Lung Segmentation Research Articles

Can Deep Adult Lung Segmentation Models Generalize to the Pediatric Population?

Lung segmentation in chest X-rays (CXRs) is an important prerequisite for improving the specificity of diagnoses of cardiopulmonary diseases in a clinical decision support system. Current deep learning models for lung segmentation are trained and evaluated on CXR datasets in which the radiographic projections are captured predominantly from the adult population. However, the shape of the lungs is reported to be significantly different across the developmental stages from infancy to adulthood. This might result in age-related data domain shifts that would adversely impact lung segmentation performance when the models trained on the adult population are deployed for pediatric lung segmentation. In this work, our goal is to (i) analyze the generalizability of deep adult lung segmentation models to the pediatric population and (ii) improve performance through a stage-wise, systematic approach consisting of CXR modality-specific weight initializations, stacked ensembles, and an ensemble of stacked ensembles. To evaluate segmentation performance and generalizability, novel evaluation metrics consisting of mean lung contour distance (MLCD) and average hash score (AHS) are proposed in addition to the multi-scale structural similarity index measure (MS-SSIM), the intersection of union (IoU), Dice score, 95% Hausdorff distance (HD95), and average symmetric surface distance (ASSD). Our results showed a significant improvement (p < 0.05) in cross-domain generalization through our approach. This study could serve as a paradigm to analyze the cross-domain generalizability of deep segmentation models for other medical imaging modalities and applications.

VGG-based discrimination for pneumonia images enhanced by Generative Adversarial Network

Due to problems such as blurring of chest X-ray image and double shadow of heart and lung, the accuracy of judging whether there is COVID-19 is not high. This study intends to use Generative Adversarial Network (GAN) to segment images, and then put the segmented image set into the classification algorithm to determine whether it is infected by the pneumonia. The data set is divided into “normal” and “pneumonia”. In order to remove the interference of rib and heart on lung contour in X-ray image, GAN technology is used to segment the original data set. The segmented image only retains the most important lung silhouette, which is convenient for the next step of recognition and information extraction. The classification algorithm uses Visual Geometry Group (VGG)-16. During the training, only fully connected layer 1 and fully connected layer 2 layers were trained, and the rest layers were frozen. When the model loads the original weight, the top level of the model will be replaced by the head model. The head model is placed on the basic model and becomes part of the actual training model to determine the best weight. After the original X-ray image is processed by GAN, the accuracy of the segmented lung image is improved from 84% to 98%. Loss decreased from 17% to 3%, and the accuracy of classification results was greatly improved. Therefore, it is proved that the segmented image can effectively remove irrelevant parts in the recognition process, and the accuracy is improved due to the removal of noise interference and clearer lung contour.

A Method of Lung Organ Segmentation in CT Images Based on Multiple Residual Structures and an Enhanced Spatial Attention Mechanism

Accurate organ segmentation is a fundamental step in disease-assisting diagnostic systems, and the precise segmentation of lung is crucial for subsequent lesion detection. Prior to this, lung segmentation algorithms had typically segmented the entire lung tissue. However, the trachea is also essential for diagnosing lung diseases. Challenges in lung parenchyma segmentation include the limited robustness of U-Net in acquiring contextual information and the small size of the trachea being mixed up with lung, making it difficult to identify and reconstruct the lungs. To overcome these difficulties, this paper proposes three improvements to U-Net: multiple concatenation modules to enhance the network’s ability to capture context, multi-scale residual learning modules to improve the model’s multi-scale learning capabilities, and an enhanced gated attention mechanism to enhance the fusion of various hierarchical features. The experimental results demonstrate that our model has achieved a significant improvement in trachea segmentation compared to existing models.

Non-local attention and multi-task learning based lung segmentation in chest X-ray

Precise segmentation of lung field is a crucial step in chest radiographic computer-aided diagnosis system. With the development of deep learning, fully convolutional network based models for lung field segmentation have achieved great effect but are poor at accurate identification of the boundary and preserving lung field consistency. To solve this problem, this paper proposed a lung segmentation algorithm based on non-local attention and multi-task learning. Firstly, an encoder-decoder convolutional network based on residual connection was used to extract multi-scale context and predict the boundary of lung. Secondly, a non-local attention mechanism to capture the long-range dependencies between pixels in the boundary regions and global context was proposed to enrich feature of inconsistent region. Thirdly, a multi-task learning to predict lung field based on the enriched feature was conducted. Finally, experiments to evaluate this algorithm were performed on JSRT and Montgomery dataset. The maximum improvement of Dice coefficient and accuracy were 1.99% and 2.27%, respectively, comparing with other representative algorithms. Results show that by enhancing the attention of boundary, this algorithm can improve the accuracy and reduce false segmentation.

Intelligent optimization-based pulmonary emphysema detection with adaptive multi-scale dilation assisted residual network with Bi-LSTM layer

Emphysema is a type of chronic disorder that occurs due to the abnormal expansion of the alveoli in the lungs. This causes several difficulties to breathe, and the final stage of emphysema leads to lung cancer. By using pulmonary function tests and Computed Tomography (CT) scans, the progressive destruction of emphysema disease can be addressed. The early-stage detection of emphysema may help to reduce the risk of the patients. The primary diagnosis should be carried out with the help of CT and spirometry analysis to reduce mortality rates. In the computer vision-based approaches, the inter-observer and intra-observer variations are difficult to analyze. To overcome these complications, an enhanced emphysema disease detection model with adaptive deep-structured architecture is implemented. At first, the input images are undergone, where histogram equalization, pre-processing stage, filtering technique, and Contrast Limited Adaptive Histogram Equalization methods are utilized. Then, the pre-processed images are fed into the lung segmentation process by using Fuzzy C-Means Clustering (FCM) and Adaptive Region Growing. Those segmented images are provided as input for the proposed emphysema detection by utilizing Adaptive Multi-Scale Dilation Assisted Residual Network with Bi-LSTM Layer (AMSD-RN-Bi-LSTM) layer, in which the constraints are optimized by using Improved Honey Badger Algorithm (IHBA). Through the experimental analysis, the proposed pulmonary emphysema detection model shows the tendency to give a rapid diagnosis of a disease that aids to identify the disease and diagnose them effectively.

Automated MRI Lung Segmentation and 3D Morphologic Features for Quantification of Neonatal Lung Disease

Automated MRI Lung Segmentation and 3D Morphologic Features for Quantification of Neonatal Lung Disease

Computed tomography imaging and clinical characteristics of pulmonary ground-glass nodules ≤2 cm with micropapillary pattern.

The aim of this study was to investigate the imaging features, lymph node metastasis, and genetic mutations in micropapillary lung adenocarcinoma (imaging with mixed ground-glass nodules) ≤2 cm, to provide a more precise and refined basis for the selection of lung segment resection. A retrospective analysis of 162 patients with surgically resected pathologically confirmed cancers ≤2.0 cm in diameter (50 cases of micropapillary mixed ground-glass nodules [mGGNs], 50 cases of nonmicropapillary mGGNs, and 62 cases of micropapillary SNs [solid nodules]) was performed. mGGNs were classified into five categories according to imaging features. The distribution of these five morphologies in micropapillary with mGGN and nonmicropapillary with mGGN was analyzed. The postoperative pathology and prognosis of lymph node metastasis were also compared between micropapillary mGGNs and micropapillary with SNs. After searching the TCGA database, we demonstrated heterogeneity, high malignancy and high risk of microcapillary lung cancer cancers. Different pathological subtypes of mGGN differed in morphological features (p < 0.05). The rate of lymph node metastasis was significantly higher in micropapillary mGGNs than in nonmicropapillary mGGNs. In the TCGA database samples, lactate transmembrane protein activity, collagen transcription score, and fibroblast EMT score were remarkably higher in micropapillary adenocarcinoma. Other pathological subtypes had a better survival prognosis and longer disease-free survival compared with micropapillary adenocarcinoma. mGGNs ≤2 cm with a micropapillary pattern have a higher risk of lymph node metastasis compared with SNs, and computed tomography (CT) imaging features can assist in their diagnosis.

Enhancing 3D Lung Infection Segmentation with 2D U-Shaped Deep Learning Variants

Accurate lung segmentation plays a vital role in generating 3D projections of lung infections, which contribute to the diagnosis and treatment planning of various lung diseases, including cases like COVID-19. This study capitalizes on the capabilities of deep learning techniques to reconstruct 3D lung projections from CT-scans. In this pursuit, we employ well-established 2D architectural frameworks like UNet, LinkNet, Attention UNet, UNet 3+, and TransUNet. The dataset used comprises 20 3D CT-scans from COVID-19 patients, resulting in over 2900 raw 2D slices. Following preprocessing, the dataset is refined to encompass 2560 2D slices tailored for modeling. Preprocessing procedures involve mask refinement, image resizing, contrast limited adaptive histogram equalization (CLAHE), and image augmentation to enhance the data quality and diversity. Evaluation metrics, including Intersection over Union (IoU) and dice scores, are used to assess the models’ performance. Among the models tested, Attention UNet stands out, demonstrating the highest performance. Its key trait of harnessing attention mechanisms enhances its ability to focus on crucial features. This translates to exceptional results, with an IoU score of 85.36% and dice score of 91.49%. These findings provide valuable insights into guiding the selection of an appropriate architecture tailored to specific requirements, considering factors such as segmentation accuracy and computational resources, in the context of 3D lung projection reconstruction.

Towards AI-augmented radiology education: a web-based application for perception training in chest X-ray nodule detection.

Artificial intelligence (AI)-based applications for augmenting radiological education are underexplored. Prior studies have demonstrated the effectiveness of simulation in radiological perception training. This study aimed to develop and make available a pure web-based application called Perception Trainer for perception training in lung nodule detection in chest X-rays. Based on open-access data, we trained a deep-learning model for lung segmentation in chest X-rays. Subsequently, an algorithm for artificial lung nodule generation was implemented and combined with the segmentation model to allow on-the-fly procedural insertion of lung nodules in chest X-rays. This functionality was integrated into an existing zero-footprint web-based DICOM viewer, and a dynamic HTML page was created to specify case generation parameters. The result is an easily accessible platform-agnostic web application available at: https://castlemountain.dk/mulrecon/perceptionTrainer.html.The application allows the user to specify the characteristics of lung nodules to be inserted into chest X-rays, and it produces automated feedback regarding nodule detection performance. Generated cases can be shared through a uniform resource locator. We anticipate that the description and availability of our developed solution with open-sourced codes may help facilitate radiological education and stimulate the development of similar AI-augmented educational tools. A web-based application applying AI-based techniques for radiological perception training was developed. The application demonstrates a novel approach for on-the-fly generation of cases in chest X-ray lung nodule detection employing deep-learning-based segmentation and lung nodule simulation.

Association of outcomes in point-of-care lung ultrasound for bronchiolitis in the pediatric emergency department

BackgroundAcute bronchiolitis (AB) is the most common lower respiratory tract infection in infants. Objective scoring tools and plain film radiography have limited application, thus diagnosis is clinical. The role of point-of-care lung ultrasound (LUS) is not well established. ObjectiveWe sought to characterize LUS findings in infants presenting to the pediatric ED diagnosed with AB, and to identify associations between LUS and respiratory support (RS) at 12 and 24 h, maximum RS during hospitalization, disposition, and hospital length of stay (LOS). MethodsInfants ≤12 months presenting to the ED and diagnosed with AB were enrolled. LUS was performed at the bedside by a physician. Lungs were divided into 12 segments and scanned, then scored and summated (min. 0, max. 36) in real time accordingly:0 - A lines with <3 B lines per lung segment.1 - ≥3 B lines per lung segment, but not consolidated.2 - consolidated B lines, but no subpleural consolidation.3 - subpleural consolidation with any findings scoring 1 or 2.Chart review was performed for all patients after discharge. RS was categorized accordingly: RS (room air), low RS (wall O2 or heated high flow nasal cannula <1 L/kg), and high RS (heated high flow nasal cannula ≥1 L/kg or positive pressure). Results82 subjects were enrolled. Regarding disposition, the mean (SD) LUS scores were: discharged 1.18 (1.33); admitted to the floor 4.34 (3.62); and admitted to the ICU was 10.84 (6.54). For RS, the mean (SD) LUS scores at 12 h were: no RS 1.56 (1.93), low RS 4.34 (3.51), and high RS 11.94 (6.17). At 24 h: no RS 2.11 (2.35), low RS 4.91 (3.86), and high RS 12.64 (6.48). Maximum RS: no RS 1.22 (1.31), low RS 4.11 (3.61), and high RS 10.45 (6.16). Mean differences for all dispositions and RS time points were statistically significant (p < 0.05, CI >95%). The mean (SD) hospital LOS was 84.5 h (SD 62.9). The Pearson correlation coefficient (r) comparing LOS and LUS was 0.489 (p < 0.0001). ConclusionHigher LUS scores for AB were associated with increased respiratory support, longer LOS, and more acute disposition. The use of bedside LUS in the ED may assist the clinician in the management and disposition of patient's diagnosed with AB.

Joint classification and segmentation for an interpretable diagnosis of acute respiratory distress syndrome from chest x-rays.

Acute respiratory distress syndrome (ARDS) is a life-threatening condition that can cause a dramatic drop in blood oxygen levels due to widespread lung inflammation. Chest radiography is widely used as a primary modality to detect ARDS due to its crucial role in diagnosing the syndrome, and the x-ray images can be obtained promptly. However, despite the extensive literature on chest x-ray (CXR) image analysis, there is limited research on ARDS diagnosis due to the scarcity of ARDS-labeled datasets. Additionally, many machine learning-based approaches result in high performance in pulmonary disease diagnosis, but their decisions are often not easily interpretable, which can hinder their clinical acceptance. This work aims to develop a method for detecting signs of ARDS in CXR images that can be clinically interpretable. To achieve this goal, an ARDS-labeled dataset of chest radiography images is gathered and annotated for training and evaluation of the proposed approach. The proposed deep classification-segmentation model, Dense-Ynet, provides an interpretable framework for automatically diagnosing ARDS in CXR images. The model takes advantage of lung segmentation in diagnosing ARDS. By definition, ARDS causes bilateral diffuse infiltrates throughout the lungs. To consider the local involvement of lung areas, each lung is divided into upper and lower halves, and our model classifies the resulting lung quadrants. The quadrant-based classification strategy yields the area under the receiver operating characteristic curve of 95.1% (95% CI 93.5 to 96.1), which allows for providing a reference for the model's predictions. In terms of segmentation, the model accurately identifies lung regions in CXR images even when lung boundaries are unclear in abnormal images. This study provides an interpretable decision system for diagnosing ARDS, by following the definition used by clinicians for the diagnosis of ARDS from CXR images.

Embolization and bleeding in patients with anticardiolipin antibodies

The patient, a 32-year-old male, was positive for anti-cardiolipin antibody syndrome and had a history of smoking for 15 years. He underwent “inferior vena cava filter implantation + catheter hemolysis” 10 years ago due to deep vein thrombosis of the right lower limb. After surgery, he continued to take farwarin and aspirin orally. Three years ago, due to thrombosis again, intravenous thrombolysis was performed in the lower extremity. After surgery, pharhualin was taken orally continuously. Massive hemoptysis occurred after intense exercise and eating spicy stimulating food 5 months ago. DSA showed abnormal bleeding in the right bronchial artery and right internal thoracic artery, so 300-500 um microsphere was used to embolize the peripheral artery. 350-560 um PVA granules embolized bronchial arteries; Three months ago, due to the recurrence of hemoptysis after oral administration of TCM blood-promoting drugs, angiography showed an increase in the number of responsible arteries, abnormal bleeding of the left bronchial artery and the right upper lobar bronchial artery of the lung segment, the formation of the middle and lower lobar bronchial artery of the right lung segment and the formation of collateral circulation of the right internal thoracic artery. Embolization of the above arteries was performed, and the patient recovered well after surgery and was discharged from hospital.

GAN inversion-based semi-supervised learning for medical image segmentation

The crux of medical image segmentation stems from learning pixel-wise semantic consistency from a mount of labeled samples through exhaustive annotation. Most existing methods rely heavily on labeled image pairs and are optimized in a fully supervised manner, often resulting in unsatisfactory results due to limited paired training samples. In this paper, we propose a GAN inversion-based semi-supervised learning framework InvSSL for medical image segmentation, which generates corresponding variants from labeled images by GAN inversion. Specifically, it first trains a StyleGAN as an inversion generator which is able to generate samples with the same distribution as training samples, and then obtains variant samples from a training sample by performing GAN inversion to strengthen the segmentation performance in a semi-supervised learning manner. In particular, inversion generator inverts a training sample into the latent space and adds interference to the latent codes to evolve variant samples. To model high-level consistency between pixels and decouple semantic components in the latent space, our InvSSL designs a multi-level dense contrastive learning mechanism. Extensive experiments on lung segmentation and skin lesion segmentation demonstrate that our InvSSL outperforms state-of-the-art methods for medical image segmentation. Our code will be released at https://github.com/funkdub/InvSSL.

Focal modulation network for lung segmentation in chest X-ray images

Focal modulation network for lung segmentation in chest X-ray images

An Amalgamated Deep Learning Approach for Lung Segmentation using X-Ray Images

The X-ray images are considered to be an affordable means to diagnose the diseases especially the cardio vascular diseases. Chest X-rays are even examined by modern doctors in order to identify the disease and cure them accordingly. Due to opacified nature of the X-Ray images and different morphology of human body, the lung segmentation of human chest becomes very tedious. In this paper, we have introduced an approach to overcome this limitation and achieve the art-of-state lung segmentation despite of different anatomy and opacity. The approach involves division formula based on U-Net. The used the previously trained models MobileNetV2 and InceptionResNetV2. The result acquired shows the increase in the efficiency of segmentation by 2.5% with respect to Dice score and 2.37% with respect to IoU when compared it with the tradition U-Net.

ОСОБЕННОСТИ ТЕЧЕНИЯ ВНЕБОЛЬНИЧНОЙ ПНЕВМОНИИ И ОЦЕНКА ЭФФЕКТИВНОСТИ КОМПЛЕКСНОЙ ТЕРАПИИ

Introduction. Community-acquired pneumonia is an urgent problem of our time due to its high prevalence, increasing resistance of microflora to antibacterial therapy, a large number of complications and high mortality. Aim. The aim of the study was features of the course of community-acquired pneumonia and evaluation of the effectiveness of complex therapy. Material and methods. A survey of 120 patients was conducted, with a diagnosis of moderate and severe pneumonia. The patients were divided into 2 groups: group 1 (70 patients), in which, along with the standard therapy of pneumonia, the drug meglumine sodium succinate was included, and group 2 (50 patients), which received standard therapy of pneumonia. In addition to the standard clinical examination of patients with pneumonia, concomitant pathology from other organs and systems was studied. In addition to anamnesis and objective research methods, laboratory and instrumental methods of studying the respiratory and hepatobiliary systems were used. Additionally, procalcitonin was determined in blood serum by immunochromatography, and hepatocyte growth factor and endothelin-1 were determined by enzyme-linked immunosorbent assay. Statistical processing of the results of the study was carried out using parametric and nonparametric methods. Results and discussion. Community- acquired pneumonia is characterized by a moderate course, segmental or lobular lung damage, and the development of a number of complications. With community-acquired pneumonia, the hepatobiliary system is affected and often community-acquired pneumonia occurs against the background of cardiovascular pathology. Conclusion. When meglumine sodium succinate was included in the complex therapy of community-acquired pneumonia, in addition to the positive dynamics of the clinical and laboratory picture of the underlying disease, and a significant reduction in the average duration of hospitalization. In the course of therapy of patients with community-acquired pneumonia with meglumine sodium succinate, a significant decrease in inflammation indicators and an increase in hepatocyte growth factor concentration were observed. When meglumine sodium succinate was included in the therapy of patients with pneumonia, there was a tendency to increase the number of endotheliocytes and a significant increase in the content of endothelin-1, which may reflect an increase in the sanogenetic ability of the body to systemic inflammation in community-acquired pneumonia.

Automatic measurement of Cardiothoracic Ratio in chest X-ray images with ProGAN-Generated dataset

CT Cardiomegaly could be identified using the CTR (Cardiothoracic Ratio), which could be assessed on a chest image or X-ray. It is determined using the link between the size of the heart and chest’s transverse dimension. When the ratio exceeds a certain threshold, cardiomegaly is diagnosed. The objective of the study is to offer an approach for calculating the ratio for categorizing cardiomegaly within chest X-ray pictures. The suggested method begins by building heart and lung segmentation models on the basis of U-Net design utilizing publicly accessible datasets containing lung and heart mask ground truth. The segmented lung and heart portion sizes are then used to calculate the ratio. Additionally, chest X-ray images from 3 classes-cardiomegaly, female,and male normal - are created using a novel dataset using PGAN (Progressive Growing of GANs). This dataset is then utilized to evaluate the suggested solution. The suggested approach is also utilized to assess the quality of PGAN-generated chest X-ray pictures. In the trials, lung and heart areas in chest images of X-rays are segmented using trained models and a self-gathered dataset. The computed values of CTR are contrasted with those that were manually assessed by specialists. The average inaccuracy is 3.08 percent. The models are then used to segment areas of the lung and heart for CTR computation on the PGAN dataset. The cardiomegaly is then measured utilizing multiple attempts with varying cut-off threshold values. The proposed approach yields 94.20 percent specificity,88.31% sensitivity, and 94.61% accuracy with the usual cut-off of 0.50. The suggested approach is shown to be resilient across hitherto unexplored datasets for computation of CTR, segmentation, as well as cardiomegaly classification, such as the PGAN dataset. To increase sensitivity, modify the cut-off value to be less than 0.50. The proposed solution is then assessed from a variety of angles, such as lung and heart segmentation, CTR computation, as well as cardiomegaly classification. Tests are also carried out on publicly available datasets, self-collected datasets, as well as ProGAN- reconstructed datasets.

External validation, radiological evaluation, and development of deep learning automatic lung segmentation in contrast-enhanced chest CT

ObjectivesThere is a need for CT pulmonary angiography (CTPA) lung segmentation models. Clinical translation requires radiological evaluation of model outputs, understanding of limitations, and identification of failure points. This multicentre study aims to develop an accurate CTPA lung segmentation model, with evaluation of outputs in two diverse patient cohorts with pulmonary hypertension (PH) and interstitial lung disease (ILD).MethodsThis retrospective study develops an nnU-Net-based segmentation model using data from two specialist centres (UK and USA). Model was trained (n = 37), tested (n = 12), and clinically evaluated (n = 176) on a diverse ‘real-world’ cohort of 225 PH patients with volumetric CTPAs. Dice score coefficient (DSC) and normalised surface distance (NSD) were used for testing. Clinical evaluation of outputs was performed by two radiologists who assessed clinical significance of errors. External validation was performed on heterogenous contrast and non-contrast scans from 28 ILD patients.ResultsA total of 225 PH and 28 ILD patients with diverse demographic and clinical characteristics were evaluated. Mean accuracy, DSC, and NSD scores were 0.998 (95% CI 0.9976, 0.9989), 0.990 (0.9840, 0.9962), and 0.983 (0.9686, 0.9972) respectively. There were no segmentation failures. On radiological review, 82% and 71% of internal and external cases respectively had no errors. Eighteen percent and 25% respectively had clinically insignificant errors. Peripheral atelectasis and consolidation were common causes for suboptimal segmentation. One external case (0.5%) with patulous oesophagus had a clinically significant error.ConclusionState-of-the-art CTPA lung segmentation model provides accurate outputs with minimal clinical errors on evaluation across two diverse cohorts with PH and ILD.Clinical relevanceClinical translation of artificial intelligence models requires radiological review and understanding of model limitations. This study develops an externally validated state-of-the-art model with robust radiological review. Intended clinical use is in techniques such as lung volume or parenchymal disease quantification.Key Points• Accurate, externally validated CT pulmonary angiography (CTPA) lung segmentation model tested in two large heterogeneous clinical cohorts (pulmonary hypertension and interstitial lung disease).• No segmentation failures and robust review of model outputs by radiologists found 1 (0.5%) clinically significant segmentation error.• Intended clinical use of this model is a necessary step in techniques such as lung volume, parenchymal disease quantification, or pulmonary vessel analysis.Graphical

Optimization of Median Modified Wiener Filter for Improving Lung Segmentation Performance in Low-Dose Computed Tomography Images

In low-dose computed tomography (LDCT), lung segmentation effectively improves the accuracy of lung cancer diagnosis. However, excessive noise is inevitable in LDCT, which can decrease lung segmentation accuracy. To address this problem, it is necessary to derive an optimized kernel size when using the median modified Wiener filter (MMWF) for noise reduction. Incorrect application of the kernel size can result in inadequate noise removal or blurring, degrading segmentation accuracy. Therefore, various kernel sizes of the MMWF were applied in this study, followed by region-growing-based segmentation and quantitative evaluation. In addition to evaluating the segmentation performance, we conducted a similarity assessment. Our results indicate that the greatest improvement in segmentation performance and similarity was at a kernel size 5 × 5. Compared with the noisy image, the accuracy, F1-score, intersection over union, root mean square error, and peak signal-to-noise ratio using the optimized MMWF were improved by factors of 1.38, 33.20, 64.86, 7.82, and 1.30 times, respectively. In conclusion, we have demonstrated that by applying the MMWF with an appropriate kernel size, the optimization of noise and blur reduction can enhance segmentation performance.

A radiomics-based logistic regression model for the assessment of emphysema severity.

The aim of this study is to develop a model that differentiates between the radiological patterns of severe and mild emphysema using radiomics parameters, as well as to examine the parameters included in the model. Over the last 12 months, a total of 354 patients were screened based on the presence of terms such as “Fleischner”, “CLE”, and “centriacinar” in their thoracic CT reports, culminating in a study population of 82 patients. The study population was divided into Group 1 (Fleischner mild and moderate; n= 45) and Group 2 (Fleischner confluent and advanced destructive; n= 37). Volumetric segmentation was performed, focusing on the upper lobe segments of both lungs. From these segmented volumes, radiomics parameters including shape, size, first-order, and second-order features were calculated. The best model parameters were selected based on the Bayesian Information Criterion and further optimized through grid search. The final model was tested using 1000 iterations of bootstrap resampling. In the training set, performance metrics were calculated with a sensitivity of 0.862, specificity of 0.870, accuracy of 0.863, and AUC of 0.910. Correspondingly, in the test set, these values were sensitivity= 0.848; specificity= 0.865; accuracy= 0.857; and AUC= 0.907. The logistic regression model, composed of radiomics parameters and trained on a limited number of cases, effectively differentiated between mild and severe radiological patterns of emphysema using computed tomography images.