Effective multiscale deep learning model for COVID19 segmentation tasks: A further step towards helping radiologist.

Infection by the SARS-CoV-2 leading to COVID-19 disease is still rising and techniques to either diagnose or evaluate the disease are still thoroughly investigated. The use of CT as a complementary tool to other biological tests is still under scrutiny as the CT scans are prone to many false positives as other lung diseases display similar characteristics on CT scans. However, fully investigating CT images is of tremendous interest to better understand the disease progression and therefore thousands of scans need to be segmented by radiologists to study infected areas. Over the last year, many deep learning models for segmenting CT-lungs were developed. Unfortunately, the lack of large and shared annotated multicentric datasets led to models that were either under-tested (small dataset) or not properly compared (own metrics, none shared dataset), often leading to poor generalization performance. To address, these issues, we developed a model that uses a multiscale and multilevel feature extraction strategy for COVID19 segmentation and extensively validated it on several datasets to assess its generalization capability for other segmentation tasks on similar organs. The proposed model uses a novel encoder and decoder with a proposed kernel-based atrous spatial pyramid pooling module that is used at the bottom of the model to extract small features with a multistage skip connection concatenation approach. The results proved that our proposed model could be applied on a small-scale dataset and still produce generalizable performances on other segmentation tasks. The proposed model produced an efficient Dice score of 90% on a 100 cases dataset, 95% on the NSCLC dataset, 88.49% on the COVID19 dataset, and 97.33 on the StructSeg 2019 dataset as compared to existing state-of-the-art models. The proposed solution could be used for COVID19 segmentation in clinic applications. The source code is publicly available at https://github.com/RespectKnowledge/Mutiscale-based-Covid-_segmentation-usingDeep-Learning-models.

A DEEP LEARNING APPROACH LUNG SEGMENTATION AND PNEUMONIA DETECTION FROM X-RAYS

Lung segmentation is a process of detection and identification of lung cancer and pneumonia with the help of image processing techniques. Deep learning algorithms can be incorporated to build the computer-aided diagnosis (CAD) system for detecting or recognizing broad objects like acute respiratory distress syndrome (ARDS), Tuberculosis, Pneumonia, Lung cancer, Covid, and several other respiratory diseases. This paper presents pneumonia detection from lung segmentation using deep learning methods on chest radiography. Chest X-ray is the most useful technique among other existing techniques, due to its lesser cost. The main drawback of a chest x-ray is that it cannot detect all problems in the chest. Thus, implementing convolutional neural networks (CNN) to perform lung segmentation and to obtain correct results. The 'lost' regions of the lungs are reconstructed by an automatic segmentation method from raw images of chest X-ray. © 2022 IEEE.

Application of Convolutional Neural Networks for COVID-19 Detection in X-ray Images Using InceptionV3 and U-Net.

COVID-19 has expanded overall across the globe after its initial cases were discovered in December 2019 in Wuhan-China. Because the virus has impacted people's health worldwide, its fast identification is essential for preventing disease spread and reducing mortality rates. The reverse transcription polymerase chain reaction (RT-PCR) is the primary leading method for detecting COVID-19 disease; it has high costs and long turnaround times. Hence, quick and easy-to-use innovative diagnostic instruments are required. According to a new study, COVID-19 is linked to discoveries in chest X-ray pictures. The suggested approach includes a stage of pre-processing with lung segmentation, removing the surroundings that do not provide information pertinent to the task and may result in biased results. The InceptionV3 and U-Net deep learning models used in this work process the X-ray photo and classifies them as COVID-19 negative or positive. The CNN model that uses a transfer learning approach was trained. Finally, the findings are analyzed and interpreted through different examples. The obtained COVID-19 detection accuracy is around 99% for the best models.

Segmentation and classification on chest radiography: a systematic survey.

Chest radiography (X-ray) is the most common diagnostic method for pulmonary disorders. A trained radiologist is required for interpreting the radiographs. But sometimes, even experienced radiologists can misinterpret the findings. This leads to the need for computer-aided detection diagnosis. For decades, researchers were automatically detecting pulmonary disorders using the traditional computer vision (CV) methods. Now the availability of large annotated datasets and computing hardware has made it possible for deep learning to dominate the area. It is now the modus operandi for feature extraction, segmentation, detection, and classification tasks in medical imaging analysis. This paper focuses on the research conducted using chest X-rays for the lung segmentation and detection/classification of pulmonary disorders on publicly available datasets. The studies performed using the Generative Adversarial Network (GAN) models for segmentation and classification on chest X-rays are also included in this study. GAN has gained the interest of the CV community as it can help with medical data scarcity. In this study, we have also included the research conducted before the popularity of deep learning models to have a clear picture of the field. Many surveys have been published, but none of them is dedicated to chest X-rays. This study will help the readers to know about the existing techniques, approaches, and their significance.

Deep Learning based Lung Segmentation Prior for Robust COVID-19 Classification

Independent of a person's race, COVID-19 is one of the most contagious diseases in the world. The World Health Organization classified the COVID-19 outbreak as a pandemic after noting its global distribution. By using (i) sample-supported analysis and (ii) image-assisted diagnosis, COVID-19 is examined and verified. Our goal is to use CT scan images to identify the COVID-19 infiltrates. The followings steps are used to carry out the suggested work: (i) Automated segmentation with CNN;(ii) Feature mining;(iii) Principal feature selection with Bat-Algorithm;(iv) Classifier implementation using mobile framework and (v) Performance evaluation. We used a variety of automatic segmentation algorithms in our experiment, and the VGG-16 produced better results. This study is evaluated using benchmark datasets gathered, and SVM based RBF kernal classifier system resulted in superior COVID-19 abnormality identification. © 2023 IEEE.

Automated semantic lung segmentation in chest CT images using deep neural network.

Lung segmentation algorithms play a significant role in segmenting theinfected regions in the lungs. This work aims to develop a computationally efficient and robust deep learning model for lung segmentation using chest computed tomography (CT) images with DeepLabV3 + networks for two-class (background and lung field) and four-class (ground-glass opacities, background, consolidation, and lung field). In this work, we investigate the performance of the DeepLabV3 + network with five pretrained networks: Xception, ResNet-18, Inception-ResNet-v2, MobileNet-v2 and ResNet-50. A publicly available database for COVID-19 that contains 750 chest CT images and corresponding pixel-labeled images are used to develop the deep learning model. The segmentation performance has been assessed using five performance measures: Intersection of Union (IoU), Weighted IoU, Balance F1 score, pixel accu-racy, and global accuracy. The experimental results of this work confirm that the DeepLabV3 + network with ResNet-18 and a batch size of 8 have a higher performance for two-class segmentation. DeepLabV3 + network coupled with ResNet-50 and a batch size of 16 yielded better results for four-class segmentation compared to other pretrained networks. Besides, the ResNet with a fewer number of layers is highly adequate for developing a more robust lung segmentation network with lesser computational complexity compared to the conventional DeepLabV3 + network with Xception. This present work proposes a unified DeepLabV3 + network to delineate the two and four different regions automatically using CT images for CoVID-19 patients. Our developed automated segmented model can be further developed to be used as a clinical diagnosis system for CoVID-19 as well as assist clinicians in providing an accurate second opinion CoVID-19 diagnosis.

Covid-19 Infection Segmentation Using Deep Learning Techniques

The rapid spread of the disease after COVID-19's emergence in 2019 has presented enormous problems to medical institutions. The diagnosis process will go more rapidly if the infected region in the COVID-19 CT image can be automatically segmented, which will aid clinicians in promptly identifying the patient's illness. Automated lung infection identification using computed tomography scans is a more general approach. However, segmenting sick areas from CT slices is quite difficult. In this work, a diagnosis system based on deep learning methods is being created to identify and quantify COVID-19 infection and screen for pneumonia using CT imaging. Here, Unet++ approaches, U-net architecture based on CNN encoder and CNN decoder, and Attention Unet segmentation techniques are used. These methods are applied for quick and accurate picture segmentation to produce segmentation models for lung and infection. Fourfold cross-validation has been used as a re-sampling method to improve skill estimate on unseen data. To enable volume ratio calculating and determine infection rate, the lung and infection volumes have been reconfigured. 20 CT scan cases were used in this study, and the data were split into two, training dataset 70% and a validation dataset 30%. In this study with three architectures it shows that basic Unet performs well compared to other two architectures. © 2022 IEEE

Optimization of CNN for Diagnosis on Lung Disease by Lung Segmentation and Rib Suppression

The discrimination of lung diseases by chest X- ray images is a clinically important tool. How to use artificial intelligence to accurately and quickly help doctors to diagnose different lung diseases is very important in the context of the current COVID-19 global pandemic. In this paper, we propose a model structure, including two U-Net, which implement lung segmentation and rib suppression for chest X-ray images respectively, image enhancement techniques such as histogram equalization, which enhances images contrast, and a Xception- based CNN, which classifies the processed images finally. The model can effectively avoid the interference of regions outside the lung to CNN for feature recognition and the influence of environmental factors such as X-ray machines on the quality of X-ray images and thus on the classification. The experimental results show that the classification accuracy of the model is higher than that of the direct use of the Xception model for classification. © 2022 IEEE.

DeepChestNet: Artificial intelligence approach for COVID-19 detection on computed tomography images

The conventional approach for identifying ground glass opacities (GGO) in medical imaging is to use a convolutional neural network (CNN), a subset of artificial intelligence, which provides promising performance in COVID-19 detection. However, CNN is still limited in capturing structured relationships of GGO as the texture and shape of the GGO can be confused with other structures in the image. In this paper, a novel framework called DeepChestNet is proposed that leverages structured relationships by jointly performing segmentation and classification on the lung, pulmonary lobe, and GGO, leading to enhanced detection of COVID-19 with findings. The performance of DeepChestNet in terms of dice similarity coefficient is 99.35%, 99.73%, and 97.89% for the lung, pulmonary lobe, and GGO segmentation, respectively. The experimental investigations on DeepChestNet-Lung, DeepChestNet-Lobe and DeepChestNet-COVID datasets, and comparison with several state-of-the-art approaches reveal the great potential of DeepChestNet for diagnosis of COVID-19 disease. © 2023 Wiley Periodicals LLC.

Automated Quantification of Pneumonia Infected Volume in Lung CT Images: A Comparison with Subjective Assessment of Radiologists.

OBJECTIVE: To help improve radiologists' efficacy of disease diagnosis in reading computed tomography (CT) images, this study aims to investigate the feasibility of applying a modified deep learning (DL) method as a new strategy to automatically segment disease-infected regions and predict disease severity. METHODS: We employed a public dataset acquired from 20 COVID-19 patients, which includes manually annotated lung and infections masks, to train a new ensembled DL model that combines five customized residual attention U-Net models to segment disease infected regions followed by a Feature Pyramid Network model to predict disease severity stage. To test the potential clinical utility of the new DL model, we conducted an observer comparison study. First, we collected another set of CT images acquired from 80 COVID-19 patients and process images using the new DL model. Second, we asked two chest radiologists to read images of each CT scan and report the estimated percentage of the disease-infected lung volume and disease severity level. Third, we also asked radiologists to rate acceptance of DL model-generated segmentation results using a 5-scale rating method. RESULTS: Data analysis results show that agreement of disease severity classification between the DL model and radiologists is >90% in 45 testing cases. Furthermore, >73% of cases received a high rating score (≥4) from two radiologists. CONCLUSION: This study demonstrates the feasibility of developing a new DL model to automatically segment disease-infected regions and quantitatively predict disease severity, which may help avoid tedious effort and inter-reader variability in subjective assessment of disease severity in future clinical practice.

3D Visualization for Lung Surface Images of Covid-19 Patients based on U-Net CNN Segmentation

The Covid-19 infection challenges medical staff to make rapid diagnoses of patients. In just a few days, the Covid-19 virus infection could affect the performance of the lungs. On the other hand, semantic segmentation using the Convolutional Neural Network (CNN) on Lung CT-scan images had attracted the attention of researchers for several years, even before the Covid-19 pandemic. Ground Glass Opacity (GGO), in the form of white patches caused by Covid-19 infection, is detected inside the patient's lung area and occasionally at the edge of the lung, but no research has specifically paid attention to the edges of the lungs. This study proposes to display a 3D visualization of the lung surface of Covid-19 patients based on CT-scan image segmentation using U-Net architecture with a training dataset from typical lung images. Then the resulting CNN model is used to segment the lungs of Covid-19 patients. The segmentation results are selected as some slices to be reconstructed into a 3D lung shape and displayed in 3D animation. Visualizing the results of this segmentation can help medical staff diagnose the lungs of Covid-19 patients, especially on the surface of the lungs of patients with GGO at the edges. From the lung segmentation experiment results on ten patients in the Zenodo dataset, we have a Mean-IoU score = of 76.86%, while the visualization results show that 7 out of 10 patients (70%) have eroded lung surfaces. It can be seen clearly through 3D visualization.

Automated Detection of Broncho-Arterial Pairs Using CT Scans Employing Different Approaches to Classify Lung Diseases.

Current research indicates that for the identification of lung disorders, comprising pneumonia and COVID-19, structural distortions of bronchi and arteries (BA) should be taken into account. CT scans are an effective modality to detect lung anomalies. However, anomalies in bronchi and arteries can be difficult to detect. Therefore, in this study, alterations of bronchi and arteries are considered in the classification of lung diseases. Four approaches to highlight these are introduced: (a) a Hessian-based approach, (b) a region-growing algorithm, (c) a clustering-based approach, and (d) a color-coding-based approach. Prior to this, the lungs are segmented, employing several image preprocessing algorithms. The utilized COVID-19 Lung CT scan dataset contains three classes named Non-COVID, COVID, and community-acquired pneumonia, having 6983, 7593, and 2618 samples, respectively. To classify the CT scans into three classes, two deep learning architectures, (a) a convolutional neural network (CNN) and (b) a CNN with long short-term memory (LSTM) and an attention mechanism, are considered. Both these models are trained with the four datasets achieved from the four approaches. Results show that the CNN model achieved test accuracies of 88.52%, 87.14%, 92.36%, and 95.84% for the Hessian, the region-growing, the color-coding, and the clustering-based approaches, respectively. The CNN with LSTM and an attention mechanism model results in an increase in overall accuracy for all approaches with an 89.61%, 88.28%, 94.61%, and 97.12% test accuracy for the Hessian, region-growing, color-coding, and clustering-based approaches, respectively. To assess overfitting, the accuracy and loss curves and k-fold cross-validation technique are employed. The Hessian-based and region-growing algorithm-based approaches produced nearly equivalent outcomes. Our proposed method outperforms state-of-the-art studies, indicating that it may be worthwhile to pay more attention to BA features in lung disease classification based on CT images.

Lung Segmentation Algorithm and SVM Classification of COVID-19 in CT Images

The analysis of COVID-19 by tomographic imaging has been a standard for pandemic management. The application of different types of artificial intelligence algorithms has proven to be an accurate method for disease detection. This study presents a method of lung segmentation and a classification algorithm that allows to discriminate between images that show signs of the disease and those that don’t. In addition, the article seeks to establish what kind of features are relevant when feeding a machine learning algorithm. Texture features extracted from Gray Label Concurrence Matrix (GLCM) and a Gabor filter are used for this purpose. Then, we trained and evaluated a SVM algorithm using different combinations of features. It is found that the features extracted from the Gabor filter work better than those extracted from the GLCM, finding that those features focused exclusively on intensity description work better than those focused on spatial description, at least in early stages. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

CT-based pathological lung opacities volume as a predictor of critical illness and inflammatory response severity in patients with COVID-19.

Objective: The aim of the study was to assess the impact of CT-based lung pathological opacities volume on critical illness and inflammatory response severity of patients with COVID-19. Methods: A retrospective, single center, single arm study was performed over a 30-day period. In total, 138 patients (85.2%) met inclusion criteria. All patients were evaluated with non-contrast enhanced chest CT scan at hospital admission. CT-based lung segmentation was performed to calculate pathological lung opacities volume (LOV). At baseline, complete blood count (CBC) and inflammation response biomarkers were obtained. The primary endpoint of the study was the occurrence of critical illness, as defined as, the need of mechanical ventilation and/or ICU admission. Mann-Whitney U test was performed for univariate analysis. Logistic regression analysis was performed to determine independent predictors of critical illness. Spearman analysis was performed to assess the correlation between inflammatory response biomarkers serum concentrations and LOV. Results: Median LOV was 28.64% (interquartile range [IQR], 6.33-47.22%). Correlation analysis demonstrated that LOV was correlated with higher levels of D-dimer (r = 0.51, p < 0.01), procalcitonin (r = 0.47, p < 0.01) and IL6 (r = 0.48, p < 0.01). Critical illness occurred in 51 patients (37%). Univariate analysis demonstrated that inflammatory response biomarkers and LOV were associated with critical illness (p < 0.05). However, multivariate analysis demonstrated that only D-dimer and LOV were independent predictors of critical illness. Furthermore, a ROC analysis demonstrated that a LOV equal or greater than 60% had a sensitivity of 82.1% and specificity of 70.2% to determine critical illness with an odds ratio of 19.4 (95% CI, 4.2-88.9). Conclusion: Critical illness may occur in up to 37% of the patients with COVID-19. Among patients with critical illness, higher levels of inflammatory response biomarkers with larger LOVs were observed. Furthermore, multivariate analysis demonstrated that pathological lung opacities volume was an independent predictor of critical illness. In fact, patients with a pathological lung opacities volume equal or greater than 60% had 19.4-fold increased risk of critical illness.

Lung Segmentation using MultiResUNet CNN based on Computed Tomography Image

Lung segmentation is the first step in medical image processing to determine various lung diseases. Currently, the image segmentation process will be more optimal by using deep learning through the convolution process. Various Convolution Neural Network (CNN) based architectures for image segmentation were created by many researchers, however U-Net is the current state of the art for medical image segmentation. Nevertheless, the modification of U-Net continues, and MultiResUNet is one of the new architectures claimed to be better. In this study, we use MultiResUNet for lung segmentation on Computed Tomography (CT) images as the first step to Covid-19 infection segmentation, and the results will be compared using the U-Net architecture. Based on the results of the segmentation experiment, we got satisfactory results. Using the Mean-IoU evaluation metric, it was concluded that the MultiResUNet score was slightly better than the U-Net score for patient lung segmentation, where there was an increase in the score of 1.33% (MultiResUNet=93.05%, U-Net=91.83%) in the dataset which we use. © 2022 IEEE.

Covid-19 Lung Segmentation using U-Net CNN based on Computed Tomography Image

In medical image analysis, lung segmentation is needed as an initial step in diagnosing various diseases in the lung area, including Covid-19 infection. Deep Learning has been used for image segmentation in recent years. One of the Deep Learning-based architectures widely used in medical image segmentation is U-Net CNN. U-Net employs a semantic segmentation approach, which has the benefit of being accurate in segmenting even though the model is trained on a limited quantity of data. Our work intends to assist radiologists in providing a more detailed visualization of COVID-19 infection on CT scans, including infection categories and lung conditions. We conduct preliminary work to segment lung regions using U-Net CNN. The dataset used is relatively small, consisting of 267 CT-scan images split into 240 (90%) images for training and 27 (10%) images for testing. The model is evaluated using the K-fold cross-validation (k=10) approach, which has been believed to be appropriate for models created with limited training data. The metric used for experiments is Mean-IoU. It is commonly used in evaluating the segmentation processes. The results achieved were satisfactory, with Mean-IoU scores ranging from 90.2% to 95.3% in each test phase (k1 – k10), with an average value of 93.3%. © 2022 IEEE.

CT Intensity Segmentation of Lungs

The early diagnosis and treatment of lung diseases is a very critical procedure and it requires the use of Computed Tomography (CT) imaging for the segmentation of lungs. Segmentation of the lung helps in the analysis of the lesions. The project proposes a CT lung and vessel segmentation model without any labels which is based on medical image processing using Python. This would assist the medical practitioners and scientists who are working in the field of CT intensity segmentation of lungs. It would make the diagnosis process easier and more convenient for patients, especially in pandemic situations like COVID. © 2022 IEEE.

Automated Lung Segmentation On Chest CT Using Deep Learning

A pandemic of coronavirus disease 2019 (COVID-19) devastated humanity before the end of 2019, which was triggered by severe acute respiratory syndrome (SARS), which originated in Wuhan, China. This sickness has claimed the lives of many people. The pandemic's consequences have been more severe in the world's most populous countries. Despite the fact that over a billion immunizations have been distributed to Indian citizens, the epidemic has not abated as of October 21, 2021. While certain limits are being eased, the threat of the dreaded "fourth wave"remains. In these situations, having technologies for swift disease testing and diagnosis is critical, since it allows for a much speedier process. Using a CT scan of the lungs, this research will provide vision into a model that may proficiently and correctly envisage the existence of COVID-19. © 2022 IEEE.

ILC-Unet++ for Covid-19 Infection Segmentation

Since the appearance of Covid-19 pandemic, in the end of 2019, Medical Imaging has been widely used to analysis this disease. In fact, CT-scans of the Lung can help to diagnosis, detect and quantify Covid-19 infection. In this paper, we address the segmentation of Covid-19 infection from CT-scans. In more details, we propose a CNN-based segmentation architecture named ILC-Unet++. The proposed ILC-Unet++ architecture, which is trained for both Covid-19 Infection and Lung Segmentation. The proposed architecture were tested using three datasets with two scenarios (intra and cross datasets). The experimental results showed that the proposed architecture performs better than three baseline segmentation architectures (Unet, Unet++ and Attention-Unet) and two Covid-19 infection segmentation architectures (SCOATNet and nCoVSegNet). © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Image data augmentation for improving performance of deep learning-based model in pathological lung segmentation

Accurate segmentation of lung fields from chest Xray (CXR) images is very important for subsequent analysis of many pulmonary diseases. Deep Neural Networks (DNN)-based methods have achieved remarkable progress in many image related tasks. However, their performance depends highly on the distribution of training and test samples, and they perform well if both training and test samples are from the same distribution. For example, DNN-based lung segmentation methods perform well on segmentation of healthy lung or lung with mild disease, however their performance is poor on lungs with severe abnormalities. Pulmonary opacification, which blurs the lung boundary, is one of the main reasons. A solution to this problem is data augmentation to increase the pool of training images, however despite the great success of traditional data augmentation techniques for natural images, they are not very effective for medical images. To simulate CXR images with opacification and low contrast, we present a novel image data augmentation technique in this study. To generate an augmented image, we first generate a random area inside the lung and then blur the area with a gaussian filter. Then, low contrast is simulated by adjusting the contrast and brightness. To evaluate the utility of the proposed augmentation technique, we applied it to images with different pulmonary diseases such as tuberculosis, pneumoconiosis and covid-19 from three public datasets as well as a private dataset and compared its effect on segmentation performance with traditional data augmentation techniques. Results suggest that the proposed technique outperforms traditional data augmentation techniques for all datasets on lung segmentation, in terms of Dice Coefficient (DC) and Jaccard Index (JI). Extensive experiments on multiple datasets validate the effectiveness of the proposed data augmentation technique.