Lung Segmentation Research Articles (Page 1)

Pneumonia remains a serious global health issue, particularly affecting vulnerable groups such as children and the elderly, where timely and accurate diagnosis is critical for effective treatment. Recent advances in deep learning have significantly enhanced pneumonia detection using chest X-rays, yet many current methods still face challenges with interpretability, efficiency, and clinical applicability. In this work, we proposed a YOLOv11-based deep learning framework designed for real-time pneumonia detection, strengthened by the integration of Grad-CAM for visual interpretability. To further enhance robustness, the framework incorporated preprocessing techniques such as Contrast Limited Adaptive Histogram Equalization (CLAHE) for contrast improvement, region-of-interest extraction, and lung segmentation, ensuring both precise localization and improved focus on clinically relevant features. Evaluation on two publicly available datasets confirmed the effectiveness of the approach. On the COVID-19 Radiography Dataset, the system reached a macro-average accuracy of 98.50%, precision of 98.60%, recall of 97.40%, and F1-score of 97.99%. On the Chest X-ray COVID-19 & Pneumonia dataset, it achieved 98.06% accuracy, with corresponding high precision and recall, yielding an F1-score of 98.06%. The Grad-CAM visualizations consistently highlighted pathologically relevant lung regions, providing radiologists with interpretable and trustworthy predictions. Comparative analysis with other recent approaches demonstrated the superiority of the proposed method in both diagnostic accuracy and transparency. With its combination of real-time processing, strong predictive capability, and explainable outputs, the framework represents a reliable and clinically applicable tool for supporting pneumonia and COVID-19 diagnosis in diverse healthcare settings.

ObjectivesTo investigate the feasibility of a deep learning model, using a transfer learning approach, for recognizing high-altitude pulmonary edema (HAPE) on chest X-ray images and exploring its capability for assessing severity.Study designRetrospective study.MethodsThis retrospective study utilized a multi-source dataset. The pretraining set was derived from the ARXIV_V5_CHESTXRAY database (3,923 images, including 2,303 with edema labels). The primary HAPE-specific training set comprised radiographs from the 950th Hospital of the Chinese People’s Liberation Army (1,003 HAPE cases and 702 normal controls; 2007–2023). An external validation set was constructed from recent cases (Jan-Dec 2023) from two hospitals (679 HAPE cases and 436 normal controls), with strict patient separation. We implemented a multi-stage pipeline: (1) A DeepLabV3_ResNet-50 model was trained for lung segmentation on a subset of the pretraining set; (2) MobileNet_V2 and VGG19 architectures underwent pretraining for general pulmonary edema severity grading on the ARXIV_V5_CHESTXRAY dataset; (3) These models were then fine-tuned on the HAPE-specific training set.ResultsThe segmentation model achieved a Dice coefficient of 99.03%. The binary classification model (VGG19) for edema detection achieved a validation AUC of 0.950. The multi-class models (MobileNet_V2) achieved macroaverage AUCs of 0.92 (3-class) and 0.89 (4-class). The model demonstrated high performance in distinguishing normal (class 0) and severe edema (class 3) (sensitivities: 0.91, 0.88). However, performance was critically low for intermediate grades (classes 1 and 2; sensitivities: 0.16, 0.37).ConclusionsTransfer learning from general to HAPE-specific edema data produced a model that accurately segments lungs and differentiates severe HAPE from normal cases with high performance. However, its failure to reliably identify intermediate grades underscores the challenges of domain shift and fine-grained radiographic assessment. This work highlights both the promise and pitfalls of using heterogeneous datasets for rare disease diagnosis.

To propose simultaneous acquisition of free-breathing, noncontrast-enhanced 3D perfusion-weighted (QW) and ventilation-weighted (VW) maps using 3D ultrashort echo-time (UTE) magnetic resonance imaging (MRI). This prospective study included 1 healthy volunteer (25 years; female) and 5 patients (65 ± 10 y; 1 female) with diffuse pulmonary diseases [2 chronic obstructive pulmonary disease (COPD), 2 interstitial lung disease (ILD), 1 asthma], conducted between January 2022 and March 2024. Three-dimensional QW and VW maps were obtained through retrospective cardiac and respiratory gating using 3D UTE MRI on a 3T clinical scanner (Magnetom Prisma; Siemens Healthineers). QW maps were generated by voxel-wise subtraction between maximum and minimum values of 8 cardiac phase-resolved images at end-expiration, and VW maps by subtraction between end-inspiration and end-expiration images. Validation of QW maps involved: (1) assessment of coefficient of variation (CV) across 12 lung segments compared with SPECT, (2) structural similarity index measure (SSIM) analysis compared with SPECT, and (3) evaluation of anteroposterior gravity-dependence by 1D coronal slice profiles. Repeatability was tested in one healthy subject with multiple scans on separate days. In patients, regional perfusion was assessed in lesions identified on CT, and V/Q match or mismatch was evaluated in asthma and emphysema-predominant COPD. Statistical analysis included SSIM and Mann-Whitney U tests (P < 0.05). UTE MRI-based QW and VW maps showed high similarity with corresponding SPECT maps [SSIM: 0.86 (QW), 0.87 (VW); P >0.05 for CV across 12 lung segments]. Both maps demonstrated gravity-dependence with high correlation to SPECT (correlation coefficient: QW = 0.91, VW = 0.96). QW maps show reduced perfusion in emphysema regions and increased perfusion in regions with consolidation, ground-glass opacity (GGO), and inflammation around fibrotic cysts. Comparing asthma and emphysema-predominant COPD, QW and VW maps demonstrated V/Q mismatch in asthma but matched defects in COPD. Simultaneous noncontrast-enhanced 3D UTE MRI effectively provides reliable regional perfusion and ventilation information for pulmonary disease evaluation without exposure to ionizing radiation. By providing perfusion and ventilation information simultaneously, the proposed method can help to provide precise and comprehensive functional assessment of pulmonary diseases, including differentiation of pathophysiological conditions and improved evaluation of disease severity and prognosis.

Due mostly to the relative lack of effective chemotherapeutic methods, the lung infection incidence by Pseudomonas aeruginosa that are categorized as multi-drug resistant has significantly raised. This study demonstrated the strong antioxidant and anti-infectious properties of copper nanoparticles (CuNPs) made using an aqueous extract of Allium sativum in vivo. We employed FE-SEM, UV-Vis, XRD, EDX, and TEM to determine the characteristics of the CuNPs that were created when an aqueous extract of Allium sativum reacted with a copper nitrate solution. The fatal dosage of P. aeruginosa is assessed in mice as part of an in vivo investigation, and the clinical manifestations-such as bacteremia, hypothermia, and weight loss-are analyzed 48h after infection. Infected mice's body temperature significantly decreased from 38.8°C (0h) to 32.5°C (at 48h), and after the trial, a 20% weight loss was noted. In comparison to day 1, when the bacterial burden was determined to be 1.5 Log10CFU/mL, the bacterial burden was 0.1 Log10CFU/mL on day 8, indicating a considerable decrease. According to histopathological findings, there was a more widespread and patchy buildup of alveolar space inflammatory cells, and infiltrates were seen in every lung segment of the infected animals. At 100µg/kg, the research unequivocally states that the CuNPs is efficient on P. aeruginosa-induced lung infections. The lung infection histopathological scores of the group treated by the CuNPs were 0, 0, 1, and 1 related to the alveoli distortion and degeneration, inflammatory cell infiltration, congestion, and inflammatory cell infiltration, respectively. Following human clinical trials, the newly developed green-formulated copper particles may be administrated as a novel anti-infectious medication or dietary supplement.

BackgroundDrug-resistant pulmonary tuberculosis (DR-TB) remains a critical public health challenge, particularly affecting vulnerable populations such as the elderly, who exhibit higher morbidity and mortality rates. This study aims to elucidate the chest CT characteristics of DR-TB in elderly patients to improve diagnostic accuracy and guide individualized treatment strategies.MethodsA retrospective analysis of 183 confirmed DR-TB cases (Huai’an Infectious Disease Hospital, June 2013-June 2023) compared chest CT findings (lesion distribution, extent, and morphology) between elderly patients (≥ 60 years) and non-elderly patients (14–59 years).ResultsKey findings reveal that elderly patients demonstrate a higher frequency of extensive lung involvement, with 76% exhibiting lesions in all lung lobes compared to 40.74% in the non-elderly group (P < 0.001). Additionally, the elderly group displayed significantly more pathological features, such as segmental and lobar shadows (61.33% vs. 45.37%, P = 0.033) and lung destruction (22.67% vs. 11.11%, P = 0.035).ConclusionThe identification of risk factors on chest CT, including the presence of pulmonary and bronchial lesions, highlights the necessity for tailored screening and management strategies for elderly DR-TB patients.

Mucus plug formation and chronic bronchitis are manifestations of mucus pathology in chronic obstructive pulmonary disease. Identifying gene expression changes related to mucus pathology could provide insight into its pathogenesis. To investigate gene expression changes in individuals with mucus plugs, identify related biological pathways, and assess whether mucus plug-related gene expression associates with clinical features of other mucus pathologies. We studied 290 participants from the Detection of Early Lung Cancer Among Military Personnel 2 study with mainstem bronchial brush bulk RNA-sequencing data (n = 204 discovery, n = 86 validation). We scored mucus plugging based on the number of lung segments with mucus plugs identified on chest computed tomography scans and used correlative analysis to identify differentially expressed genes and examine their association with chronic bronchitis symptoms. 76 participants (37%) in the discovery set had mucus plugs. Differentially expressed genes were broadly epithelial- or immune-related. Epithelial-related genes show decreased expression of genes involved in cilia maintenance and microtubule function and increased expression of genes related to epithelial maintenance and protection. Expression patterns of epithelial-related genes are associated with chronic bronchitis symptoms. Immune-related genes are enriched for innate and adaptive pathways. Expression of immune genes varies by lung function and was more weakly associated with mucus plugs than that of epithelial-related genes. Findings were replicated in an independent validation set. Several distinct gene expression patterns are linked to the presence of mucus plugs, highlighting biological pathways involved in mucus pathophysiology. Variability in gene expression suggests a spectrum of mucus pathophysiology contributes to mucus plugs and chronic bronchitis symptoms.

Segmentation of medical images is critical for the evaluation, diagnosis, and treatment of various medical conditions. While deep learning-based approaches are the dominant methodology, they rely heavily on abundant labeled data and face significant challenges when data is limited. Semi-supervised learning methods mitigate this issue but there are still some challenges associated with them. Additionally, these approaches can be improved specifically for medical images considering their unique properties (e.g., smooth boundaries). In this work, we adapt and enhance the well-established pseudo-labeling approach specifically for medical image segmentation. Our exploration consists of modifying the network’s loss function, pruning the pseudo-labels, and refining pseudo-labels by integrating traditional image processing methods with semi-supervised learning. This integration enables traditional segmentation techniques to complement deep semi-supervised methods, particularly in capturing fine edges where deep models often struggle. It also incorporates the smoothness of the edges in the segmentation and achieves a balance between deep learning and traditional methods through tunable parameters. Moreover, to address the problem of noisy or unreliable pseudo-labels, we utilize uncertainty-based pixel-level and image-level pruning of the pseudo-labels using a specific loss function, thereby improving the accuracy and robustness of the segmentation. We evaluated our approach on three different datasets from two imaging modalities (CT and MRI) and demonstrated its superior performance, highlighting its accuracy and robustness in the presence of limited labeled data. With only 15% of the labeled data, on the Sunnybrook Cardiac dataset, our approaches increased endocardium segmentation accuracy from 82.1% to 87.5%, and epicardium segmentation from 82.5% to 86.7%. On the COVID-19 CT lung and infection segmentation dataset, our approach improved left lung segmentation accuracy from 72.5% to 79.3%, and right lung segmentation from 75.8% to 81.6% when using only 15% of labeled data. On the Automated Cardiac Diagnostic Challenge dataset, with just 10% of labeled data, our approach increased endocardium segmentation from 91% to 93.7%, myocardium from 69.8% to 74.5%, and right ventricle from 76.7% to 82.1%. Our codes will be published in https://github.com/behnam-rahmati.

Purpose: Monitoring disease development or viruses that invade our bodies, such as Coronavirus Disease of 2019 (COVID-19), can be effectively carried out using Computed Tomography (CT) imaging tools. However, manual assessment of CT images by consultants is often insufficient for determining the extent of lung damage in COVID-19 patients. Automated evaluation of lung damage addresses this limitation by optimizing healthcare resource utilization. It reduces the workload on radiologists, allowing them to concentrate on more complex cases. Additionally, it ensures accurate and consistent assessments of lung damage, minimizing variability and the potential for human error inherent in manual evaluations. Materials and Methods: In this study, a new approach was presented for improving CT images of the lung and specifying further lesions. This will help calculate the extent of damage without human intervention. The structure of the proposed technique draws upon four phases (data collection, improvement, segmentation and extraction lung damage region and evaluation). Firstly, 100 patients were recruited between September 29 2020 and July 10, 2022, of whom tested positive for COVID-19 and CT images were collected, then composite technique is implemented to extract the percentage of lung damage of COVID-19 patients. Results: The study results demonstrated an efficient method for quickly and practically calculating the percentage of lung damage. There is a clear convergence between manual evaluation, done by radiologists, and automatic evaluation using the proposed method, suggesting its potential as an alternative in the absence of a specialist doctor. The differences in the arithmetic mean between the proposed technique and the radiologists' evaluations were 3.5%, 10%, 18%, and 0.98% for radiologists 1, 2, 3, and 4, respectively. Additionally, the findings indicated that individuals aged 20-60 years are the most affected by COVID-19. Conclusion: This method serves as a potent tool for swiftly and practically assessing the percentage of lung damage caused by COVID-19. By eliminating the need for human intervention, it enables the evaluation of lung damage autonomously. This feature makes it particularly valuable in telemedicine applications and emergency situations where specialist medical expertise may not be readily available.

Robust T-Loss for medical image segmentation.

IntroductionBronchoscopic thermal vapour ablation (BTVA) is a bronchoscopic lung volume reduction technique, also recommended by the global initiative for chronic obstructive lung disease (GOLD) guidelines. Previous studies on BTVA have primarily focused on segmental treatment, targeting the most severely affected lung segments while preserving healthier areas. However, there is considerable variability in the severity of emphysema within subsegments of these lung segments, suggesting that a more precise approach could potentially improve treatment outcomes. This study aims to evaluate the efficacy and safety of subsegmental BTVA, which may better preserve healthy lung tissue while more accurately targeting the most severely affected regions in patients with severe emphysema.Methods and analysisThis is a prospective, multicentre, randomised, controlled, open-label clinical trial conducted in China. A total of 100 patients with severe emphysema, who continue to experience significant symptoms despite optimal medical therapy according to GOLD guidelines, will be enrolled. Participants will be randomly assigned in a 1:1 ratio to either the experimental group (subsegmental BTVA) or the control group (segmental BTVA), both receiving optimal medical therapy. BTVA will be performed in two separate procedures, with the second procedure occurring at least 6 weeks and no more than 6 months after the first. The primary endpoint is the change in forced expiratory volume in 1 s at 6 months following the second procedure. Secondary endpoints include changes in lung volumes, quality of life (measured by the St. George’s respiratory questionnaire for chronic obstructive pulmonary disease patients), other parameters in pulmonary function tests, and 6 min walk distance at 6 and 12 months, etc.Ethics and disseminationThe trial protocol was approved by the Ethics Committee of Shanghai Chest Hospital (IS23073). Additionally, study approval was obtained from local regulatory boards at each site before the study commenced. The results will be published in a peer-reviewed journal.Trial registration numberThis trial was registered on ClinicalTrials.gov on 16 November 2023 (NCT06152107).

Early detection of lung cancer is critical for improving treatment outcomes, and automatic lung image segmentation plays a key role in diagnosing lung-related diseases such as cancer, COVID-19, and respiratory disorders. Challenges include overlapping anatomical structures, complex pixel-level feature fusion, and intricate morphology of lung tissues all of which impede segmentation accuracy. To address these issues, this paper introduces GEANet, a novel framework for lung segmentation in CT images. GEANet utilizes an encoder-decoder architecture enriched with radiomics-derived features. Additionally, it incorporates Graph Neural Network (GNN) modules to effectively capture the complex heterogeneity of tumors. Additionally, a boundary refinement module is incorporated to improve image reconstruction and boundary delineation accuracy. The framework utilizes a hybrid loss function combining Focal Loss and IoU Loss to address class imbalance and enhance segmentation robustness. Experimental results on benchmark datasets demonstrate that GEANet outperforms eight state-of-the-art methods across various metrics, achieving superior segmentation accuracy while maintaining computational efficiency.

Decisions regarding veno-venous extracorporeal membrane oxygenation (vv-ECMO) in patients with acute respiratory distress syndrome (ARDS) are often based solely on clinical and physiological parameters, which may insufficiently reflect severity and heterogeneity of lung injury. This study aimed to develop a predictive model integrating machine learning-derived quantitative features from admission chest computed tomography (CT) with selected clinical variables to support early individualized decision-making regarding vv-ECMO therapy. In this retrospective single-center cohort study, 375 consecutive patients with COVID-19-associated ARDS admitted to the ICU between March 2020 and April 2022 were included. Lung segmentation from initial CTs was performed using a convolutional neural network (CNN) to generate high-resolution, anatomically accurate masks of the lungs. Subsequently, 592 radiomic features, quantifying lung aeration, density and morphology, were extracted. Four clinical parameters - age, mean airway pressure, lactate, and C-reactive protein, were selected on the basis of clinical relevance. Three logistic regression models were developed: (1) Imaging Model, (2) Clinical Model, and (3) Combined Model integrating different features. Predictive performance was assessed via the area under the receiver operating characteristic curve (AUROC), accuracy, sensitivity, and specificity. A total of 375 patients were included: 172 in the training and 203 in the validation cohort. In the training cohort, the AUROCs were 0.743 (Imaging), 0.828 (Clinical), and 0.842 (Combined). In the validation cohort, the Combined Model achieved the highest AUROC (0.705), outperforming the Clinical (0.674) and Imaging (0.639) Models. Overall accuracy in the validation cohort was 64.0% (Combined), 66.5% (Clinical), and 59.1% (Imaging). The Combined Model showed 68.1% sensitivity and 58.9% specificity. Kaplan-Meier analysis confirmed a significantly greater cumulative incidence of ECMO therapy in patients predicted as high risk (p < 0.001), underscoring its potential to support individualized, timely ECMO decisions in ARDS by providing clinicians with objective data-driven risk estimates. Quantitative CT features based on machine learning-derived lung segmentation allow early individualized prediction of the need for vv-ECMO in ARDS. While clinical data remain essential, radiomic markers enhance prognostic accuracy. The Combined Model demonstrates considerable potential to support timely and evidence-based ECMO initiation, facilitating individualized critical care in both specialized and general ICU environments.Trial registration: The study is registered with the German Clinical Trials Register under the number DRKS00027856. Registered 18.01.2022, retrospectively registered due to retrospective design of the study.

Bronchopulmonary vascular malformations constitute a broad spectrum of developmental disorders in which a part of the lung is perfused exclusively from the systemic arterial tree with or without tracheobronchial communication. Pseudosequestration is a rare congenital pulmonary anomaly involving a lung segment with abnormal systemic arterial supply but a preserved connection to the bronchial tree. We present the case of a 5-month-old male referred for evaluation of a cardiac murmur. Initial investigations, including laboratory tests and chest X-ray, revealed ground-glass opacities in the left lung, prompting further imaging. Thoracic computed tomography (CT) and CT angiography demonstrated an aberrant systemic artery originating from the descending aorta and supplying the left lower lobe. Digital subtraction angiography (DSA) was subsequently performed to rule out vascular malformations and confirmed the diagnosis of pseudosequestration. Endovascular treatment was successfully completed with coil embolization and occlusion of the aberrant vessel using an Amplatzer Vascular Plug. The patient remained asymptomatic during a 20-month follow-up period. The case underscores the critical role of early recognition and noninvasive imaging in managing rare congenital pulmonary vascular anomalies.

This study investigates the comparative performance of foundational models, advanced large-kernel architectures, and traditional deep learning approaches for hyperpolarized gas MRI segmentation across progressive data reduction scenarios. Chronic obstructive pulmonary disease (COPD) remains a leading global health concern, and advanced imaging techniques are crucial for its diagnosis and management. Hyperpolarized gas MRI, utilizing helium-3 (3He) and xenon-129 (129Xe), offers a non-invasive way to assess lung function. We evaluated foundational models (Segment Anything Model and MedSAM), advanced architectures (UniRepLKNet and TransXNet), and traditional deep learning models (UNet with VGG19 backbone, Feature Pyramid Network with MIT-B5 backbone, and DeepLabV3 with ResNet152 backbone) using four data availability scenarios: 100%, 50%, 25%, and 10% of the full training dataset (1640 2D MRI slices from 205 participants). The results demonstrate that foundational and advanced models achieve statistically equivalent performance across all data scenarios (p > 0.01), while both significantly outperform traditional architectures under data constraints (p < 0.001). Under extreme data scarcity (10% training data), foundational and advanced models maintained DSC values above 0.86, while traditional models experienced catastrophic performance collapse. This work highlights the critical advantage of architectures with large effective receptive fields in medical imaging applications where data collection is challenging, demonstrating their potential to democratize advanced medical imaging analysis in resource-limited settings.

IntroductionEmphysema is a debilitating form of chronic obstructive pulmonary disease (COPD) that causes lung hyperinflation and air trapping, leading to reduced quality of life. Bronchoscopic lung volume reduction (BLVR) with endobronchial valves (EBVs) offers a minimally invasive treatment option by collapsing diseased lung segments to improve respiratory mechanics. While both CT and pulmonary function testing (PFT) are used to assess outcomes, it remains unclear whether post-procedural volume changes measured by CT correlate with functional improvements seen on PFT.Materials and methodsWe performed a single-center chart review of patients who underwent BLVR with EBV from January 2019 to December 2023 who had both pre-and post-procedure PFT and CT volume analysis performed. Data recorded included clinical and demographic characteristics, post-valve analysis, and PFT.ResultsA total of 14 patients were included in our study. A comparison of total lung volume change between post-BLVR CT analysis and PFT showed a -3% total volume change in post-BLVR CT analysis versus -6% total volume change in PFT. Post-BLVR CT analysis showed a median volume change of -247 mL versus PFT with -490 mL volume change. The volume change in PFT and post-BLVR CT analysis did not show a linear correlation (p = 0.315, R2 = 0.101). ConclusionsOur study found that post-BLVR pulmonary function testing showed greater volume reduction and functional improvement compared to CT analysis, suggesting a discrepancy between anatomical and physiological assessments. These findings indicate that CT may underestimate therapeutic response, and PFT may be a more reliable tool for evaluating clinical outcomes after EBV placement.

BackgroundWith the popularity of LDCT screening, more and more small lung cancers have been found, and segmentectomy has been widely used because of its advantages in the treatment of early lung cancer, but the feasibility of segmentectomy is still controversial because of the increased complexity of the operation. Especially in complex lung segment surgery, it is more controversial. Preoperative three-dimensional reconstruction (3DR) is one of the effective methods to ensure the smooth operation, but its role in complex segmentectomy has not yet been verified. This article aims to evaluate the value of preoperative three-dimensional reconstruction in complex pulmonary segmentectomy by retrospective analysis of preoperative three-dimensional reconstruction assisted single-port video-assisted thoracoscopic complex pulmonary segmentectomy and comparison of surgical related indicators.MethodsThe clinical data of patients with lung cancers who underwent single-port thoracoscopic complex segmentectomy (n = 299) from August 2015 to February 2019 were retrospectively analyzed, including 156 patients in the preoperative three-dimensional reconstruction group and 143 patients in the non-three-dimensional reconstruction group. Perioperative outcomes were compared between the two groups after comparative propensity score matching analysis (PSM) according to patient age, gender, BMI, lung function, smoking history, major tumor components, and tumor size.ResultsThere were 125 patients in each group after PSM, and the baseline characteristics of patients were comparable. There were no significant differences in age, sex, BMI, smoking history, tumor histology and tumor size between the two groups (all P > 0.05). Lymph node dissection (9.4 ± 5.1 vs. 10.6 ± 7.0), postoperative drainage volume (510.5 ± 279.4 ml vs. 528.7 ± 379.4 ml), indwelling time of chest tube (2.3 ± 1.1d vs. 2.5 ± 1.6 d), and the incidence of chronic air leaks (0.8% vs. 0.8%), total complications (2.4% vs. 5.6%), 30-day postoperative mortality (0% vs. 0%) were not statistically different between the two groups. The operative time of preoperative 3DR group (178.6 ± 50.5 min vs. 202.1 ± 51.4 min), intraoperative blood loss (47.6 ± 37.9 ml vs. 58.4 ± 36.2 ml) compared with the control group (non-3DR), the difference was statistically significant (P < 0.05).ConclusionThis is the first study to evaluate the usefulness of preoperative 3D reconstruction in complex segmentectomy. The results showed that the use of preoperative three-dimensional reconstruction for complex pulmonary segmentectomy was a safe and effective method, which could significantly reduce the operation time and blood loss.

This research introduces a unique dataset targeting Silicosis, a significant global occupational lung disease, and a member of the Pneumoconiosis family. Addressing the challenges in healthcare data collection and the need for expert annotation, this dataset aims to aid AI algorithms in medical applications. The comprehensive dataset includes not only Silicosis cases but also related conditions, such as tuberculosis and silicotuberculosis, alongside healthy lung images, addressing the diagnostic complexity due to symptom overlap. As the first public dataset of its kind, it offers detailed annotations for lung and disease region segmentation, as well as disease prediction, provided by multiple radiologists. Baseline experiments and findings demonstrate that current AI models have limited predictive accuracy for these disease classes, emphasizing the critical need for dedicated research. It is our assertion that the proposed Silicodata can be a key dataset in designing automated Silicosis detection tools and addressing challenges associated with small sample sizes in medical AI research.

The COVID-19 pandemic has been the most catastrophic global health emergency of the 21^{st} century, resulting in hundreds of millions of reported cases and five million deaths. Chest X-ray (CXR) images are highly valuable for early detection of lung diseases in monitoring and investigating pulmonary disorders such as COVID-19, pneumonia, and tuberculosis. These CXR images offer crucial features about the lung’s health condition and can assist in making accurate diagnoses. Manual interpretation of CXR images is challenging even for expert radiologists due to the overlapping radiological features. Therefore, Artificial Intelligence (AI) based image processing took over the charge in healthcare. But still it is uncertain to trust the prediction results by an AI model. However, this can be resolved by implementing explainable artificial intelligence (XAI) tools that transform a black-box AI into a glass-box model. In this research article, we have proposed a novel XAI-TRANS model with inception based transfer learning addressing the challenge of overlapping features in multiclass classification of CXR images. Also, we proposed an improved U-Net Lung segmentation dedicated to obtaining the radiological features for classification. The proposed approach achieved a maximum precision of 98% and accuracy of 97% in multiclass lung disease classification. By leveraging XAI techniques with the evident improvement of 4.75%, specifically LIME and Grad-CAM, to provide detailed and accurate explanations for the model’s prediction.

Dose Reduction in 4-Dimensional Computed Tomography Imaging: Breathing Signal-Guided Deep Learning-Driven Data Acquisition.

Introduction: Lung diseases, such as COPD, pneumonia, and ILD, are significant global health concerns. Accurate lung volume quantification is crucial for effective diagnosis, treatment planning, and disease monitoring. Chest Computed Tomography (CT) offers detailed anatomical and functional insights, acting as a potential biomarker for early detection and prognosis. Aim: This study aimed to develop and validate a Chest CT-based approach for quantifying lung volume and airway changes in patients with various lung diseases, exploring its utility as a diagnostic and monitoring tool. Methodology: Conducted over six months at Aster CMI Hospital, Bengaluru, the study included 30 patients. A Philips Ingenuity 128 slices CT scanner was utilized for imaging. Patient data, including demographics, medical history, complaints, and diagnosis, were collected alongside CT scans. The methodology involved analysing existing CT scans and clinical data, assessing patient conditions, employing a standardized CT scan protocol, using semi-automated segmentation via the Philips IntelliSpace Portal system to isolate lungs, and automatically calculating lung volumes. Results: The study revealed significant variations in lung volume. COPD and ILD patients showed markedly larger lung volumes compared to those with asthma and lower respiratory tract infections. A strong correlation existed between lung volume and disease severity, most evident in COPD and ILD cases (e.g., mild COPD: 2.56L vs. severe: 3.89L; mild ILD: 2.34L vs. severe: 3.21L). Patients with asthma and lower respiratory tract infections generally had smaller lung volumes. The findings highlight lung volume measurement as a valuable biomarker for detecting disease progression and guiding treatment, with higher volumes correlating with increased hospitalization and mortality risks. Conclusion: This study effectively demonstrates the utility of semi-automated lung segmentation and lung volume measurement in lung disease patients. The results underscore its potential as a crucial biomarker for diagnosis and monitoring. Further research is necessary to fully explore its clinical applications and standardize protocols across diverse lung diseases.