Lung Computed Tomography Images Research Articles

End-to-end Prediction of EGFR Mutation Status with Denseformer.

Accurate genotyping of the epidermal growth factor receptor (EGFR) is critical for the treatment planning of lung adenocarcinoma. Currently, clinical identification of EGFR genotyping highly relies on biopsy and sequence testing which is invasive and complicated. Recent advancements in the integration of computed tomography (CT) imagery with deep learning techniques have yielded a non-invasive and straightforward way for identifying EGFR profiles. However, there are still many limitations for further exploration: 1) most of these methods still require physicians to annotate tumor boundaries, which are time-consuming and prone to subjective errors; 2) most of the existing methods are simply borrowed from computer vision field which does not sufficiently exploit the multi-level features for final prediction. To solve these problems, we propose a Denseformer framework to identify EGFR mutation status in a real end-to-end fashion directly from 3D lung CT images. Specifically, we take the 3D whole-lung CT images as the input of the neural network model without manually labeling the lung nodules. This is inspired by the medical report that the mutational status of EGFR is associated not only with the local tumor nodules but also with the microenvironment surrounded by the whole lung. Besides, we design a novel Denseformer network to fully explore the distinctive information across the different level features. The Denseformer is a novel network architecture that combines the advantages of both convolutional neural network (CNN) and Transformer. Denseformer directly learns from the 3D whole-lung CT images, which preserves the spatial location information in the CT images. To further improve the model performance, we designed a combined Transformer module. This module employs the Transformer Encoder to globally integrate the information of different levels and layers and use them as the basis for the final prediction. The proposed model has been tested on a lung adenocarcinoma dataset collected at the Affiliated Hospital of Zunyi Medical University. Extensive experiments demonstrated the proposed method can effectively extract meaningful features from 3D CT images to make accurate predictions. Compared with other state-of-the-art methods, Denseformer achieves the best performance among current methods using deep learning to predict EGFR mutation status based on a single modality of CT images.

An efficient lung image classification and detection using spiral‐optimized Gabor filter with convolutional neural network

AbstractLung cancer has a high death rate of around seven million cases every year worldwide. A computed tomography (CT) scan provides certain essential data concerning lung diseases and their diagnosis. The main objective of this work is to classify various lung diseases such as Normal, Bronchiectasis, and Pleural Effusion. The proposed approach consists of three stages, namely pre‐processing, feature extraction, and classification. At first, CT lung images are collected from the dataset and pre‐processed. After pre‐processing, the important texture features are extracted from each image. For feature extraction, spiral‐optimized Gabor filter (SOGF) is utilized. The proposed SOGF is a combination of spiral optimization algorithm (SOA) and Gabor filter (GF). Then, the extracted features are given to the convolutional neural network (CNN) to classify different types of lung diseases. For comparison, we use different classifiers, namely artificial neural network (ANN), Random Tree, and the Naïve Bayes. The experimental results show that our proposed approach attained the maximum accuracy of 93.67% which is high compared to other methods.

Examination of COVID-19 pneumonia severity and lung computed tomography findings

Purpose: Early recognition of corona virus disease 2019 (COVID-19) cases is the key to reducing mortality. In this study, we aimed to describe the relationship between the clinical, laboratory, and lung computed tomography (LCT) characteristics of patients with COVID-19 pneumonia and determine the severity of pneumonia in these patients. Methods: The pneumonia severity index (PSI) score system, LCT images, and laboratory parameters at the time of first presentation to the emergency department were examined to assess the severity of COVID-19 pneumonia in 225 adult patients. Results: In this study, a significant relationship was found between COVID-19-associated mortality and male gender (p=0.045), advanced age (p

Blend U-Net: Redesigning Skip Connections to Obtain Multiscale Features for Lung CT Images Segmentation.

Lung cancer is a pervasive and persistent issue worldwide, with the highest morbidity and mortality among all cancers for many years. In the medical field, computer tomography (CT) images of the lungs are currently recognized as the best way to help doctors detect lung nodules and thus diagnose lung cancer. U-Net is a deep learning network with an encoder-decoder structure, which is extensively employed for medical image segmentation and has derived many improved versions. However, these advancements do not utilize various feature information from all scales, and there is still room for future enhancement. In this study, we proposed a new model called Blend U-Net, which incorporates nested structures, redesigned long and short skip connections, and deep supervisions. The nested structures and the long and short skip connections combined characteristic information of different levels from feature maps in all scales, while the deep supervision learning hierarchical representations from all-scale concatenated feature maps. Additionally, we employed a mixed loss function to obtain more accurate results. We evaluated the performance of the Blend U-Net against other architectures on the publicly available Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI) dataset. Moreover, the accuracy of the segmentation was verified by using the dice coefficient. Blend U-Net with a boost of 0.83 points produced the best outcome in a number of baselines. Based on the results, our method achieves superior performance in terms of dice coefficient compared to other methods and demonstrates greater proficiency in segmenting lung nodules of varying sizes.

Abstract 16513: Decreased Pulmonary Blood Flow and Airway Volumes in Patients With Long COVID Syndrome Assessed by Functional Respiratory Imaging

Introduction: In contrast to normal chest X-ray, lung computed tomography (CT), and physiological lung and cardiac functions, many patients with long COVID syndrome suffer from shortness of breath. Hypothesis: The aim of this study was to quantify the pulmonary blood and airway volumes of long COVID patients compared with that of healthy controls. Methods: Patients with long COVID syndromes were included if they had PCR-verified previous (≥3 months) SARS-CoV-2 infection, had normal laboratory (e.g. inflammation, coagulation, cardiac or other organ) parameter, normal pulmonary morphology (chest X-ray and CT) and function (spirometry and body plethysmography). The lung CT images were postprocessed by Functional Respiratory imaging analysis by using 3D reconstruction with automated lung vessel segmentation algorithm. Data of the quantitative images were compared with age, gender, and BMI-matched healthy controls. Results: Thirty patients (45±13 years, 37% male, 25.9±4.3 kg/m^2) at a median time of 256 (118-574) days after a confirmed COVID infection and 30 healthy controls (55±7y, 37% male, 26.3±2.7 kg/m^2) were included. All long COVID patients complained of dyspnoea and 14 (48.3%) patients reported thoracic pain. The total pulmonary blood volume was significantly lower in the long COVID patients compared to controls (190±24.3 mL/m^2 vs230.6±26.2 ml/m^2, p<0.001). Similarly, the capillary-small vessel blood flow (vessel cross sectional area <5 mm^2) was reduced in the long COVID population (118±19 mL vs 132±23 mL, p=0.011). (Figure). The specific image-based airway volume of the distal lung regions was lower than that of the healthy population (11.1±6.74 mL vs 17.33±7.7 mL, p<0.05). Conclusions: Both the reduced global and capillary pulmonary blood flow, and distal airway volumes indicate impaired gas exchange and might explain the pulmonary complaints of patient with long COVID syndromes even severe months after Coronavirus infection.

Abstract 12452: Risk Stratification in Cardiogenic Shock Patients With Venoarterial Extracorporeal Membrane Oxygenation Using Lung Computed Tomography Density and Model for Endstage Liver Disease Score

Background: End-organ dysfunction often occur in patients receiving femoro-femoral veno-arterial extracorporeal membrane oxygenation (VA ECMO) and it has been reported to be associated with poor prognosis. This study aimed to elucidate the ability to risk stratification in patients with cardiogenic shock (CS) who received VA ECMO using lung computed tomography (CT) density and Model for End-stage Liver Disease exclude INR (MELD-XI) score. Methods: This is a single-center retrospective observational study targeted 65 CS patients who has had available chest CT images on VA ECMO support followed by MCS escalation from 2012 to 2021. The average density of lung CT images in the lung field at the level of tracheal bifurcation was measured using region-of-interest methods. MELD-XI scores were also calculated. The primary endpoint was 180-day all cause death after MCS escalation. Results: Twenty-two patients (34%) died within 180 days during the follow-up period [659 (78-1,544) days]. According to the Cox regression analysis higher MELD-XI scores (≥ 18) (HR; 2.68; 95% CI, 1.05-6.86; p=0.039), and higher lung CT density [≥ -481 Hounsfield unit (HU)] (HR, 5.64; 95% CI, 1.95-16.28; p=0.001) were independently associated with all cause death. Moreover, combination of both lung CT density and MELD-XI score further refine risk stratification in current patient population : group A, high lung CT density (≥-481 HU) and high MELD-XI score (≥18); group B, high lung CT density (≥-481 HU) or high MELD-XI score (≥18); and group C, low lung CT density (<-481 HU) and low MELD-XI score (<18) (group A:31.3% vs. group B:65.2% vs. group C:88.5%, log-rank p<0.001). Conclusions: Combination of higher lung CT density and higher MELD-XI score is a useful predictor of prognosis in patients with VA ECMO requiring escalation of MCS.

“Three medicines and three formulas” in COVID-19: From bench to bedside

Chinese guideline has been proven effective in the fight against Coronavirus disease 2019 (COVID-19) during the epidemic spread globally. Traditional Chinese medicine (TCM) has been widely recognized for its effectiveness in alleviating symptoms, inhibiting disease deterioration, reducing mortality, and improving cure rate of COVID-19 patients. During the pandemic, “three medicines and three formulas” stood out from hundreds of registered clinical studies and became the highly recommended TCM for COVID-19 treatment. The “three medicines and three formulas” not only effectively relieve the clinical symptoms of fever, cough, fatigue, and phlegm, but also significantly shorten the time of nucleic acid negative conversion, improve lung computed tomography imaging feature and inflammation, ameliorate clinical biochemical indicators, and reduce sequelae. The potential pharmacological mechanisms of them are mainly relevant with the crosstalk of viral toxicity, endothelial damage, cytokine storm, immune response, and microthrombus. In brief, the clinical effects as well as the potential mechanisms of “three medicines and three formulas” on COVID-19 were systematically analyzed and summarized covering the whole stages of disease development, including virus invasion and replication, immune response and cytokine storm, and acute respiratory distress syndrome and multiple organ dysfunction syndrome. We hope that this review could provide theoretical basis and reference for in-depth understanding the positive role of “three medicines and three formulas” for COVID-19 treatment.

Hybrid unsupervised paradigm based deformable image fusion for 4D CT lung image modality

Deformable image registration plays a critical role in various clinical applications (e.g., image fusion, atlas creation, and tumors targeting). In radiation therapy, especially in the context of fast registration of computed tomography (CT) lung image modalities, it is used to determine the geometric transformation by relating the anatomic points in two images. The main challenge lies in effectively addressing the nonlinear large and small deformation between the inspiration and expiration phases. In this work, we propose an unsupervised hybrid paradigm-based registration network (HPRN) for the registration of 4D CT lung images without relying on ground truth data. The proposed HPRN exhibits effective learning of multi-scale and multi-resolution features, leading to the computation of a more accurate Deformation Vector Field (DVF). Furthermore, we incorporate the regularization, image similarity and Jacobian determinant loss functions, which results in improving capability in dealing with complex large and small deformations. We evaluate the effectiveness of the proposed model on the publicly accessible DIRLab 4DCT lung image dataset, which shows the effectiveness of the proposed framework by achieving better Target Registration Error (2.04±1.42mm) compared to other state-of-the-art unsupervised image registration algorithms.

Improved donor lung size matching by estimation of lung volumes based on chest X-ray measurements.

Organ size matching is an important determinant of successful allocation and outcomes in lung transplantation. While computed tomography (CT) is the gold standard, it is rarely used in an organ-donor context, and chest X-ray (CXR) may offer a practical and accurate solution in estimating lung volumes for donor and recipient size matching. We compared CXR lung measurementsto CT-measured lung volumes and traditional estimates of lung volume in the same subjects. Our retrospective study analyzed clinically obtained CXR and CT lung images of 250 subjects without evidence of lung disease (mean age 9.9 ± 7.8 years; 129 M/121F). From CT, each lung was semi-automatically segmented and total lung volumes were quantified. From anterior-posterior CXR view, each lung was manually segmented and areas were measured. Lung lengths from the apices to the mid-basal regions of each lung were measured from CXR. Quantified CT lung volumes were compared to the corresponding CXR lung lengths, CXR lung areas, height, weight, and predicted total lung capacity (pTLC). There are strong and significant correlations between CT volumes and CXR lung areas in the right lung (R2 = .89, p < .0001), left lung (R2 = .87, p < .0001), and combined lungs (R2 = .89, p < .0001). Similar correlations were seen between CT volumes and CXR measured lung lengths in the right lung (R2 = .79, p < .0001) and left lung (R2 = .81, p < .0001). This correlation between anatomical lung volume (CT) and CXR was stronger than lung-volume correlation to height (R2 = .66, p < .0001), weight (R2 = .43, p < .0001), or pTLC (R2 = .66, p < .0001). CXR measures correlate much more strongly with true lung volumes than height, weight, or pTLC. The ability to obtain efficient and more accurate lung volume via CXR has the potential to change our current listing practices of using height as a surrogate for lung size, with a case example provided.

Attention U-net for automated pulmonary fissure integrity analysis in lung computed tomography images

Computed Tomography (CT) imaging is routinely used for imaging of the lungs. Deep learning can effectively automate complex and laborious tasks in medical imaging. In this work, a deep learning technique is utilized to assess lobar fissure completeness (also known as fissure integrity) from pulmonary CT images. The human lungs are divided into five separate lobes, divided by the lobar fissures. Fissure integrity assessment is important to endobronchial valve treatment screening. Fissure integrity is known to be a biomarker of collateral ventilation between lobes impacting the efficacy of valves designed to block airflow to diseased lung regions. Fissure integrity is also likely to impact lobar sliding which has recently been shown to affect lung biomechanics. Further widescale study of fissure integrity’s impact on disease susceptibility and progression requires rapid, reproducible, and noninvasive fissure integrity assessment. In this paper we describe IntegrityNet, an attention U-Net based automatic fissure integrity analysis tool. IntegrityNet is able to predict fissure integrity with an accuracy of 95.8%, 96.1%, and 89.8% for left oblique, right oblique, and right horizontal fissures, compared to manual analysis on a dataset of 82 subjects. We also show that our method is robust to COPD severity and reproducible across subject scans acquired at different time points.

Lung and colon cancer detection from CT images using Deep Learning

Cancer is a deadly disease that has gained a reputation as a global health concern. Further, lung cancer has been widely reported as the most deadly cancer type globally, while colon cancer comes second. Meanwhile, early detection is one of the primary ways to prevent lung and colon cancer fatalities. To aid the early detection of lung and colon cancer, we propose a computer-aided diagnostic approach that employs a Deep Learning (DL) architecture to enhance the detection of these cancer types from Computed Tomography (CT) images of suspected body parts. Our experimental dataset (LC25000) contains 25000 CT images of benign and malignant lung and colon cancer tissues. We used weights from a pre-trained DL architecture for computer vision, EfficientNet, to build and train a lung and colon cancer detection model. EfficientNet is a Convolutional Neural Network architecture that scales all input dimensions such as depth, width, and resolution at the same time. Our research findings showed detection accuracies of 99.63%, 99.50%, and 99.72% for training, validation, and test sets, respectively.

Bilgisayarlı Tomografi Görüntülerinden Derin Öğrenme ve Makine Öğrenmesi ile covid-19 Hastalığının Teşhisi

Abstract&#x0D; The new virus disease (COVID-19) first came to China towards the end of December 2019 and became a pandemic all over the world. The disease caused a large number of people to be infected and die. Rapid diagnosis of the disease is of great importance in controlling transmission. A computed Tomography device provides successful results in the diagnosis of COVID-19 disease. In this study, two-class (COVID-19 and normal) data sets were created from 7200 lung Computed Tomography images diagnosed between March 2020 and November 2020 in a private hospital with the help of specialist physicians. Verification and testing processes were carried out on Artificial Neural Network (ANN), Support Vector Machine (SVM), K-Nearest Neighbour (KNN) algorithms from Machine Learning algorithms, and ResNet-50, DenseNet-201, InceptionResNetV2, Inceptionv3, VGG-16, Xception architectures from Deep Learning models. As a result of the studies, the DenseNet-201 architecture obtained the highest result from deep learning models with %99,35 training and test %98,75 accuracy rates, respectively. ANN %97,6, KNN %97,4 and SVM %96,9 accuracy rates were obtained from machine learning.

Medical image segmentation method based on multi‐scale feature and U‐net network

AbstractIn medical image segmentation (MIS), better segmentation results can be obtained by training the deeper neural network. However, directly building too deep network will cause problems such as gradient disappearance, which will affect the segmentation effect. Therefore, a dilated inception U‐Net (DIU)‐net network is constructed by combining the multi‐scale feature fusion (MSFF) method and the concept of Inception in Google net based on U‐net, and its effectiveness is verified by experiments. The DIU‐net network's training accuracy has been improved in the lung computed tomography (CT) and fundus vascular CT image data sets. And the attenuation of the loss function is relatively stable, with the highest accuracy of 99.6%. In comparison of evaluation indicators, the values of different indicators of DIU‐net in the two data sets are higher than those of the comparison network. The DICE coefficient of DIU‐net in the lung CT image in the experiment is 0.986 on average, which is 0.2% higher than that of ResU‐net. SE value is 0.985, which is 1.9% higher than SegNet. Specificity value is slightly higher than the second segmentation effect. F1 score is 0.985, 0.6% higher than ResU‐net, area under curve value is 0.99, 0.7% higher than FCN‐8 s. In general, the DIU‐net network proposed in the study will not cause gradient disappearance and other problems in the experiment. At the same time, this method also shows high efficiency and has strong feasibility for the actual MIS.

Accurate Quantification of Small Pulmonary Nodules Using 3D Reconstruction of 2D Computed Tomography Lung Images

Accurate Quantification of Small Pulmonary Nodules Using 3D Reconstruction of 2D Computed Tomography Lung Images

An advanced perceptual U-Net segmentation based DDO-SDN classification system for lung pulmonary cancer detection

AbstractDeveloping an automated lung disease diagnosis framework is still remains one of the most challenging and demanding tasks in recent days. Most of the medical experts highly preferring the Computed Tomography (CT) lung images for an accurate disease detection. For this purpose, various segmentation, optimization, and classification techniques are developed in the conventional works for lung pulmonary disease detection. However, the existing techniques have the major problems of over segmentation, inaccurate ROI extraction, reduced accuracy, computational complexity, and high false positives. Thus, this research work intends to a simple and efficient segmentation based classification framework for an accurate lung nodules detection and pulmonary disease classification. Here, the tanh normalization technique is applied for preprocessing the input lung CT image with reduced noise and increased quality. After that, the perceptual U-Net segmentation algorithm is employed to accurately segment the lung nodules from the preprocessed CT images with simple computational operations. Moreover, the Decked Dragonfly Optimization (DDO) technique is used for choosing the relevant features based on the best optimal solution, which supports to obtain an increased detection accuracy and reduced classification error rate. Finally, the Speculative Deceptive Network (SDN) based classification algorithm is deployed to exactly detect the pulmonary lung cancer according to the optimal features. During evaluation, the performance of the proposed segmentation based DDO-SDN mechanism is validated and compared by using various evaluation parameters.

ALBAE feature extraction based lung pneumonia and cancer classification.

Lung cancer is a deadly disease showing uncontrolled proliferation of malignant cells in the lungs. If the lung cancer is detected in early stages, it can be cured before critical stage. In recent years, new technologies have gained much attention in the healthcare industry however, the unpredictable appearance of tumors, finding their presence, determining its shape, size and high discrepancy in medical images are the challenging tasks. To overcome this issue a novel Ant lion-based Autoencoders (ALbAE) model is proposed for efficient classification of lung cancer and pneumonia. Initially Computed Tomography (CT) images are pre-processed using median filters to remove noise artifacts and improving the quality of the images. Consequently, the relevant features such as image edges, pixel rates of the images and blood clots are extracted by ant lion-based autoencoder (ALbAE) technique. Finally, in classification stage, the lung CT images are classified into three different categories such as normal lung, cancer affected lung and pneumonia affected lung using Random forest technique. The effectiveness of the implemented design is estimated by different parameters such as precision, recall, Accuracy and F1-measure. The proposed approach attains 97% accuracy; 98% of recall and F-measure rate is attained through the developed design and the proposed model gains 96% of precision score. Experimental outcomes show that the proposed model performs better than existing SVM, ELM, and MLP in classifying lung cancer and pneumonia.

Deep learning techniques for prediction of pneumonia from lung CT images

Pneumonia disease is caused by viruses and bacteria which affect one or both lungs. It is the most dangerous disease that causes huge cancer death worldwide. Early detection of Pneumonia is the only way to improve a patient’s chance for survival. We can detect this disease from X-ray or computed tomography (CT) lung images using deep learning techniques. This research paper provides a solution to medical practitioners in predicting the impact of virus as high-risk, low-risk and medium-risk among the population being tested through various deep learning techniques such as convolutional neural networks, artificial neural network (ANN) and recurrent neural networks using long short-term memory cells. We observed 3000 CT images of Pneumonia confirmed patients and achieved the accuracy resulting 98–99%. The performance of the classifiers is evaluated using parameters such as confusion matrix, accuracy, F-measure, precision and recall. The results prove that deep learning affords a fitting tool for fast screening of Pneumonia and discovering high-risk patients and preventing them by providing suitable medical remedies.

Enhanced Elman spike Neural network optimized with flamingo search optimization algorithm espoused lung cancer classification from CT images

At present, researchers have been try to enhance the CAD system performance utilizing deep learning techniques in lung cancer screening and computed tomography (CT), but none of them attain the adequate accuracy. Therefore, an Enhanced Elman Spike Neural Network optimized with flamingo search optimization algorithm espoused Lung cancer classification from CT images (EESNN-FSOA-LCC) is proposed in this article. Initially, the input lung CT images are gathered through IQ-OTH/NCCD Lung Cancer Dataset. Then the input lung CT images are pre-processed using anisotropic diffusion Kuwahara filtering. These pre-processed lung CT images are fed to Hesitant Fuzzy Linguistic Bi-objective Clustering process for ROI region of lung cancer. Then the significant features present under ROI region of lung cancer segmentation are clarified through the help of Gray level co-occurrence matrix (GLCM) window adaptive approach. The extracted features are presented to EESNN for categorizing lung CT images image as normal, Benign, and Malignant. In general, EESNN classifier not divulges any adaption of optimization methods for determining the optimum parameters and to assure exact classification of lung cancer. Hence, a flamingo search OptimizationAlgorithmis proposed to optimize the EESNN classifier, which precisely classifies the lung cancer. The proposed EESNN-FSOA-LCC approach is activated in python. The proposed EESNN-FSOA-LCC approach achieves 38.58%, 25.69% and 43.87%, high accuracy when comparing to the existing CTI-LCC-SVM, CTI-LCC-GoogleNet-DNN and CTI-LCC-CNN respectively.

Quantitative CT Lung Imaging and Machine Learning Improves Prediction of Emergency Room Visits and Hospitalizations in COPD.

Predicting increased risk of future healthcare utilization in chronic obstructive pulmonary disease (COPD) patients is an important goal for improving patient management. Our objective was to determine the importance of computed tomography (CT) lung imaging measurements relative to other demographic and clinical measurements for predicting future health services use with machine learning in COPD. In this retrospective study, lung function measurements and chest CT images were acquired from Canadian Cohort of Obstructive Lung Disease study participants from 2010 to 2017 (https://clinicaltrials.gov, NCT00920348). Up to two follow-up visits (1.5- and 3-year follow-up) were performed and participants were asked for details related to healthcare utilization. Healthcare utilization was defined as any COPD hospitalization or emergency room visit due to respiratory problems in the 12 months prior to the follow-up visits. CT analysis was performed (VIDA Diagnostics Inc.); a total of 108 CT quantitative emphysema, airway and vascular measurements were investigated. A hybrid feature selection method with support vector machine classifier was used to predict healthcare utilization. Performance was determined using accuracy, F1-measure and area under the receiver operating characteristic curve (AUC) and Matthews's correlation coefficient (MC). Of the 527 COPD participants evaluated, 179 (35%) used healthcare services at follow-up. There were no significant differences between the participants with or without healthcare utilization at follow-up for age (p=0.50), sex (p=0.44), BMI (p=0.05) or pack-years (p=0.76). The accuracy for predicting subsequent healthcare utilization was 80% ± 3% (F1-measure=74%, AUC=0.80, MC=0.6) when all measurements were considered, 76% ± 6% (F1-measure=72%, AUC=0.77, MC=0.55) for CT measurements alone and 65% ± 5% (F1-measure=60%, AUC=0.67, MC=0.34) for demographic and lung function measurements alone. The combination of CT lung imaging and conventional measurements leads to greater prediction accuracy of subsequent health services use than conventional measurements alone, and may provide needed prognostic information for patients suffering from COPD.

DACov: a deeper analysis of data augmentation on the computed tomography segmentation problem

ABSTRACT Due to the COVID-19 global pandemic, computer-assisted diagnoses of medical images have gained much attention, and robust methods of semantic segmentation of Computed Tomography (CT) images have become highly desirable. In this work, we present a deeper analysis of how data augmentation techniques improve segmentation performance on this problem. We evaluate traditional augmentation techniques on five public datasets. Six different probabilities of applying each augmentation technique on an image were evaluated. We also assess a different training methodology where the training subsets are combined into a single larger set. All networks were evaluated through a -fold cross-validation strategy, resulting in over experiments. We also propose a novel data augmentation technique based on Generative Adversarial Networks (GANs) to create new healthy and unhealthy lung CT images, evaluating four variations of our approach with the same six probabilities of the traditional methods. Our findings show that GAN-based techniques and spatial-level transformations are the most promising for improving the learning of deep models on this problem, with the StarGAN v2 + F with a probability of achieving the highest F-score value on the Ricord1a dataset in the unified training strategy. Our code is publicly available at https://github.com/VRI-UFPR/DACov2022.