Lung Computed Tomography Images Research Articles

Asbestosis diagnosis algorithm combining the lung segmentation method and deep learning model in computed tomography image

BackgroundEarly detection of asbestosis is important; hence, quick and accurate diagnostic tools are essential. This study aimed to develop an algorithm that combines lung segmentation and deep learning models that can be utilized as a clinical decision support system (CDSS) for diagnosing patients with asbestosis in segmented computed tomography (CT) images. MethodsWe accurately segmented the lungs in CT images of patients examined at Seoul St. Mary’s Hospital using a threshold-based method. Lungs with asbestosis and normal lungs were classified by applying the segmented image to the long-term recurrent convolutional network deep learning model. Performance was evaluated using the area under the receiver operating characteristic curve (AUROC) and F1 score from the test data. ResultsThe algorithm developed using the DenseNet201pre-trained model showed excellent performance, with a sensitivity of 0.962, specificity of 0.975, accuracy of 0.970, AUROC of 0.968, and F1 score of 0.961. ConclusionsWe developed an algorithm with significantly better diagnostic accuracy than a radiologist (0.970 vs. 0.73–0.79). Our developed algorithm is expected to be an excellent support tool if used as a CDSS to diagnose asbestosis using CT images.

VGG19 Network Assisted Joint Segmentation and Classification of Lung Nodules in CT Images.

Pulmonary nodule is one of the lung diseases and its early diagnosis and treatment are essential to cure the patient. This paper introduces a deep learning framework to support the automated detection of lung nodules in computed tomography (CT) images. The proposed framework employs VGG-SegNet supported nodule mining and pre-trained DL-based classification to support automated lung nodule detection. The classification of lung CT images is implemented using the attained deep features, and then these features are serially concatenated with the handcrafted features, such as the Grey Level Co-Occurrence Matrix (GLCM), Local-Binary-Pattern (LBP) and Pyramid Histogram of Oriented Gradients (PHOG) to enhance the disease detection accuracy. The images used for experiments are collected from the LIDC-IDRI and Lung-PET-CT-Dx datasets. The experimental results attained show that the VGG19 architecture with concatenated deep and handcrafted features can achieve an accuracy of 97.83% with the SVM-RBF classifier.

Diagnostic Value of Deep Learning-Based CT Feature for Severe Pulmonary Infection.

The study aimed to explore the diagnostic value of computed tomography (CT) images based on cavity convolution U-Net algorithm for patients with severe pulmonary infection. A new lung CT image segmentation algorithm (U-Net+ deep convolution (DC)) was proposed based on U-Net network and compared with convolutional neural network (CNN) algorithm. Then, it was applied to CT image diagnosis of 100 patients with severe lung infection in The Second Affiliated Hospital of Fujian Medical University hospital and compared with traditional methods, and its sensitivity, specificity, and accuracy were compared. It was found that the single training time and loss of U-Net + DC algorithm were reduced by 59.4% and 9.8%, respectively, compared with CNN algorithm, while Dice increased by 3.6%. The lung contour segmented by the proposed model was smooth, which was the closest to the gold standard. Fungal infection, bacterial infection, viral infection, tuberculosis infection, and mixed infection accounted for 28%, 18%, 7%, 7%, and 40%, respectively. 36%, 38%, 26%, 17%, and 20% of the patients had ground-glass shadow, solid shadow, nodule or mass shadow, reticular or linear shadow, and hollow shadow in CT, respectively. The incidence of various CT characteristics in patients with fungal and bacterial infections was statistically significant (P < 0.05). The specificity (94.32%) and accuracy (97.22%) of CT image diagnosis based on U-Net + DC algorithm were significantly higher than traditional diagnostic method (75.74% and 74.23%), and the differences were statistically significant (P < 0.05). The network of the algorithm in this study demonstrated excellent image segmentation effect. The CT image based on the U-Net + DC algorithm can be used for the diagnosis of patients with severe pulmonary infection, with high diagnostic value.

Prognostic Implication of Direct Cardiac Invasion from Lung Cancer in Non-Operatively Treated Patients Based on Lung Computed Tomography Imaging

Lung cancer with direct cardiac invasion (LCCI+) exerts a significant influence on the survival of patients. There is a paucity of comparative research into the prognosis of advanced lung cancer with and without direct cardiac invasion. In this study, 50 LCCI+ patients and 50 sex-, age-, and TNM stage-matched patients without direct cardiac invasion (LCCI-) were retrospectively analysed. LCCI+ was defined as lung cancer directly invading the heart by penetrating mediastinum or extending into the atrium via the pulmonary vein. The study endpoint was all-cause death. In this study, the survival time was defined as the time from the first detection of direct cardiac invasion to the end of the event. During a median follow-up period of 31 months, all-cause death occurred in 44 patients (88.0%) in the LCCI+ group and in 36 patients (72.0%) in the LCCI- group; the overall survival (OS) time among patients in the LCCI+ group was significantly lower compared with those in the LCCI- group (5.0 [interquartile range (IQR), 2.0-12.0] vs 13.8 [IQR, 4.0-18.4] mo; p<0.001); the OS rate in the LCCI+ group was significantly lower compared with patients in the LCCI- group (log-rank, p=0.0002). Multivariate Cox regression analysis showed that direct cardiac invasion was an independent predictor of survival in patients with advanced lung cancer (hazard ratio, 2.255; 95% confidence interval, 1.443-3.524). Further analysis indicated that in patients with small cell lung cancer, the survival rate between the LCCI+ group and LCCI- group was insignificant (log-rank, p=0.075; survival time: 4.0 [IQR, 2.0-11.5] vs 11.5 [IQR, 5.0-18.3] mo); in patients with non-small cell lung cancer (NSCLC), the survival rate in the LCCI+ group was lower than that of the LCCI- group (log-rank, p=0.01; survival time: 6.0 [IQR, 3.0-13.3] vs 16.3 [IQR, 10.4-27.2] mo). Direct cardiac invasion from lung cancer was an independent prognostic factor for survival time in patients with lung cancer. Patients with direct cardiac invasion by NSCLC have a poorer clinical outcome than those without direct cardiac invasion. A careful preoperative evaluation is mandatory and appropriate management of cardiac involvement should be considered in the treatment of NSCLC.

Phase congruency-based filtering approach combined with a convolutional network for lung CT image analysis

ABSTRACT Computed tomography (CT) imaging was confirmed as one of the important evolving radiological features of disease diagnosis. In this work, we propose an image filtering method for CT images that fuses the spatial and the transform domain. Filters in the transform domain work well in restoring low-contrast details that are usually smoothed by spatial domain filters. The new filter is based on a modified anisotropic diffusion approach combined with the phase congruency (PC) feature. The PC feature is a scale-invariant compared to the classical gradient, where weak edges are usually omitted and undetected with gradient-based feature detectors. This feature is incorporated in the diffusion function to enhance image edges while eliminating noise and texture background. In a further step, a U-net convolutional network is used to segment the lung-infected area. Comparative tests and results show the efficiency of our method, which permits to enhance image features while removing noise perfectly.

Predicting lung adenocarcinoma invasiveness by measurement of pure ground-glass nodule roundness by using multiplanar reformation: a retrospective analysis

To explore the value of roundness measurement based on thin-section axial, coronal, and sagittal section computed tomography (CT) images for predicting pure ground-glass nodule (pGGN) invasiveness. A total of 168 pGGNs in 155 patients (44 male, 111 females; mean age, 55.74±10.57 years), and confirmed by surgery and histopathology, were analysed retrospectively and divided into pre-invasive (n=72) and invasive (n=96) groups. Photoshop (CS6) software was used to measure pGGN roundness based on conventional axial section, as well as coronal and sagittal sections generated by multiplanar reformation, from thin-section (1-mm-thick) CT lung images. pGGN roundness values, measured in axial, coronal, and sagittal thin-section CT sections from the pre-invasive group were 0.8±0.049, 0.816±0.05, and 0.818±0.043, respectively, while those in the invasive group were 0.745±0.077, 0.684±0.106, and 0.678±0.106; differences between the two groups were significant (all p<0.001). Binary logistic regression analysis showed that roundness values based on coronal and sagittal sections (p<0.001) were better than those from axial sections (p>0.05) in predicting pGGN invasiveness, with odds ratio (OR) values of 14.858 and 23.315, respectively. ROC analysis showed that evaluation of roundness measured in sagittal sections was better at predicting pGGN invasiveness than when coronal sections were used (AUC 0.870 versus 0.832). Roundness is useful for predicting pGGN invasiveness, with measurements from coronal and sagittal sections better than those from conventional axial sections, with sagittal section images having the best predictive value.

Optimized Backprojection Filtration Algorithm for Postoperative Reduction and Analysis of Respiratory Infection-Related Factors of Pelvic Fractures by CT Imaging

To explore the computed tomography (CT) imaging characteristics and BPF algorithm fine lung CT image efficiency for the diagnosis of pelvic fracture patients and assist clinicians to carry out the disease care and treatment, CT images based on optimized back-projection filtering (BPF) algorithm were utilized to diagnose postoperative reduction of pelvic fractures and penetrating lung infection caused by long-term bed rest. A total of 100 patients with pelvic fracture were selected and all of them underwent pelvic fracture surgery and were rolled into conventional CT diagnosis group (conventional group) and BPF algorithm optimized CT image diagnosis group (BPF group). One group used conventional CT images to guide pelvic reduction and detect lung infections, and the other used BPF algorithm to optimize the images. The results showed that the BPF group was superior to the conventional CT group in both image clarity and shadow area, and the peak signal-to-noise ratio (PSNR) was significantly better than that of the conventional group ( P &lt; 0.05 ). Nine more cases were detected in the algorithm group than in the conventional group, and the incidence of complications was 48% in the conventional group and 28% in the BPF group, with a statistical difference of 20% between the two groups ( P &lt; 0.05 ). In addition, the satisfaction of returning patients was 96% in the BPF group and 77% in the conventional group ( P &lt; 0.05 ). The diagnosis of pulmonary infection was more obvious in the BPF group, indicating that BPF optimization of the CT image was suitable for clinical diagnosis and had a practical application value.

Automated vessel segmentation in lung CT and CTA images via deep neural networks

The distribution of pulmonary vessels in computed tomography (CT) and computed tomography angiography (CTA) images of lung is important for diagnosing disease, formulating surgical plans and pulmonary research. Based on the pulmonary vascular segmentation task of International Symposium on Image Computing and Digital Medicine 2020 challenge, this paper reviews 12 different pulmonary vascular segmentation algorithms of lung CT and CTA images and then objectively evaluates and compares their performances. First, we present the annotated reference dataset of lung CT and CTA images. A subset of the dataset consisting 7,307 slices for training and 3,888 slices for testing was made available for participants. Second, by analyzing the performance comparison of different convolutional neural networks from 12 different institutions for pulmonary vascular segmentation, the reasons for some defects and improvements are summarized. The models are mainly based on U-Net, Attention, GAN, and multi-scale fusion network. The performance is measured in terms of Dice coefficient, over segmentation rate and under segmentation rate. Finally, we discuss several proposed methods to improve the pulmonary vessel segmentation results using deep neural networks. By comparing with the annotated ground truth from both lung CT and CTA images, most of 12 deep neural network algorithms do an admirable job in pulmonary vascular extraction and segmentation with the dice coefficients ranging from 0.70 to 0.85. The dice coefficients for the top three algorithms are about 0.80. Study results show that integrating methods that consider spatial information, fuse multi-scale feature map, or have an excellent post-processing to deep neural network training and optimization process are significant for further improving the accuracy of pulmonary vascular segmentation.

Computed Tomography Image under Convolutional Neural Network Deep Learning Algorithm in Pulmonary Nodule Detection and Lung Function Examination.

The objective of this study was to perform segmentation and extraction of CT images of pulmonary nodules based on convolutional neural networks (CNNs). The Mask-RCNN algorithm model is a typical end-to-end image segmentation model, which uses the R-FCN structure for nodule detection. The effect of applying the two algorithm models to the computed tomography (CT) diagnosis of pulmonary nodules was analyzed, and different indexes of pulmonary nodule CT images in lung function examination after algorithm optimization were compared. A total of 56 patients diagnosed with pulmonary nodules by surgery or puncture were taken as the research objects. Based on the Mask-RCNN algorithm, a model for CT image segmentation processing of pulmonary nodules was proposed. Subsequently, the 3D Faster-RCNN model was used to label the nodules in the pulmonary nodules. The experimental results showed that the trained Mask-RCNN algorithm model can effectively complete the segmentation task of lung CT images, but there was a little jitter at the boundary. The speed of R-FCN algorithm for nodular detection was 0.172 seconds/picture, and the accuracy was 88.9%. CT scans were performed on the 56 patients based on a deep learning algorithm. The results showed that 30 cases of malignant pulmonary nodules were confirmed, and the diagnostic accuracy was 93.75%. There were 22 benign lesions, the diagnostic accuracy was 91.67%, and the overall diagnostic accuracy was 92.85%. This study effectively improved the diagnostic efficiency of CT images of pulmonary nodules, and the accuracy of CT images in the diagnosis of pulmonary nodules was analyzed and evaluated. It provided theoretical support for the follow-up diagnosis of pulmonary nodules and the treatment of lung cancer. It also significantly improved the diagnostic effect and detection efficiency of pulmonary nodules.

COVID-19 pneumonia: pathophysiology and management.

Coronavirus disease 2019 (COVID-19) pneumonia is an evolving disease. We will focus on the development of its pathophysiologic characteristics over time, and how these time-related changes determine modifications in treatment. In the emergency department: the peculiar characteristic is the coexistence, in a significant fraction of patients, of severe hypoxaemia, near-normal lung computed tomography imaging, lung gas volume and respiratory mechanics. Despite high respiratory drive, dyspnoea and respiratory rate are often normal. The underlying mechanism is primarily altered lung perfusion. The anatomical prerequisites for PEEP (positive end-expiratory pressure) to work (lung oedema, atelectasis, and therefore recruitability) are lacking. In the high-dependency unit: the disease starts to worsen either because of its natural evolution or additional patient self-inflicted lung injury (P-SILI). Oedema and atelectasis may develop, increasing recruitability. Noninvasive supports are indicated if they result in a reversal of hypoxaemia and a decreased inspiratory effort. Otherwise, mechanical ventilation should be considered to avert P-SILI. In the intensive care unit: the primary characteristic of the advance of unresolved COVID-19 disease is a progressive shift from oedema or atelectasis to less reversible structural lung alterations to lung fibrosis. These later characteristics are associated with notable impairment of respiratory mechanics, increased arterial carbon dioxide tension (PaCO2), decreased recruitability and lack of response to PEEP and prone positioning.

Low-Dose Lung Radiation Therapy for COVID-19 Lung Disease: A Preclinical Efficacy Study in a Bleomycin Model of Pneumonitis

PurposeLow-dose whole lung radiation therapy (LDLR) has been proposed as a treatment for patients with acute respiratory distress syndrome associated with SARS-CoV-2 infection, and clinical trials are underway. There is an urgent need for preclinical evidence to justify this approach and inform dose, scheduling, and mechanisms of action.Methods and MaterialsFemale C57BL/6 mice were treated with intranasal bleomycin sulfate (7.5 or 11.25 units/kg, day 0) and then exposed to whole lung radiation therapy (0.5, 1.0, or 1.5 Gy, or sham; day 3). Bodyweight was measured daily, and lung tissue was harvested for histology and flow cytometry on day 10. Computed tomography lung imaging was performed before radiation (day 3) and pre-endpoint (day 10).ResultsBleomycin caused pneumonitis of variable severity, which correlated with weight loss. LDLR at 1.0 Gy was associated with a significant increase in the proportion of mice recovering to 98% of initial bodyweight, and a proportion of these mice exhibited less severe histopathologic lung changes. Mice experiencing moderate initial weight loss were more likely to respond to LDLR than those experiencing severe initial weight loss. In addition, LDLR (1.0 Gy) significantly reduced bleomycin-induced increases in interstitial macrophages, CD103+ dendritic cells (DCs), and neutrophil-DC hybrids. Overall, bleomycin-treated mice exhibited significantly higher percentages of nonaerated lung in left than right lungs, and LDLR (1.0 Gy) limited further reductions in aerated lung volume in right but not left lungs. LDLR at 0.5 and 1.5 Gy did not improve bodyweight, flow cytometric, or radiologic readouts of bleomycin-induced pneumonitis.ConclusionsOur data support the concept that LDLR can ameliorate acute inflammatory lung injury, identify 1.0 Gy as the most effective dose, and provide evidence that it is more effective in the context of moderate than severe pneumonitis. Mechanistically, LDLR at 1.0 Gy significantly suppressed bleomycin-induced accumulation of pulmonary interstitial macrophages, CD103+ DCs, and neutrophil-DC hybrids.

Preliminary Report on Computed Tomography Radiomics Features as Biomarkers to Immunotherapy Selection in Lung Adenocarcinoma Patients.

Simple SummaryThe objective of the study was to assess the radiomics features obtained by computed tomography (CT) examination as biomarkers in order to select patients with lung adenocarcinoma who would benefit from immunotherapy. We demonstrated that specific radiomic features could be used to select patients with lung adenocarcinoma who would benefit from immunotherapy by predicting OS or PFS time.Purpose: To assess the efficacy of radiomics features obtained by computed tomography (CT) examination as biomarkers in order to select patients with lung adenocarcinoma who would benefit from immunotherapy. Methods: Seventy-four patients (median age 63 years, range 42–86 years) with histologically confirmed lung cancer who underwent immunotherapy as first- or second-line therapy and who had baseline CT studies were enrolled in this approved retrospective study. As a control group, we selected 50 patients (median age 66 years, range 36–86 years) from 2005 to 2013 with histologically confirmed lung adenocarcinoma who underwent chemotherapy alone or in combination with targeted therapy. A total of 573 radiomic metrics were extracted: 14 features based on Hounsfield unit values specific for lung CT images; 66 first-order profile features based on intensity values; 43 second-order profile features based on lesion shape; 393 third-order profile features; and 57 features with higher-order profiles. Univariate and multivariate statistical analysis with pattern recognition approaches and the least absolute shrinkage and selection operator (LASSO) method were used to assess the capability of extracted radiomics features to predict overall survival (OS) and progression free survival (PFS) time. Results: A total of 38 patients (median age 61; range 41–78 years) with confirmed lung adenocarcinoma and subjected to immunotherapy satisfied inclusion criteria, and 50 patients in a control group were included in the analysis The shift in the center of mass of the lesion due to image intensity was significant both to predict OS in patients subjected to immunotherapy and to predict PFS in patients subjected to immunotherapy and in patients in the control group. With univariate analysis, low diagnostic accuracy was reached to stratify patients based on OS and PFS time. Regarding multivariate analysis, considering the robust (two morphological features, three textural features and three higher-order statistical metrics) application of the LASSO approach and all patients, a support vector machine reached the best results for stratifying patients based on OS (area under curve (AUC) of 0.89 and accuracy of 81.6%). Alternatively, considering the robust predictors (six textural features and one higher-order statistical metric) and application of the LASSO approach including all patients, a decision tree reached the best results for stratifying patients based on PFS time (AUC of 0.96 and accuracy of 94.7%). Conclusions: Specific radiomic features could be used to select patients with lung adenocarcinoma who would benefit from immunotherapy because a subset of imaging radiomic features useful to predict OS or PFS time were different between the control group and the immunotherapy group.

3D multi‐resolution deep learning model for diagnosis of multiple pathological types on pulmonary nodules

AbstractTo accurately diagnose multiple pathological types of pulmonary nodules based on lung computed tomography (CT) images, a multi‐resolution three‐dimensional (3D) multi‐classification deep learning model (Mr‐Mc) was proposed. The Mr‐Mc model was constructed by using our own constructed lung CT image dataset of pulmonary nodules with clinical pathological information (LCID‐CPI), which can accurately diagnose inflammation, squamous cell carcinoma, adenocarcinoma, and other benign diseases. In order to process nodules with different sizes, a multi‐resolution extraction method was proposed to extract 3D volume data with different resolutions from lung CT images. The Mr‐Mc was composed of three different resolution networks, each of which has input volume data of a specific resolution. Experiments showed that the constructed Mr‐Mc model can achieve an average accuracy of 0.81 on LCID‐CPI. Besides, the Mr‐Mc model can also achieve a high accuracy of 0.87 on the Lung Image Database Consortium and Image Database Resource Initiative dataset.

Texture Analysis of Computed Tomography Images in the Lung of Patients With Breast Cancer.

The aim of this study was to investigate whether the texture features of lung computed tomography images were altered by primary breast cancer without pulmonary metastasis. Texture analysis was performed on the regions of interest of lung computed tomography images from 36 patients with breast cancer and 36 healthy controls. Texture parameters between subjects with different clinical stages and hormone receptor (HR) statuses in patients with breast cancer were analyzed. Three texture parameters (mean, SD, and variance) were significantly different between patients with breast cancer and healthy controls and between preoperative and postoperative stages in patients with breast cancer. All 3 parameters showed an increasing trend under the tumor-bearing state. These parameters were significantly higher in the stage III + IV group than in the stage I + II group. The variance parameter was significantly higher in the HR-negative group than in the HR-positive group. Texture analysis may serve as a novel additional tool for discovering conventionally invisible changes in the lung tissue of patients with breast cancer.

Intelligent predictions of Covid disease based on lung CT images using machine learning strategy.

Covid or Corona Virus, a term ruling the world from past two years and causes a huge destruction in all countries. One of the most important Covid disease identification method is Lung based Computed Tomography (CT) image scanning, in which it provides an effective disease identification means in clear manner. However, this Lung CT image based disease detection principles are complex to health care representatives and doctors to predict the Covid disease accurately. Several manual errors and medical flaws are raised day-by-day, so that a new systematic methodology is required to identify the Covid disease effectively with respect to machine learning principles. The machine learning principles are most popular to identify the respective disease efficiently as well as classify the disease in accurate manner without any time consumption. The infected portions of the chest are identified accurately and report to the respective person without any delay. In this paper, a new machine learning strategy is introduced called Hybrid Disease Detection Principle (HDDP), in which it is derived from the two classical machine learning algorithms called Convolutional Neural Network (CNN) and the AdaBoost Classifier. Both these algorithms are integrated together to produce a new strategy called HDDP, in which it process the lung CT image based on the machine learning factors such as pre-processing, feature extraction and classification. Based on these effective image processing strategies the proposed algorithm handles the CT images to predict the Covid disease and report to the respective user with proper accuracy ratio. This paper intends to provide effcient disease predictions as well as provide a sufficient support to medical people and patients in fine manner to assist them with modern classification algorithms.

An effective deep network for automatic segmentation of complex lung tumors in CT images.

Accurate segmentation of complex tumors in lung computed tomography (CT) images is essential to improve the effectiveness and safety of lung cancer treatment. However, the characteristics of heterogeneity, blurred boundaries, and large-area adhesion to tissues with similar gray-scale features always make the segmentation of complex tumors difficult. This study proposes an effective deep network for the automatic segmentation of complex lung tumors (CLT-Net). The network architecture uses an encoder-decoder model that combines long and short skip connections and a global attention unit to identify target regions using multiscale semantic information. A boundary-aware loss function integrating Tversky loss and boundary loss based on the level-set calculation is designed to improve the network's ability to perceive boundary positions of difficult-to-segment (DTS) tumors. We use a dynamic weighting strategy to balance the contributions of the two parts of the loss function. The proposed method was verified on a dataset consisting of 502lung CT images containing DTS tumors. The experiments show that the Dice similarity coefficient and Hausdorff distance metric of the proposed method are improved by 13.2% and 8.5% on average, respectively, compared with state-of-the-art segmentation models. Furthermore, we selected three additional medical image datasets with different modalities to evaluate the proposed model. Compared with mainstream architectures, the Dice similarity coefficient is also improved to a certain extent, which demonstrates the effectiveness of our method for segmenting medical images. Quantitative and qualitative results show that our method outperforms current mainstream lung tumor segmentation networks in terms of Dice similarity coefficient and Hausdorff distance. Note that the proposed method is not limited to the segmentation of complex lung tumors but also performs in different modalities of medical image segmentation.

Analysis of Image Features and TCM Syndrome Types of Lobar Pneumonia in Children Based on Mean Square Deviation Lung CT Image Registration Algorithm

The aim was to analyze the application of computed tomography (CT) images in the diagnosis of lesions, and the composition of traditional Chinese medicine (TCM) syndromes in children with lobar pneumonia. Lung CT image registration algorithm was constructed based on optimized mean square deviation (OMSD) algorithm, which was applied to CT images of 188 children patients with lobar pneumonia before and after treatment. Besides, free-form deformation (FFD) algorithm and mean square deviation (MSD) algorithm were introduced for comparison with OMSD. Results showed that sum of squared differences (SSD) of OMSD was significantly lower than that of MSD and FFD ( P &lt; 0.05 ). The mutual information (MI), relative overlap rate (ROR), and rcc of OMSD were markedly higher than those of MSD and FFD ( P &lt; 0.05 ). After treatment, the number of pulmonary interstitial thickening, lobular interstitial thickening, ground-glass shadow, patchy shadow, consolidation shadow, pleural thickening, pleural effusion, lymphadenopathy, pneumothorax, and mediastinal emphysema decreased sharply in contrast to before treatment ( P &lt; 0.05 ). Among the selected children, there were 184 children patients with empirical TCM syndrome, accounting for 97.85%, and its dominant syndrome was wind-heat closed lung syndrome (52.75%) and phlegm-heat closed lung syndrome (46.11%). The main symptoms of wind-heat closed and phlegm-heat closed lung syndrome were fever, cough, and pulmonary rales. In conclusion, OMSD was superior to MSD and FFD in lung CT image registration. CT registration image based on OMSD could clearly display the intrapulmonary and extrapulmonary manifestations of children patients, so as to enhance the clinical diagnosis effect. Besides, wind-heat closed lung syndrome and phlegm-heat closed lung syndrome were the common types of TCM syndrome of children pneumonia with the common symptoms of fever and cough.

Computed Tomography Image Segmentation Using Edge Correction Algorithm for Refractory Mycoplasma Pneumonia in Children

Objective. It is to study the application of edge correction algorithm (ECA) in computed tomography (CT) medical image segmentation, explore its guiding significance in the analysis of clinical characteristics of children with refractory mycoplasma pneumoniae (RMPP), and discuss the therapeutic value of fiberoptic bronchoscopy bronchoalveolar lavage (BAL) for RMPP. Methods. The accuracy of ECA in CT medical image segmentation of children with RMPP was compared with that of the watershed segmentation algorithm (WSA) and swarm intelligence optimization algorithm (SIOA). The clinical characteristics and the imaging characteristics of 80 children with RMPP admitted to hospital from January 2018 to January 2020 were retrospectively analyzed based on the ECA. All children were divided into a lavage group (BAL group, n = 69) and a nonlavage group (non-BAL group, n = 11) according to whether fiberoptic bronchoscopy and BAL were performed. Bronchoscopy was adopted to analyze the cytological characteristics of BAL fluid (BALF) in children, and the recovery rate and the total effective rate of the two groups of children were observed and compared. Results. The overall accuracies (OAs) of the three ECAs (Roberts operator (RO), Sobel operator (SO), and Prewitt operator (PO)) were higher than that of WSA and SIOA, their false negative rate (FNR) and false positive rate (FPR) were small, and their denoising performance was superior to that of WSA and SIOA. The main clinical manifestations of all children were high fever, irritating dry cough, and few early signs. The results of chest CT examination were mainly manifested as patchy or large-scale consolidation, two lung mesh or small nodular shadows, and atelectasis. 69 cases with fiberoptic bronchoscopy showed swelling and congestion of the bronchial mucosa at the lesion site with visible viscous secretions, which was consistent with the imaging changes. The total number of cells in the BALF of children increased ( P &lt; 0.05 ), which mainly represented the increase of neutrophils ( P &lt; 0.05 ). The recovery rate of children with lavage (81.16%) was higher dramatically than that of the nonlavage group (45.45%). Conclusion. The ECA had good accuracy and denoising performance in lung CT image segmentation. The clinical characteristics, imaging characteristics, and cytological components of children had changed when they suffered from the RMPP, and fiberoptic bronchoscopy lavage had a therapeutic effect on it.

Extension of Collagen Deposition in COVID-19 Post Mortem Lung Samples and Computed Tomography Analysis Findings.

Lung fibrosis has specific computed tomography (CT) findings and represents a common finding in advanced COVID-19 pneumonia whose reversibility has been poorly investigated. The aim of this study was to quantify the extension of collagen deposition and aeration in postmortem cryobiopsies of critically ill COVID-19 patients and to describe the correlations with qualitative and quantitative analyses of lung CT. Postmortem transbronchial cryobiopsy samples were obtained, formalin fixed, paraffin embedded and stained with Sirius red to quantify collagen deposition, defining fibrotic samples as those with collagen deposition above 10%. Lung CT images were analyzed qualitatively with a radiographic score and quantitatively with computer-based analysis at the lobe level. Thirty samples from 10 patients with COVID-19 pneumonia deceased during invasive mechanical ventilation were included in this study. The median [interquartile range] percent collagen extension was 6.8% (4.6–16.2%). In fibrotic compared to nonfibrotic samples, the qualitative score was higher (260 (250–290) vs. 190 (120–270), p = 0.036) while the gas fraction was lower (0.46 (0.32–0.47) vs. 0.59 (0.37–0.68), p = 0.047). A radiographic score above 230 had 100% sensitivity (95% confidence interval, CI: 66.4% to 100%) and 66.7% specificity (95% CI: 41.0% to 92.3%) to detect fibrotic samples, while a gas fraction below 0.57 had 100% sensitivity (95% CI: 66.4% to 100%) and 57.1% specificity (95% CI: 26.3% to 88.0%). In COVID-19 pneumonia, qualitative and quantitative analyses of lung CT images have high sensitivity but moderate to low specificity to detect histopathological fibrosis. Pseudofibrotic CT findings do not always correspond to increased collagen deposition.

Detection of lung tumor using dual tree complex wavelet transform and co‐active adaptive neuro fuzzy inference system classification approach

AbstractThe automatic detection and location of the tumor regions in lung images is more important to provide timely medical treatments to patients in order to save their lives. In this article, machine learning‐based lung tumor detection, classification and segmentation algorithm is proposed. The tumor classification phase first smooth the source lung computed tomography image using adaptive median filter and then discrete time complex wavelet transform (DT‐CWT) is applied on this smoothed lung image to decompose the entire image into a number of sub‐bands. Along with the decomposed sub‐bands, DWT, pattern, and co‐occurrence features are computed and classified using co‐active adaptive neuro fuzzy inference system (CANFIS). The tumor segmentation phase uses morphological functions on this classified abnormal lung image to locate the tumor regions. The multi‐evaluation parameters are used to evaluate the proposed method. This method is compared with the other state‐of‐the‐art methods on the same lung image from open‐access dataset.