyze the supervised deep learning methods, open resource datasets, data augmentation methods, and loss functions used for various segment shapes of COVID-19 infection from computerized tomography (CT) chest images. We have selected 56 primary studies relevant to the topic of the paper. We have compared different aspects of the algorithms used to segment infected areas in the CT images. Limitations to deep learning in the segmentation of infected areas still need to be developed to predict smaller regions of infection at the beginning of their appearance.

Acute-Onset Achalasia Following a Recent COVID-19 Infection: A Case Report.

Achalasia is a rare esophageal motility disorder that leads to dysphagia, regurgitation, and several other symptoms. While the etiology of achalasia is not completely understood, studies have suggested an immune reaction to viral infections, including severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), as a potential cause. Here, we present a case report of a previously healthy 38-year-old male who presented to the emergency room with severe shortness of breath, recurrent vomiting, and dry cough, that had progressively worsened over five days. The patient was diagnosed with coronavirus disease 2019 (COVID-19), and a chest CT also revealed prominent features of achalasia with a markedly dilated esophagus and areas of narrowing at the distal esophagus. The initial management of the patient included IV fluids, antibiotics, anticholinergics, and corticosteroid inhalers which improved his symptoms. This case report highlights the importance of considering the acute-onset of achalasia in COVID-19 patients and the need for further research on the potential association between SARS-CoV-2 and achalasia.

Oximetry at admission as a predictor of tomographic and functional impairment after 3-6 months in hospitalized patients with COVID-19.

OBJECTIVE: To investigate characteristics that may be associated with radiologic and functional findings following discharge in patients with severe coronavirus disease 2019 (COVID-19). METHODS: This single-center, prospective, observational cohort study comprised patients aged >18 years who were hospitalized with COVID-19 pneumonia, between May and October 2020. After 3 to 6 months of discharge, patients were clinically evaluated and underwent spirometry, a 6-minute walk test (6MWT), and chest computed tomography (CT). Statistical analysis was performed using association and correlation tests. RESULTS: A total of 134 patients were included (25/114 [22%] were admitted with severe hypoxemia). On the follow-up chest CT, 29/92 (32%) had no abnormalities, regardless of the severity of the initial involvement, and the mean 6MWT distance was 447 m. Patients with desaturation on admission had an increased risk of remaining CT abnormalities: patients with SpO2 between 88 and 92% had a 4.0-fold risk, and those with SpO2 < 88% had a 6.2-fold risk. The group with SpO2 < 88% also walked shorter distances than patients with SpO2 between 88 and 92%. CONCLUSION: Initial hypoxemia was found to be a good predictor of persistent radiological abnormalities in follow-up and was associated with low performance in 6MWT.

The importance of chest CT severity score and lung CT patterns in risk assessment in COVID-19-associated pneumonia: a comparative study.

Introduction: Chest computed tomography (CT) is suitable to assess morphological changes in the lungs. Chest CT scoring systems (CCTS) have been developed and use in order to quantify the severity of pulmonary involvement in COVID-19. CCTS has also been correlated with clinical outcomes. Here we wished to use a validated, relatively simple CTSS to assess chest CT patterns and to correlate CTSS with clinical outcomes in COVID-19. Patients and methods: Altogether 227 COVID-19 cases underwent chest CT scanning using a 128 multi-detector CT scanner (SOMATOM Go Top, Siemens Healthineers, Germany). Specific pathological features, such as ground-glass opacity (GGO), crazy-paving pattern, consolidation, fibrosis, subpleural lines, pleural effusion, lymphadenopathy and pulmonary embolism were evaluated. CTSS developed by Pan et al. (CTSS-Pan) was applied. CTSS and specific pathologies were correlated with demographic, clinical and laboratory data, A-DROP scores, as well as outcome measures. We compared CTSS-Pan to two other CT scoring systems. Results: The mean CTSS-Pan in the 227 COVID-19 patients was 14.6 ± 6.7. The need for ICU admission (p < 0.001) and death (p < 0.001) were significantly associated with higher CTSS. With respect to chest CT patterns, crazy-paving pattern was significantly associated with ICU admission. Subpleural lines exerted significant inverse associations with ICU admission and ventilation. Lymphadenopathy was associated with all three outcome parameters. Pulmonary embolism led to ICU admission. In the ROC analysis, CTSS>18.5 significantly predicted admission to ICU (p = 0.026) and CTSS>19.5 was the cutoff for increased mortality (p < 0.001). CTSS-Pan and the two other CTSS systems exerted similar performance. With respect to clinical outcomes, CTSS-Pan might have the best performance. Conclusion: CTSS may be suitable to assess severity and prognosis of COVID-19-associated pneumonia. CTSS and specific chest CT patterns may predict the need for ventilation, as well as mortality in COVID-19. This can help the physician to guide treatment strategies in COVID-19, as well as other pulmonary infections.

Pulmonary Functional Outcomes at 3 Months in Critical COVID-19 Survivors Hospitalized during the First, Second, and Third Pandemic Waves.

INTRODUCTION: Despite improved management of patients with COVID-19, we still ignore whether pharmacologic treatments and improved respiratory support have modified outcomes for intensive care unit (ICU) surviving patients of the three first consecutive waves (w) of the pandemic. The aim of this study was to evaluate whether developments in the management of ICU COVID-19 patients have positively impacted respiratory functional outcomes, quality of life (QoL), and chest CT scan patterns in ICU COVID-19 surviving patients at 3 months, according to pandemic waves. METHODS: We prospectively included all patients admitted to the ICU of two university hospitals with acute respiratory distress syndrome (ARDS) related to COVID-19. Data related to hospitalization (disease severity, complications), demographics, and medical history were collected. Patients were assessed 3 months post-ICU discharge using a 6 min walking distance test (6MWT), a pulmonary function test (PFT), a respiratory muscle strength (RMS) test, a chest CT scan, and a Short Form 36 (SF-36) questionnaire. RESULTS: We included 84 ARDS COVID-19 surviving patients. Disease severity, complications, demographics, and comorbidities were similar between groups, but there were more women in wave 3 (w3). Length of stay at the hospital was shorter during w3 vs. during wave 1 (w1) (23.4 ± 14.2 days vs. 34.7 ± 20.8 days, p = 0.0304). Fewer patients required mechanical ventilation (MV) during the second wave (w2) vs. during w1 (33.3% vs. 63.9%, p = 0.0038). Assessment at 3 months after ICU discharge revealed that PFTs and 6MWTs scores were worse for w3 > w2 > w1. QoL (SF-36) deteriorated (vitality and mental health) more for patients in w1 vs. in w3 (64.7 ± 16.3 vs. 49.2 ± 23.2, p = 0.0169). Mechanical ventilation was associated with reduced forced expiratory volume (FEV1), total lung capacity (TLC), diffusing capacity for carbon monoxide (DLCO), and respiratory muscle strength (RMS) (w1,2,3, p < 0.0500) on linear/logistic regression analysis. The use of glucocorticoids, as well as tocilizumab, was associated with improvements in the number of affected segments in chest CT, FEV1, TLC, and DLCO (p < 0.01). CONCLUSIONS: With better understanding and management of COVID-19, there was an improvement in PFT, 6MWT, and RMS in ICU survivors 3 months after ICU discharge, regardless of the pandemic wave during which they were hospitalized. However, immunomodulation and improved best practices for the management of COVID-19 do not appear to be sufficient to prevent significant morbidity in critically ill patients.

Pneumopathie sévère a SARS-CoV-2: aspects TDM à 6 mois

RÉSUMÉ Introduction – Depuis le début de la pandémie Covid-19, la TDM thoracique a joué un rôle clé dans le diagnostic, l'évaluation pronostique et le suivi de la pneumonie virale à SARS-CoV-2. Données récentes – La persistance de lésions pulmonaires sur la TDM de suivi d'une pneumopathie à Covid-19 grave concerne environ 60 % des patients. L'aspect TDM à 6 mois peut varier d'une résolution radiologique complète à des lésions fibrosantes pouvant imiter des pathologies interstitielles connues. Conclusion – Cet article illustre, à partir des images TDM de la cohorte Recovery from Covid-19 ardS (RECOVIDS) les aspects caractéristiques observés au suivi à 6 mois d'une pneumopathie sévère à SARS-CoV-2. En s'appuyant sur la présence de fibrose et le profil lésionnel prédominant, les radiologues peuvent reconnaître et intégrer ces images dans le diagnostic différentiel d'autres pneumopathies de présentations cliniques et TDM proches. SUMMARY Introduction- Since the beginning of the COVID-19 pandemic, chest computed tomography (CT) has played a key role in the diagnosis, prognostic evaluation, and follow-up of severe SARS-COVD-2 viral pneumonia. Recent Findings- Approximately 60 % of patients with severe COVID-19 exhibit persistent lung lesions on follow-up chest CT. The chest CT appearance at 6-month follow-up can range from complete resolution to severe fibrotic changes that may mimic known interstitial lung diseases. Conclusion- This article illustrates the typical appearance at 6-month using Chest-CT images from the "Recovery from COVID-19 ARDS” (RECOVIDS) cohort. An approach based on the presence of fibrosis and the predominant pattern will enable radiologists to recognize and incorporate these aspects into the differential diagnosis of other interstitial lung diseases that may present with similar clinical and CT features.

EVALUATION OF AUTOMATED QUANTIFICATION OF CT PATTERNS ASSOCIATED WITH COVID-19 FROM CHEST CT

Introduction: Computed Tomography (CT) is rapid and sensitive enough to identify COVID-19 pneumonia in its early stages. But because of the disease's high case load, it is difficult for the talented radiologists to report the cases. Therefore, using Artificial Intelligence (AI) to support radiologists' work will be crucial for producing prompt and precise results. Objective: To determine diagnostic effectiveness of AI in identifying different COVID-19 CT patterns and to correlate the AI findings with the findings appreciated by skilled Radiologists. Material and Methods: A prospective study consisting of 500 patients with RT-PCR positive COVID- 19 patients were evaluated, after obtaining informed consent. Data was analysed and represented in the form of frequencies and proportions. Collected data were analysed by Pearson's correlation coefficient (r), Intra Class Correlation (ICC) coefficient, Bland--Altman analysis. Results: AI can assess the severity of disease quickly and with good accuracy compared to manual analysis by decreasing the time taken to analyse the scan by 50%, and overall accuracy of approximately 90%. Conclusion: We conclude that as manual analysis of Chest CT in COVID-19 high case load scenario is comparatively more time-consuming, there is a need for a quick, accurate, and automated technique for identification and quantification of common findings in COVID-19. [ FROM AUTHOR] Copyright of Journal of Cardiovascular Disease Research (Journal of Cardiovascular Disease Research) is the property of Journal of Cardiovascular Disease Research 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.)

CT chest findings in patients infected with COVID-19: review of literature

BackgroundCoronavirus disease 2019 (COVID-19) is a highly infectious disease causing severe respiratory distress syndrome that was first discovered by the end of 2019 in Wuhan, China.Main textA wide variety of CT findings in COVID-19 have been reported in different studies, and the CT findings differ according to the stage of the disease and disease severity and associated co-morbidities. We will discuss each sign separately and its importance in diagnosis and prognosis.ConclusionCT plays a pivotal role in the diagnosis and management of COVID-19 pneumonia. The typical appearance of COVID-19 pneumonia is bilateral patchy areas of ground glass infiltration, more in the lower lobes. The appearance of other signs like consolidation, air bronchogram, crazy pavement appearance, and air bubble signs appear during the course of the disease. In the context of pandemic, the CT chest can be used as a screening tool in symptomatic patients as it is cheaper, available, and time saving.

Chest CT findings of COVID-19-infected patients, are there differences between pediatric and adult patients? A systematic review

BackgroundPurpose of this study was to deliver a report of chest CT findings of COVID-19-infected pediatric and adult patients and to make an age-based comparison. A systematic search was conducted in accordance with PRISMA guidelines to identify relevant studies in the electronic databases of PubMed, Scopus, ProQuest, ScienceDirect, and Web of Sciences from January 1, 2020 to March 27, 2020 using search terms in the titles and s. Based on our inclusion and exclusion criteria, 762 articles were screened. Finally, 15 eligible articles which had adequate data on chest CT findings of COVID-19-infected patients were enrolled in this systematic review.ResultsIn pediatric patients (15 years old or younger), peripheral distribution was found in 100% of cases, ground glass opacities (GGO) in 55.2%, bilateral involvement in 50%, halo sign in 50%, unilateral involvement in 30%, consolidation in 22.2%, crazy paving pattern in 20%, nodular opacities in 15%, pleural effusion in 4.2%, lymphadenopathy in none, and normal imaging in 20.8% of cases. On the other hand, in adult patients, bilateral involvement was reported in 76.8%, GGO in 68.4%, peripheral distribution in 62.2%, mixed GGO and consolidation in 48.7%, consolidation in 33.7%, crazy paving pattern in 27.7%, mixed central and peripheral distribution in 25.0%, unilateral involvement in 15.2%, nodular opacities in 9.2%, pleural effusion in 5.5%, central distribution of lesions in 5.4%, lymphadenopathy in 2.4%, and normal imaging in 9.8% of cases.ConclusionAccording to the findings of this systematic review, children infected with COVID-19 can present with normal or atypical findings (nodular opacities/unilateral involvement) in chest imaging more frequently than adult patients. Therefore, more caution should be taken to avoid misdiagnosis or missed diagnosis in infected children. Besides, clinical and laboratory findings need to be considered more decision-making for pediatric patients with normal or atypical chest CT scan but high suspicion of COVID-19.

Reducing radiation dose for NN-based COVID-19 detection in helical chest CT using real-time monitored reconstruction.

Computed tomography is a powerful tool for medical examination, which plays a particularly important role in the investigation of acute diseases, such as COVID-19. A growing concern in relation to CT scans is the radiation to which the patients are exposed, and a lot of research is dedicated to methods and approaches to how to reduce the radiation dose in X-ray CT studies. In this paper, we propose a novel scanning protocol based on real-time monitored reconstruction for a helical chest CT using a pre-trained neural network model for COVID-19 detection as an expert. In a simulated study, for the first time, we proposed using per-slice stopping rules based on the COVID-19 detection neural network output to reduce the frequency of projection acquisition for portions of the scanning process. The proposed method allows reducing the total number of X-ray projections necessary for COVID-19 detection, and thus reducing the radiation dose, without a significant decrease in the prediction accuracy. The proposed protocol was evaluated on 163 patients from the COVID-CTset dataset, providing a mean dose reduction of 15.1% while the mean decrease in prediction accuracy amounted to only 1.9% achieving a Pareto improvement over a fixed protocol.

Chest CT Findings after Mild COVID-19 Do Not Explain Persisting Respiratory Symptoms: An Explanatory Study.

(1) Background: Lung tissue involvement is frequently observed in acute COVID-19. However, it is unclear whether CT findings at follow-up are associated with persisting respiratory symptoms after initial mild or moderate infection. (2) Methods: Chest CTs of patients with persisting respiratory symptoms referred to the post-COVID-19 outpatient clinic were reassessed for parenchymal changes, and their potential association was evaluated. (3) Results: A total of 53 patients (31 female) with a mean (SD) age of 46 (13) years were included, of whom 89% had mild COVID-19. Median (quartiles) time from infection to CT was 139 (86, 189) days. Respiratory symptoms were dyspnea (79%), cough (42%), and thoracic pain (64%). Furthermore, 30 of 53 CTs showed very discrete and two CTs showed medium parenchymal abnormalities. No severe findings were observed. Mosaic attenuation (40%), ground glass opacity (2%), and fibrotic-like changes (25%) were recorded. No evidence for an association between persisting respiratory symptoms and chest CT findings was found. (4) Conclusions: More than half of the patients with initially mild or moderate infection showed findings on chest CT at follow-up. Respiratory symptoms, however, were not related to any chest CT finding. We, therefore, do not suggest routine chest CT follow-up in this patient group if no other indications are given.

Correction: CT-Angiographic Aspects of Pulmonary Embolism on SARS COV-2.

[This corrects the article DOI: 10.5334/jbsr.3021.].

Relationship of severity of COVID-19 disease with lung-CT-scan imaging

Objectives: Lung-CT-scan imaging is known as important diagnostic technique for evaluation of the effectiveness and infectious involvement of the lungs. In this study, we evaluated and analyzed lung CT images in patients with Coronavirus Disease 2019 (COVID-19) and its relationship with some important clinical and laboratory factors as well as the patient's condition in the worst disease conditions. Methods: In this retrospective descriptive-analytical study, 375 patients with complete information have been considered. Among these patients, CT scans of patients' lungs was carefully reevaluated. Other radiologist reviewed the images and recorded the final score of the patients' lung involvement. Results: Data showed that lung and cardiac involvement have high prevalence among studies patients. Among demographic variables, there was significant relationship between age and recovery. Evaluating the relationship of recovery with CT variables showed that CT score, bilateral lung involvement, and Crazy paving had significant effect on recovery rate. Conclusion: According to this study, evaluation of CT variables can be used as potent factors for evaluation of disease status and design of suitable treatment strategy.

CO-RADS score and its correlation with clinical and laboratory parameters in patients with COVID-19

Background Polymerase chain reaction (PCR) based SARS-CoV-2 RNA detection and serological antibody tests give a proof of Coronavirus Disease 2019 (COVID-19) infection. Several variables can influence the consequences of these tests. Inflammatory markers among mild and severe patients of COVID-19 showed dissimilarity in inflammatory markers while computed tomography (CT) in patients infected with COVID-19 used to evaluate infection severity. The aim of this study is to investigate the application of the COVID-19 Reporting and Data System (CO-RADS) classification in COVID-19 patients and its relation to clinical and laboratory finding. Results One hundred patients suspected to have COVID-19 infection were involved. Their age was 49.6 & PLUSMN;14.7. Fever and cough were the frequent presenting symptoms. Patients with positive PCR were significantly associated with dyspnea and higher inflammatory markers. Lymphopenia had sensitivity of 63.6% and specificity of 91.7%. Combination of PCR and lymphopenia increased both sensitivity and specificity. CT findings in relation to PCR showed sensitivity of 90.5% and specificity of 25%. CO-RADS score showed positive correlation with age and inflammatory biomarkers and negative correlation with absolute lymphocyte count (ALC). Conclusions CT finding was more prominent in older patients with COVID-19 and associated with higher inflammatory biomarkers and lower ALC which were correlated with CO-RADS score. Patients with positive PCR had more symptoms and inflammatory marker. Combination of PCR with either lymphopenia or CT finding had more sensitivity, specificity and accuracy in diagnosis

Development of an Artificial Intelligence Method to Detect COVID-19 Pneumonia in Computed Tomography Images

Introduction: This study aimed to construct an artificial intelligence system to detect Coronavirus disease-2019 (COVID-19) pneumonia on computed tomography (CT) images and to test its diagnostic performance. Methods: Data were acquired between March 18-April 17, 2020. CT data of 269 reverse tran-scriptase-polymerase chain reaction proven patients were extracted, and 173 studies (122 for training, 51 testing) were finally used. Most typical lesions of COVID-19 pneumonia were la-beled by two radiologists using a custom tool to generate multiplanar ground-truth masks. Us-ing a patch size of 128x128 pixels, 18,255 axial, 71,458 coronal, and 72,721 sagittal patches were generated to train the datasets with the U-Net network. Lesions were extracted in the or-thogonal planes and filtered by lung segmentation. Sagittal and coronal predicted masks were reconverted to the axial plane and were merged into the intersect-ed axial mask using a voting scheme. Results: Based on the axial predicted masks, the sensitivity and specificity of the model were found as 91.4% and 99.9%, respectively. The total number of positive predictions has increased by 3.9% by the use of intersected predicted masks, whereas the total number of negative predic-tions has only slightly decreased by 0.01%. These changes have resulted in 91.5% sensitivity, 99.9% specificity, and 99.9% accuracy. Conclusion: This study has shown the reliability of the U-Net architecture in diagnosing typical pulmonary lesions of COVID-19 in CT images. It also showed a slightly favorable effect of the intersection method to increase the model's performance. Based on the performance level pre-sented, the model may be used in the rapid and accurate detection and characterization of the typical COVID-19 pneumonia to assist radiologists.

Mathematical Morphology Adopted Automatic Detection of Ground Glass Opacities in Lung CT Images of COVID-19 Patients

The severity of the nCOVID infection relies on the presence of Ground Glass Opacities (GGO) present in the patient's chest CT scan images. Although, detecting and delineating the precise boundaries of GGO in the chest CT images is challenging. Here, we proposed a fast and novel technique to automatically segment the regions containing GGO in lung CT images using mathematical morphology. We have tested our algorithm on the chest CT images of 145 Covid-positive cases. This unique segmentation approach correctly segments the lung field from chest CT images and identifies GGO with average sensitivity, specificity, and accuracy of 96.89%, 95.23%, and 97.22%, respectively. We used expert radiologists' hand-curated segmentation of GGO as ground truth for quantificational performance analysis. Our research results indicate that this algorithm performs well found in the literature. © 2023 IEEE.