A 3D-CNN model with CT-based parametric response mapping for classifying COPD subjects

Development and validation of a deep learning-based automated computed tomography image segmentation and diagnostic model for infectious hydronephrosis: a retrospective multicentre cohort study

Background

Accurately diagnosing whether hydronephrosis is complicated by infection is crucial for guiding appropriate clinical treatment. This study aimed to develop a fully automated segmentation and non-invasive diagnostic model for infectious hydronephrosis (IH) using CT images and a deep learning algorithm.

Methods

A retrospective analysis of clinical information and annotated cross-sectional CT images from patients diagnosed with hydronephrosis between June 2, 2019 and June 30, 2024 at the Sun Yat-Sen Memorial Hospital (SYSMH), Heyuan People's Hospital (HPH), and Ganzhou People's Hospital (GPH) was performed. Data on cases of hydronephrosis were extracted from the hospital's medical record system. The SYSMH cohort was randomly divided into two subsets: the SYSMH training set (n = 279) and the SYSMH validation set (n = 93) in a 3:1 ratio. The HPH cohort and GPH cohort serve as external validation sets. A hydronephrosis segmentation model (HRSM) was developed using the Improved U-Net algorithm, and the segmentation accuracy evaluated by the Dice Similarity Coefficient (DSC). Using 3D Convolutional Neural Network established an IH risk score (IHRS) based on segmented images. Independent risk clinical data for IH were screened by logistic regression. An IH diagnostic model (IHDM) was then developed, incorporating the IHRS and clinical data, using five machine learning algorithms (Random Forests, K-Nearest Neighbor, Decision Tree, Logistic Regression and Support Vector Machine). The diagnostic performance of the IHDM was assessed by the Receiver Operating Characteristic (ROC) curve.

Findings

The study initially included 1464 potential eligible cases, of which 864 were deemed qualified after preliminary examination. Ultimately, a total of 615 patients (363 female and 252 male) with hydronephrosis (including 5876 annotated cross-sectional CT images) were included in the study, 372 of whom were from SYSMH, 123 from HPH, and 120 from GPH. Based on bacterial culture results from percutaneous nephrostomy drainage of hydronephrosis, 291 cases were classified as IH, while 324 were non-IH. The DSC for the HRSM in the internal and two external validation cohorts were 0.922 (95% CI: 0.895, 0.949), 0.906 (95% CI: 0.869, 0.943), and 0.883 (95% CI: 0.857, 0.909), respectively, indicating high segmentation accuracy. The IHRS achieved a prediction accuracy of 78.5% (95% CI: 78.1%-78.9%) in the internal validation set. The IHDM developed using Support Vector Machine (SVM) combination with blood neutrophil count, fever within one week of history and IHRS performed best, demonstrated areas under the ROC curve of 0.919 (95% CI: 0.859-0.980), 0.902 (95% CI: 0.849-0.955), and 0.863 (95% CI: 0.800-0.926) in three cohorts, respectively.

Interpretation

The automated HRSM demonstrated excellent segmentation performance for hydronephrosis, while the non-invasive IHDM provided significant diagnostic efficacy, facilitating infection assessment in patients with hydronephrosis. However, more diverse real-world multicenter validation studies are needed to verify the robustness of the model before it can be incorporated into clinical practice.

Funding

The Key-Area Research and Development Program of Guangdong Province, and the National Natural Science Foundation of China.

Deep learning and machine learning in CT-based COPD diagnosis: Systematic review and meta-analysis

BACKGROUND

With advancements in medical technology and science, chronic obstructive pulmonary disease (COPD), one of the world's three major chronic diseases, has seen numerous remarkable outcomes when combined with artificial intelligence, particularly in disease diagnosis. However, the diagnostic performance of these AI models still lacks comprehensive evidence. Therefore, this study quantitatively analyzed the diagnostic performance of AI models in CT images of COPD patients, aiming to promote the development of related research in the future.

METHODS

PubMed, Cochrane Library, Web of Science, and Embase were retrieved up to September 1, 2024. The QUADAS-2 evaluation tool was used to assess the quality of the included studies. Meta-analysis of the included researches was performed using Stata18, RevMan 5.4, and Meta-Disc 1.4 software to merge sensitivity, specificity and plot a summary receiver operating characteristic curve (SROC). Heterogeneity was assessed using the Q statistic, and sources of inter-study heterogeneity were explored through meta-regression analysis.

RESULTS

Twenty-two of 3280 identified studies were eligible. Meta-analysis was performed on 15 of these studies, encompassing a total of 22,817 patients for which statistical metrics were reported or could be calculated. Seven studies were based on deep learning (DL) model, three on machine learning (ML) model, and five on DL model with multiple-instance learning (MIL) mechanisms. One study evaluated both DL and ML models. The meta-analysis results showed that the pooled sensitivity of all DL and ML models was 86 % (95 %CI 78-91 %), specificity was 87 % (95 %CI 83-91 %), and area under the curve was 93 % (95 %CI 90-95 %). Subgroup analyses revealed no significant difference in diagnostic sensitivity and specificity between DL and ML models (sensitivity 82 % (95 %CI 76-87 %), 93 % (95 %CI 85-97 %); specificity 87 % (95 %CI 79-91 %), 84 % (95 %CI 79-88 %), and the DL model with MIL (sensitivity 87 % (95 %CI 61-96 %); specificity 89 % (95 %CI 78-95 %) improved the performance of DL model, but this improvement was not statistically significant (p > 0.05).

CONCLUSION

Both DL and ML models for diagnosing COPD using CT images exhibited high accuracy. There was no significant difference in diagnostic efficacy between the two types of AI models, and the addition of the MIL mechanism may enhance the performance of the DL model.

Association of the characteristics of brain magnetic resonance imaging with genes related to disease onset in schizophrenia patients

BACKGROUND

Schizophrenia (SCH) is a complex neurodevelopmental disorder, whose pathogenesis is not fully elucidated. This article aims to reveal disease-specific brain structural and functional changes and their potential genetic basis by analyzing the characteristics of brain magnetic resonance imaging (MRI) in SCH patients and related gene expression patterns.

METHODS

Differentially expressed genes (DEGs) between SCH and healthy control (NC) groups in the GSE48072 dataset were identified and functionally analyzed, and a protein-protein interaction (PPI) network was fabricated to screen for core genes (CGs). Meanwhile, MRI data from the COBRE, the Human Connectome Project (HCP), the 1000 Functional Connectomes Project (FCP), and the Consortium for Reliability and Reproducibility (CoRR) were utilized to explore differences in brain activity patterns between SCH patients and NC group using a 3D deep aggregation network (3D DANet) machine learning approach. A correlation analysis was performed between the identified CGs and MRI imaging characteristics.

RESULTS

82 DEGs were collected from the GSE48072 dataset, primarily involved in cytotoxic granules, growth factor binding, and graft-versus-host disease pathways. The construction of the PPI network revealed KLRD1, KLRF1, CD244, GZMH, GZMA, GZMB, PRF1, and SLAMF6 as CGs. SCH patients exhibited relatively enhanced activity patterns in the frontoparietal attention network (FAN) and default mode network (DMN) across four datasets, while showing a trend of weakening in most other networks. The 3D DANet demonstrated higher accuracy, specificity, and sensitivity in brain image classification. The correlation between enhancement of the DMN and genetic abnormalities was the strongest, followed by the enhancement of the frontal and parietal attention networks. In contrast, the correlation between the weakening of the sensory-motor network and occipital network and genetic abnormalities was relatively weak. The strongest correlation was observed between MRI characteristics and the KLRD1 and CD244 genes.

CONCLUSION

The granzyme-mediated programmed cell death signaling pathway is related to pathogenesis of SCH, and CD244 may serve as potential biological markers for diagnosing SCH. The correlation between enhancement of the DMN and genetic abnormalities was the strongest, followed by the enhancement of the frontal and parietal attention networks. In contrast, the correlation between weakening of the sensory-motor network and occipital network and genetic abnormalities was relatively weak. Additionally, the strongest correlation was observed between MRI features and the KLRD1 and CD244 genes. The use of the 3D DANet method has improved the detection precision of brain structural and functional changes in SCH patients, providing a new perspective for understanding the biological basis of the disease.

Optimizing convolutional neural networks for Chronic Obstructive Pulmonary Disease detection in clinical computed tomography imaging

We aim to optimize the binary detection of Chronic Obstructive Pulmonary Disease (COPD) based on emphysema presence in the lung with convolutional neural networks (CNN) by exploring manually adjusted versus automated window-setting optimization (WSO) on computed tomography (CT) images. 7194 contrast-enhanced CT images (3597 with COPD; 3597 healthy controls) from 78 subjects were selected retrospectively (01.2018-12.2021) and preprocessed. For each image, intensity values were manually clipped to the emphysema window setting and a baseline 'full-range' window setting. Class-balanced train, validation, and test sets contained 3392, 1114, and 2688 images. The network backbone was optimized by comparing various CNN architectures. Furthermore, automated WSO was implemented by adding a customized layer to the model. The image-level area under the Receiver Operating Characteristics curve (AUC) [lower, upper limit 95% confidence] was utilized to compare model variations. Repeated inference (n = 7) on the test set showed that the DenseNet was the most efficient backbone and achieved a mean AUC of 0.80 [0.76, 0.85] without WSO. Comparably, with input images manually adjusted to the emphysema window, the DenseNet model predicted COPD with a mean AUC of 0.86 [0.82, 0.89]. By adding a customized WSO layer to the DenseNet, an optimal window in the proximity of the emphysema window setting was learned automatically, and a mean AUC of 0.82 [0.78, 0.86] was achieved. Detection of COPD with DenseNet models was improved by WSO of CT data to the emphysema window setting range.

Evaluating the Cumulative Benefit of Inspiratory CT, Expiratory CT, and Clinical Data for COPD Diagnosis and Staging through Deep Learning

Purpose To measure the benefit of single-phase CT, inspiratory-expiratory CT, and clinical data for convolutional neural network (CNN)-based chronic obstructive pulmonary disease (COPD) staging. Materials and Methods This retrospective study included inspiratory and expiratory lung CT images and spirometry measurements acquired between November 2007 and April 2011 from 8893 participants (mean age, 59.6 years ± 9.0 [SD]; 53.3% [4738 of 8893] male) in the COPDGene phase I cohort (ClinicalTrials.gov: NCT00608764). CNNs were trained to predict spirometry measurements (forced expiratory volume in 1 second [FEV1], FEV1 percent predicted, and ratio of FEV1 to forced vital capacity [FEV1/FVC]) using clinical data and either single-phase or multiphase CT. Spirometry predictions were then used to predict Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage. Agreement between CNN-predicted and reference standard spirometry measurements and GOLD stage was assessed using intraclass correlation coefficient (ICC) and compared using bootstrapping. Accuracy for predicting GOLD stage, within-one GOLD stage, and GOLD 0 versus 1-4 was calculated. Results CNN-predicted and reference standard spirometry measurements showed moderate to good agreement (ICC, 0.66-0.79), which improved by inclusion of clinical data (ICC, 0.70-0.85; P ≤ .04), except for FEV1/FVC in the inspiratory-phase CNN model with clinical data (P = .35) and FEV1 in the expiratory-phase CNN model with clinical data (P = .33). Single-phase CNN accuracies for GOLD stage, within-one stage, and diagnosis ranged from 59.8% to 84.1% (682-959 of 1140), with moderate to good agreement (ICC, 0.68-0.70). Accuracies of CNN models using inspiratory and expiratory images ranged from 60.0% to 86.3% (684-984 of 1140), with moderate to good agreement (ICC, 0.72). Inclusion of clinical data improved agreement and accuracy for both the single-phase CNNs (ICC, 0.72; P ≤ .001; accuracy, 65.2%-85.8% [743-978 of 1140]) and inspiratory-expiratory CNNs (ICC, 0.77-0.78; P ≤ .001; accuracy, 67.6%-88.0% [771-1003 of 1140]), except expiratory CNN with clinical data (no change in GOLD stage ICC; P = .08). Conclusion CNN-based COPD diagnosis and staging using single-phase CT provides comparable accuracy with inspiratory-expiratory CT when provided clinical data relevant to staging. Keywords: Convolutional Neural Network, Chronic Obstructive Pulmonary Disease, CT, Severity Staging, Attention Map Supplemental material is available for this article. © RSNA, 2024.

Advancements in automated classification of chronic obstructive pulmonary disease based on computed tomography imaging features through deep learning approaches

Chronic Obstructive Pulmonary Disease (COPD) represents a global public health issue that significantly impairs patients' quality of life and overall health. As one of the primary causes of chronic respiratory diseases and global mortality, effective diagnosis and classification of COPD are crucial for clinical management. Pulmonary function tests (PFTs) are standard for diagnosing COPD, yet their accuracy is influenced by patient compliance and other factors, and they struggle to detect early disease pathologies. Furthermore, the complexity of COPD pathological changes poses additional challenges for clinical diagnosis, increasing the difficulty for physicians in practice. Recently, deep learning (DL) technologies have demonstrated significant potential in medical image analysis, particularly for the diagnosis and classification of COPD. By analyzing key radiological features such as airway alterations, emphysema, and vascular characteristics in Computed Tomography (CT) scan images, DL enhances diagnostic accuracy and efficiency, providing more precise treatment plans for COPD patients. This article reviews the latest research advancements in DL methods based on principal radiological features of COPD for its classification and discusses the advantages, challenges, and future research directions of DL in this field, aiming to provide new perspectives for the personalized management and treatment of COPD.

BranchLabelNet: Anatomical Human Airway Labeling Approach using a Dividing-and-Grouping Multi-Label Classification

Anatomical airway labeling is crucial for precisely identifying airways displaying symptoms such as constriction, increased wall thickness, and modified branching patterns, facilitating the diagnosis and treatment of pulmonary ailments. This study introduces an innovative airway labeling methodology, BranchLabelNet, which accounts for the fractal nature of airways and inherent hierarchical branch nomenclature. In developing this methodology, branch-related parameters, including position vectors, generation levels, branch lengths, areas, perimeters, and more, are extracted from a dataset of 1000 chest computed tomography (CT) images. To effectively manage this intricate branch data, we employ an n-ary tree structure that captures the complicated relationships within the airway tree. Subsequently, we employ a divide-and-group deep learning approach for multi-label classification, streamlining the anatomical airway branch labeling process. Additionally, we address the challenge of class imbalance in the dataset by incorporating the Tomek Links algorithm to maintain model reliability and accuracy. Our proposed airway labeling method provides robust branch designations and achieves an impressive average classification accuracy of 95.94% across fivefold cross-validation. This approach is adaptable for addressing similar complexities in general multi-label classification problems within biomedical systems.

Detection of chronic obstructive pulmonary disease with deep learning using inspiratory and expiratory chest computed tomography and clinical information

Background

In recent years, more and more patients with chronic obstructive pulmonary disease (COPD) have remained undiagnosed despite having undergone medical examination. This study aimed to develop a convolutional neural network (CNN) model for automatically detecting COPD using double-phase (inspiratory and expiratory) chest computed tomography (CT) images and clinical information.

Methods

A total of 2,047 participants, including never-smokers, ex-smokers, and current smokers, were prospectively recruited from three hospitals. The double-phase CT images and clinical information of each participant were collected for training the proposed CNN model which integrated a sequence of residual feature extracting blocks network (RFEBNet) for extracting CT image features and a fully connected feed-forward network (FCNet) for extracting clinical features. In addition, the RFEBNet utilizing double- or single-phase CT images and the FCNet using clinical information were conducted for comparison.

Results

The proposed CNN model, which utilized double-phase CT images and clinical information, outperformed other models in detecting COPD with an area under the receiver operating characteristic curve (AUC) of 0.930 [95% confidence interval (CI): 0.913-0.951] on an internal test set (n=307). The AUC was higher than the RFEBNet using double-phase CT images (AUC =0.912, 95% CI: 0.891-0.932), single inspiratory CT images (AUC =0.888, 95% CI: 0.863-0.915), single expiratory CT images (AUC =0.897, 95% CI: 0.874-0.925), and FCNet using clinical information (AUC =0.805, 95% CI: 0.777-0.841). The proposed model also achieved the best performance on an external test (n=516) with an AUC of 0.896 (95% CI: 0.871-0.931).

Conclusions

The proposed CNN model using double-phase CT images and clinical information can automatically detect COPD with high accuracy.

Artificial intelligence in COPD CT images: identification, staging, and quantitation

Chronic obstructive pulmonary disease (COPD) stands as a significant global health challenge, with its intricate pathophysiological manifestations often demanding advanced diagnostic strategies. The recent applications of artificial intelligence (AI) within the realm of medical imaging, especially in computed tomography, present a promising avenue for transformative changes in COPD diagnosis and management. This review delves deep into the capabilities and advancements of AI, particularly focusing on machine learning and deep learning, and their applications in COPD identification, staging, and imaging phenotypes. Emphasis is laid on the AI-powered insights into emphysema, airway dynamics, and vascular structures. The challenges linked with data intricacies and the integration of AI in the clinical landscape are discussed. Lastly, the review casts a forward-looking perspective, highlighting emerging innovations in AI for COPD imaging and the potential of interdisciplinary collaborations, hinting at a future where AI doesn't just support but pioneers breakthroughs in COPD care. Through this review, we aim to provide a comprehensive understanding of the current state and future potential of AI in shaping the landscape of COPD diagnosis and management.

COPD stage detection: leveraging the auto-metric graph neural network with inspiratory and expiratory chest CT images

Chronic obstructive pulmonary disease (COPD) is a common lung disease that can lead to restricted airflow and respiratory problems, causing a significant health, economic, and social burden. Detecting the COPD stage can provide a timely warning for prompt intervention in COPD patients. However, existing methods based on inspiratory (IN) and expiratory (EX) chest CT images are not sufficiently accurate and efficient in COPD stage detection. The lung region images are autonomously segmented from IN and EX chest CT images to extract the 1 , 781 × 2 lung radiomics and 13 , 824 × 2 3D CNN features. Furthermore, a strategy for concatenating and selecting features was employed in COPD stage detection based on radiomics and 3D CNN features. Finally, we combine all the radiomics, 3D CNN features, and factor risks (age, gender, and smoking history) to detect the COPD stage based on the Auto-Metric Graph Neural Network (AMGNN). The AMGNN with radiomics and 3D CNN features achieves the best performance at 89.7 % of accuracy, 90.9 % of precision, 89.5 % of F1-score, and 95.8 % of AUC compared to six classic machine learning (ML) classifiers. Our proposed approach demonstrates high accuracy in detecting the stage of COPD using both IN and EX chest CT images. This method can potentially establish an efficient diagnostic tool for patients with COPD. Additionally, we have identified radiomics and 3D CNN as more appropriate biomarkers than Parametric Response Mapping (PRM). Moreover, our findings indicate that expiration yields better results than inspiration in detecting the stage of COPD.

Anatomically aware dual-hop learning for pulmonary embolism detection in CT pulmonary angiograms

Pulmonary Embolisms (PE) represent a leading cause of cardiovascular death. While medical imaging, through computed tomographic pulmonary angiography (CTPA), represents the gold standard for PE diagnosis, it is still susceptible to misdiagnosis or significant diagnosis delays, which may be fatal for critical cases. Despite the recently demonstrated power of deep learning to bring a significant boost in performance in a wide range of medical imaging tasks, there are still very few published researches on automatic pulmonary embolism detection. Herein we introduce a deep learning based approach, which efficiently combines computer vision and deep neural networks for pulmonary embolism detection in CTPA. Our method brings novel contributions along three orthogonal axes: (1) automatic detection of anatomical structures; (2) anatomical aware pretraining, and (3) a dual-hop deep neural net for PE detection. We obtain state-of-the-art results on the publicly available multicenter large-scale RSNA dataset.

Human-airway surface mesh smoothing based on graph convolutional neural networks

BACKGROUND AND OBJECTIVE

A detailed representation of the airway geometry in the respiratory system is critical for predicting precise airflow and pressure behaviors in computed tomography (CT)-image-based computational fluid dynamics (CFD). The CT-image-based geometry often contains artifacts, noise, and discontinuities due to the so-called stair step effect. Hence, an advanced surface smoothing is necessary. The existing smoothing methods based on the Laplacian operator drastically shrink airway geometries, resulting in the loss of information related to smaller branches. This study aims to introduce an unsupervised airway-mesh-smoothing learning (AMSL) method that preserves the original geometry of the three-dimensional (3D) airway for accurate CT-image-based CFD simulations.

METHOD

The AMSL method jointly trains two graph convolutional neural networks (GCNNs) defined on airway meshes to filter vertex positions and face normal vectors. In addition, it regularizes a combination of loss functions such as reproducibility, smoothness and consistency of vertex positions, and normal vectors. The AMSL adopts the concept of a deep mesh prior model, and it determines the self-similarity for mesh restoration without using a large dataset for training. Images of the airways of 20 subjects were smoothed by the AMSL method, and among them, the data of two subjects were used for the CFD simulations to assess the effect of airway smoothing on flow properties.

RESULTS

In 18 of 20 benchmark problems, the proposed smoothing method delivered better results compared with the conventional or state-of-the-art deep learning methods. Unlike the traditional smoothing, the AMSL successfully constructed 20 smoothed airways with airway diameters that were consistent with the original CT images. Besides, CFD simulations with the airways obtained by the AMSL method showed much smaller pressure drop and wall shear stress than the results obtained by the traditional method.

CONCLUSIONS

The airway model constructed by the AMSL method reproduces branch diameters accurately without any shrinkage, especially in the case of smaller airways. The accurate estimation of airway geometry using a smoothing method is critical for estimating flow properties in CFD simulations.

The role of chest computed tomography in the evaluation and management of chronic obstructive pulmonary disease

PURPOSE OF REVIEW

The purpose of this review is to compile recent data on the clinical associations of computed tomography (CT) scan findings in the literature and potential avenues for implementation into clinical practice.

RECENT FINDINGS

Airways dysanapsis, emphysema, chronic bronchitis, and pulmonary vascular metrics have all recently been associated with poor chronic obstructive pulmonary disease (COPD) outcomes when controlled for clinically relevant covariables, including risk of mortality in the case of emphysema and chronic bronchitis. Other authors suggest that CT scan may provide insight into both lung parenchymal damage and other clinically important comorbidities in COPD.

SUMMARY

CT scan findings in COPD relate to clinical outcomes. There is a continued need to develop processes to best implement the results of these studies into clinical practice.

[The contribution of artificial intelligence (AI) subsequent to the processing of thoracic imaging]

The contribution of artificial intelligence (AI) to medical imaging is currently the object of widespread experimentation. The development of deep learning (DL) methods, particularly convolution neural networks (CNNs), has led to performance gains often superior to those achieved by conventional methods such as machine learning. Radiomics is an approach aimed at extracting quantitative data not accessible to the human eye from images expressing a disease. The data subsequently feed machine learning models and produce diagnostic or prognostic probabilities. As for the multiple applications of AI methods in thoracic imaging, they are undergoing evaluation. Chest radiography is a practically ideal field for the development of DL algorithms able to automatically interpret X-rays. Current algorithms can detect up to 14 different abnormalities present either in isolation or in combination. Chest CT is another area offering numerous AI applications. Various algorithms have been specifically formed and validated for the detection and characterization of pulmonary nodules and pulmonary embolism, as well as segmentation and quantitative analysis of the extent of diffuse lung diseases (emphysema, infectious pneumonias, interstitial lung disease). In addition, the analysis of medical images can be associated with clinical, biological, and functional data (multi-omics analysis), the objective being to construct predictive approaches regarding disease prognosis and response to treatment.

Application of 3D computed tomography in emphysematous parenchyma patients scheduled for bronchoscopic lung volume reduction

Bronchoscopic lung volume reduction (BLVR) is a feasible, safe, effective and minimally invasive technique to significantly improve the quality of life of advanced severe chronic obstructive pulmonary disease (COPD). In this study, three-dimensional computed tomography (3D-CT) automatic analysis software combined with pulmonary function test (PFT) was used to retrospectively evaluate the postoperative efficacy of BLVR patients. The purpose is to evaluate the improvement of lung function of local lung tissue after operation, maximize the benefits of patients, and facilitate BLVR in the treatment of patients with advanced COPD. All the reported cases of advanced COPD patients treated with BLVR with one-way valve were collected and analysed from 2017 to 2020. Three-dimensional-CT image analysis software system was used to analyse the distribution of low-density areas <950 Hounsfield units in both lungs pre- and post- BLVR. Meanwhile, all patients performed standard PFT pre- and post-operation for retrospective analysis. We reported six patients that underwent unilateral BLVR with 1 to 3 valves according to the range of emphysema. All patients showed a median increase in forced expiratory volume in 1 second (FEV1) of 34%, compared with baseline values. Hyperinflation was reduced by 16.6% (range, 4.9%-47.2%). The volumetric measurements showed a significant reduction in the treated lobe volume among these patients. Meanwhile, the targeted lobe volume changes were inversely correlated with change in FEV1/FEV1% in patients with heterogeneous emphysematous. We confirm that 3D-CT analysis can quantify the changes of lung volume, ventilation and perfusion, to accurately evaluate the distribution and improvement of emphysema and rely less on the observer.

Deep learning parametric response mapping from inspiratory chest CT scans: a new approach for small airway disease screening

OBJECTIVES

Parametric response mapping (PRM) enables the evaluation of small airway disease (SAD) at the voxel level, but requires both inspiratory and expiratory chest CT scans. We hypothesize that deep learning PRM from inspiratory chest CT scans can effectively evaluate SAD in individuals with normal spirometry.

METHODS

We included 537 participants with normal spirometry, a history of smoking or secondhand smoke exposure, and divided them into training, tuning, and test sets. A cascaded generative adversarial network generated expiratory CT from inspiratory CT, followed by a UNet-like network predicting PRM using real inspiratory CT and generated expiratory CT. The performance of the prediction is evaluated using SSIM, RMSE and dice coefficients. Pearson correlation evaluated the correlation between predicted and ground truth PRM. ROC curves evaluated predicted PRMfSAD (the volume percentage of functional small airway disease, fSAD) performance in stratifying SAD.

RESULTS

Our method can generate expiratory CT of good quality (SSIM 0.86, RMSE 80.13 HU). The predicted PRM dice coefficients for normal lung, emphysema, and fSAD regions are 0.85, 0.63, and 0.51, respectively. The volume percentages of emphysema and fSAD showed good correlation between predicted and ground truth PRM (|r| were 0.97 and 0.64, respectively, p < 0.05). Predicted PRMfSAD showed good SAD stratification performance with ground truth PRMfSAD at thresholds of 15%, 20% and 25% (AUCs were 0.84, 0.78, and 0.84, respectively, p < 0.001).

CONCLUSION

Our deep learning method generates high-quality PRM using inspiratory chest CT and effectively stratifies SAD in individuals with normal spirometry.

Development of machine learning model to predict pulmonary function with low-dose CT-derived parameter response mapping in a community-based chest screening cohort

PURPOSE

To construct and evaluate the performance of a machine learning-based low dose computed tomography (LDCT)-derived parametric response mapping (PRM) model for predicting pulmonary function test (PFT) results.

MATERIALS AND METHODS

A total of 615 subjects from a community-based screening population (40-74 years old) with PFT parameters, including the ratio of the first second forced expiratory volume to forced vital capacity (FEV1/FVC), the percentage of forced expiratory volume in the one second predicted (FEV1%), and registered inspiration-to-expiration chest CT scanning were enrolled retrospectively. Subjects were classified into a normal, high risk, and COPD group based on PFT. Data of 72 PRM-derived quantitative parameters were collected, including volume and volume percentage of emphysema, functional-small airways disease, and normal lung tissue. A machine-learning with random forest regression model and a multilayer perceptron (MLP) model were constructed and tested on PFT prediction, which was followed by evaluation of classification performance based on the PFT predictions.

RESULTS

The machine-learning model based on PRM parameters showed better performance for predicting PFT than MLP, with a coefficient of determination (R2 ) of 0.749 and 0.792 for FEV1/FVC and FEV1%, respectively. The Mean Squared Errors (MSE) for FEV1/FVC and FEV1% are 0.0030 and 0.0097 for the random forest model, respectively. The Root Mean Squared Errors (RMSE) for FEV1/FVC and FEV1% are 0.055 and 0.098, respectively. The sensitivity, specificity, and accuracy for differentiating between the normal group and high-risk group were 34/40 (85%), 65/72 (90%), and 99/112 (88%), respectively. For differentiating between the non-COPD group and COPD group, the sensitivity, specificity, and accuracy were 8/9 (89%), 112/112 (100%), 120/121 (99%), respectively.

CONCLUSIONS

The machine learning-based random forest model predicts PFT results in a community screening population based on PRM, and it identifies high risk COPD from normal populations with high sensitivity and reliably predicts of high-risk COPD.

Machine Learning-Assisted Diagnosis Model for Chronic Obstructive Pulmonary Disease

Chronic obstructive pulmonary disease (COPD) is a long-term, irreversible, and progressive respiratory disease that often leads to lung function decline. Pulmonary function tests (PFTs) provide valuable information for diagnosing COPD; however, they are underutilised in clinical practice, with only a subset of test values being used for decision making. The final clinical diagnosis requires combining PFT results with patient information, symptoms, and other tests, such as imaging and blood analysis. This study aims to comprehensively utilise all the testing information in PFTs to assist in the diagnosis of COPD. Various machine learning models, such as logistic regression, support vector machine (SVM), k-nearest neighbour (KNN), random forest, decision tree, and XGBoost, have been employed to establish COPD diagnosis assistance models. The XGBoost model, trained with features extracted by the group LASSO algorithm, achieved the best performance, with an area under the receiver operating characteristic curve (ROC) of 0.90, 88.6% accuracy, and 98.5% sensitivity. This model can assist doctors in the clinical diagnosis and early prediction of COPD.

Radiomics for Improved Detection of Chronic Obstructive Pulmonary Disease in Low-Dose and Standard-Dose Chest CT Scans

Background Approximately half of adults with chronic obstructive pulmonary disease (COPD) remain undiagnosed. Chest CT scans are frequently acquired in clinical practice and present an opportunity to detect COPD. Purpose To assess the performance of radiomics features in COPD diagnosis using standard-dose and low-dose CT models. Materials and Methods This secondary analysis included participants enrolled in the Genetic Epidemiology of COPD, or COPDGene, study at baseline (visit 1) and 10 years after baseline (visit 3). COPD was defined by a forced expiratory volume in the 1st second of expiration to forced vital capacity ratio less than 0.70 at spirometry. The performance of demographics, CT emphysema percentage, radiomics features, and a combined feature set derived from inspiratory CT alone was evaluated. CatBoost (Yandex), a gradient boosting algorithm, was used to perform two classification experiments to detect COPD; the two models were trained and tested on standard-dose CT data from visit 1 (model I) and low-dose CT data from visit 3 (model II). Classification performance of the models was evaluated using area under the receiver operating characteristic curve (AUC) and precision-recall curve analysis. Results A total of 8878 participants (mean age, 57 years ± 9 [SD]; 4180 female, 4698 male) were evaluated. Radiomics features in model I achieved an AUC of 0.90 (95% CI: 0.88, 0.91) in the standard-dose CT test cohort versus demographics (AUC, 0.73; 95% CI: 0.71, 0.76; P < .001), emphysema percentage (AUC, 0.82; 95% CI 0.80, 0.84; P < .001), and combined features (AUC, 0.90; 95% CI: 0.89, 0.92; P = .16). Model II, trained on low-dose CT scans, achieved an AUC of 0.87 (95% CI: 0.83, 0.91) on the 20% held-out test set for radiomics features compared with demographics (AUC, 0.70; 95% CI: 0.64, 0.75; P = .001), emphysema percentage (AUC, 0.74; 95% CI: 0.69, 0.79; P = .002), and combined features (AUC, 0.88; 95% CI: 0.85, 0.92; P = .32). Density and texture features were the majority of the top 10 features in the standard-dose model, whereas shape features of lungs and airways were significant contributors in the low-dose CT model. Conclusion A combination of features representing parenchymal texture and lung and airway shape on inspiratory CT scans can be used to accurately detect COPD. ClinicalTrials.gov registration no. NCT00608764 © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Vliegenthart in this issue.

An unsupervised image registration method employing chest computed tomography images and deep neural networks

BACKGROUND

Deformable image registration is crucial for multiple radiation therapy applications. Fast registration of computed tomography (CT) lung images is challenging because of the large and nonlinear deformation between inspiration and expiration. With advancements in deep learning techniques, learning-based registration methods are considered efficient alternatives to traditional methods in terms of accuracy and computational cost.

METHOD

In this study, an unsupervised lung registration network (LRN) with cycle-consistent training is proposed to align two acquired CT-derived lung datasets during breath-holds at inspiratory and expiratory levels without utilizing any ground-truth registration results. Generally, the LRN model uses three loss functions: image similarity, regularization, and Jacobian determinant. Here, LRN was trained on the CT datasets of 705 subjects and tested using 10 pairs of public CT DIR-Lab datasets. Furthermore, to evaluate the effectiveness of the registration technique, target registration errors (TREs) of the LRN model were compared with those of the conventional algorithm (sum of squared tissue volume difference; SSTVD) and a state-of-the-art unsupervised registration method (VoxelMorph).

RESULTS

The results showed that the LRN with an average TRE of 1.78 ± 1.56 mm outperformed VoxelMorph with an average TRE of 2.43 ± 2.43 mm, which is comparable to that of SSTVD with an average TRE of 1.66 ± 1.49 mm. In addition, estimating the displacement vector field without any folding voxel consumed less than 2 s, demonstrating the superiority of the learning-based method with respect to fiducial marker tracking and the overall soft tissue alignment with a nearly real-time speed.

CONCLUSIONS

Therefore, this proposed method shows significant potential for use in time-sensitive pulmonary studies, such as lung motion tracking and image-guided surgery.

CT-based COPD identification using multiple instance learning with two-stage attention

BACKGROUND AND OBJECTIVE

Chronic obstructive pulmonary disease (COPD) is one of the leading causes of morbidity and mortality worldwide. However, COPD remains underdiagnosed globally. Spirometry is currently the primary tool for diagnosing COPD, but it has unneglected difficulties in detecting mild COPD. Chest computed tomography (CT) has been validated for COPD diagnosis and quantification. Whereas many CT-based deep learning approaches have been developed to identify COPD, it remains challenging to characterize CT-based pathological alternations of COPD which are multidimensional and highly spatially heterogeneous, and the diagnosis performance still needs to be improved.

METHODS

A multiple instance learning (MIL) with two-stage attention (TSA-MIL) is proposed to identify COPD using CT images. Based on transfer learning, a Resnet-50 model pre-trained on natural images is used to extract multicomponent and multidimensional features of COPD abnormalities, in which a pseudo-color method is designed to transfer single-channel CT slices to RGB-like three channels and meanwhile increase the richness of feature representations. To generate more robust attention score for each instance, a two-stage attention module is utilized with the first stage aiming at discovering the key instance while the second stage correcting the attention score for each instance by calculating its average relative distance to the key instances; besides, an instance-level clustering over feature domain is exploited to further improve feature separability and therefore facilitate the subsequent attention module. CT scans, spirometry and demographic data of a total of 800 participants were collected from a large public hospital, with 720 and 80 participants used for model development and evaluation, respectively. In addition, data of 260 participants from another large hospital were also collected for external validation.

RESULTS AND CONCLUSIONS

The proposed TSA-MIL approach outperforms not only most of the advanced MIL models, but also other up-to-date COPD identification methods, with an accuracy of 0.9200 and an area under curve (AUC) of 0.9544 on the test set, and with an accuracy of 0.8115 and an AUC of 0.8737 on the external validation set without multicenter effect reduction, which is clinically acceptable. Therefore, this approach is promising to be a powerful tool for COPD diagnosis in clinical practice.

Detection of abnormal extraocular muscles in small datasets of computed tomography images using a three-dimensional variational autoencoder

We sought to establish an unsupervised algorithm with a three-dimensional (3D) variational autoencoder model (VAE) for the detection of abnormal extraocular muscles in small datasets of orbital computed tomography (CT) images. 334 CT images of normal orbits and 96 of abnormal orbits diagnosed as thyroid eye disease were used for training and validation; 24 normal and 11 abnormal orbits were used for the test. A 3D VAE was developed and trained. All images were preprocessed to emphasize extraocular muscles and to suppress background noise (e.g., high signal intensity from bones). The optimal cut-off value was identified through receiver operating characteristic (ROC) curve analysis. The ability of the model to detect muscles of abnormal size was assessed by visualization. The model achieved a sensitivity of 79.2%, specificity of 72.7%, accuracy of 77.1%, F1-score of 0.667, and AUROC of 0.801. Abnormal CT images correctly identified by the model showed differences in the reconstruction of extraocular muscles. The proposed model showed potential to detect abnormalities in extraocular muscles using a small dataset, similar to the diagnostic approach used by physicians. Unsupervised learning could serve as an alternative detection method for medical imaging studies in which annotation is difficult or impossible to perform.

A Computer-based Analysis for Identification and Quantification of Small Airway Disease in Lung Computed Tomography Images: A Comprehensive Review for Radiologists

Computed tomography (CT) imaging is being increasingly used in clinical practice for detailed characterization of lung diseases. Respiratory diseases involve various components of the lung, including the small airways. Evaluation of small airway disease on CT images is challenging as the airways cannot be visualized directly by a CT scanner. Small airway disease can manifest as pulmonary air trapping (AT). Although AT may be sometimes seen as mosaic attenuation on expiratory CT images, it is difficult to identify diffuse AT visually. Computer technology advances over the past decades have provided methods for objective quantification of small airway disease on CT images. Quantitative CT (QCT) methods are being rapidly developed to quantify underlying lung diseases with greater precision than subjective visual assessment of CT images. A growing body of evidence suggests that QCT methods can be practical tools in the clinical setting to identify and quantify abnormal regions of the lung accurately and reproducibly. This review aimed to describe the available methods for the identification and quantification of small airway disease on CT images and to discuss the challenges of implementing QCT metrics in clinical care for patients with small airway disease.

A Survey on AI Techniques for Thoracic Diseases Diagnosis Using Medical Images

Thoracic diseases refer to disorders that affect the lungs, heart, and other parts of the rib cage, such as pneumonia, novel coronavirus disease (COVID-19), tuberculosis, cardiomegaly, and fracture. Millions of people die every year from thoracic diseases. Therefore, early detection of these diseases is essential and can save many lives. Earlier, only highly experienced radiologists examined thoracic diseases, but recent developments in image processing and deep learning techniques are opening the door for the automated detection of these diseases. In this paper, we present a comprehensive review including: types of thoracic diseases; examination types of thoracic images; image pre-processing; models of deep learning applied to the detection of thoracic diseases (e.g., pneumonia, COVID-19, edema, fibrosis, tuberculosis, chronic obstructive pulmonary disease (COPD), and lung cancer); transfer learning background knowledge; ensemble learning; and future initiatives for improving the efficacy of deep learning models in applications that detect thoracic diseases. Through this survey paper, researchers may be able to gain an overall and systematic knowledge of deep learning applications in medical thoracic images. The review investigates a performance comparison of various models and a comparison of various datasets.

Structural and functional alterations of subjects with cement dust exposure: A longitudinal quantitative computed tomography-based study

Cement dust exposure (CDE) can be a risk factor for pulmonary disease, causing changes in segmental airways and parenchymal lungs. This study investigates longitudinal alterations in quantitative computed tomography (CT)-based metrics due to CDE. We obtained CT-based airway structural and lung functional metrics from CDE subjects with baseline CT and follow-up CT scans performed three years later. From the CT, we extracted wall thickness (WT) and bifurcation angle (θ) at total lung capacity (TLC) and functional residual capacity (FRC), respectively. We also computed air volume (V_air), tissue volume (V_tissue), global lung shape, percentage of emphysema (Emph%), and more. Clinical measures were used to associate with CT-based metrics. Three years after their baseline, the pulmonary function tests of CDE subjects were similar or improved, but there were significant alterations in the CT-based structural and functional metrics. The follow-up CT scans showed changes in θ at most of the central airways; increased WT at the subgroup bronchi; smaller V_air at TLC at all except the right upper and lower lobes; smaller V_tissue at all lobes in TLC and FRC except for the upper lobes in FRC; smaller global lung shape; and greater Emph% at the right upper and lower lobes. CT-based structural and functional variables are more sensitive to the early identification of CDE subjects, while most clinical lung function changes were not noticeable. We speculate that the significant long-term changes in CT are uniquely observed in CDE subjects, different from smoking-induced structural changes.

Early detection of COPD based on graph convolutional network and small and weakly labeled data

Chronic obstructive pulmonary disease (COPD) is a common disease with high morbidity and mortality, where early detection benefits the population. However, the early diagnosis rate of COPD is low due to the absence or slight early symptoms. In this paper, a novel method based on graph convolution network (GCN) for early detection of COPD is proposed, which uses small and weakly labeled chest computed tomography image data from the publicly available Danish Lung Cancer Screening Trial database. The key idea is to construct a graph using regions of interest randomly selected from the segmented lung parenchyma and then input it into the GCN model for COPD detection. In this way, the model can not only extract the feature information of each region of interest but also the topological structure information between regions of interest, that is, graph structure information. The proposed GCN model achieves an acceptable performance with an accuracy of 0.77 and an area under a curve of 0.81, which is higher than the previous studies on the same dataset. GCN model also outperforms several state-of-the-art methods trained at the same time. As far as we know, it is also the first time using the GCN model on this dataset for COPD detection.

Quantitative computed tomography imaging-based classification of cement dust-exposed subjects with an artificial neural network technique

BACKGROUND AND OBJECTIVE

Cement dust exposure is likely to affect the structural and functional alterations in segmental airways and parenchymal lungs. This study develops an artificial neural network (ANN) model for identifying cement dust-exposed (CDE) subjects using quantitative computed tomography-based airway structural and functional features.

METHODS

We obtained the airway features in five central and five sub-grouped segmental regions and the lung features in five lobar regions and one total lung region from 311 CDE and 298 non-CDE (NCDE) subjects. The five-fold cross-validation method was adopted to train the following classification models:ANN, support vector machine (SVM), logistic regression (LR), and decision tree (DT). For all the classification models, linear discriminant analysis (LDA) and genetic algorithm (GA) were applied for dimensional reduction and hyperparameterization, respectively. The ANN model without LDA was also optimized by the GA method to observe the effect of the dimensional reduction.

RESULTS

The genetically optimized ANN model without the LDA method was the best in terms of the classification accuracy. The accuracy, sensitivity, and specificity of the GA-ANN model with four layers were greater than those of the other classification models (i.e., ANN, SVM, LR, and DT using LDA and GA methods) in the five-fold cross-validation. The average values of accuracy, sensitivity, and specificity for the five-fold cross-validation were 97.0%, 98.7%, and 98.6%, respectively.

CONCLUSIONS

We demonstrated herein that a quantitative computed tomography-based ANN model could more effectively detect CDE subjects when compared to their counterpart models. By employing the model, the CDE subjects may be identified early for therapeutic intervention.

Application of Various YOLO Models for Computer Vision-Based Real-Time Pothole Detection

Pothole repair is one of the paramount tasks in road maintenance. Effective road surface monitoring is an ongoing challenge to the management agency. The current pothole detection, which is conducted image processing with a manual operation, is labour-intensive and time-consuming. Computer vision offers a mean to automate its visual inspection process using digital imaging, hence, identifying potholes from a series of images. The goal of this study is to apply different YOLO models for pothole detection. Three state-of-the-art object detection frameworks (i.e., YOLOv4, YOLOv4-tiny, and YOLOv5s) are experimented to measure their performance involved in real-time responsiveness and detection accuracy using the image set. The image set is identified by running the deep convolutional neural network (CNN) on several deep learning pothole detectors. After collecting a set of 665 images in 720 × 720 pixels resolution that captures various types of potholes on different road surface conditions, the set is divided into training, testing, and validation subsets. A mean average precision at 50% Intersection-over-Union threshold (mAP_0.5) is used to measure the performance of models. The study result shows that the mAP_0.5 of YOLOv4, YOLOv4-tiny, and YOLOv5s are 77.7%, 78.7%, and 74.8%, respectively. It confirms that the YOLOv4-tiny is the best fit model for pothole detection.

Deep Learning Techniques for Fatty Liver Using Multi-View Ultrasound Images Scanned by Different Scanners: Development and Validation Study

BACKGROUND

Fat fraction values obtained from magnetic resonance imaging (MRI) can be used to obtain an accurate diagnosis of fatty liver diseases. However, MRI is expensive and cannot be performed for everyone.

OBJECTIVE

In this study, we aim to develop multi-view ultrasound image-based convolutional deep learning models to detect fatty liver disease and yield fat fraction values.

METHODS

We extracted 90 ultrasound images of the right intercostal view and 90 ultrasound images of the right intercostal view containing the right renal cortex from 39 cases of fatty liver (MRI-proton density fat fraction [MRI-PDFF] ≥ 5%) and 51 normal subjects (MRI-PDFF < 5%), with MRI-PDFF values obtained from Good Gang-An Hospital. We obtained combined liver and kidney-liver (CLKL) images to train the deep learning models and developed classification and regression models based on the VGG19 model to classify fatty liver disease and yield fat fraction values. We employed the data augmentation techniques such as flip and rotation to prevent the deep learning model from overfitting. We determined the deep learning model with performance metrics such as accuracy, sensitivity, specificity, and coefficient of determination (R2).

RESULTS

In demographic information, all metrics such as age and sex were similar between the two groups-fatty liver disease and normal subjects. In classification, the model trained on CLKL images achieved 80.1% accuracy, 86.2% precision, and 80.5% specificity to detect fatty liver disease. In regression, the predicted fat fraction values of the regression model trained on CLKL images correlated with MRI-PDFF values (R2=0.633), indicating that the predicted fat fraction values were moderately estimated.

CONCLUSIONS

With deep learning techniques and multi-view ultrasound images, it is potentially possible to replace MRI-PDFF values with deep learning predictions for detecting fatty liver disease and estimating fat fraction values.

Artificial Intelligence has Similar Performance to Subjective Assessment of Emphysema Severity on Chest CT

RATIONALE AND OBJECTIVES

To compare an artificial intelligence (AI)-based prototype and subjective grading for predicting disease severity in patients with emphysema.

METHODS

Our IRB approved HIPAA-compliant study included 113 adults (71±8 years; 47 females, 66 males) who had both non-contrast chest CT and pulmonary function tests performed within a span of 2 months. The disease severity was classified based on the forced expiratory volume in 1 second (FEV1 as % of predicted) into mild, moderate, and severe. 2 thoracic radiologists (RA), blinded to the clinical and AI results, graded severity of emphysema on a 5-point scale suggested by the Fleischner Society for each lobe. The whole lung scores were derived from the summation of lobar scores. Thin-section CT images were processed with the AI-Rad Companion Chest prototype (Siemens Healthineers) to quantify low attenuation areas (LAA < - 950 HU) in whole lung and each lobe separately. Bronchial abnormality was assessed by both radiologists and a fully automated software (Philips Healthcare).

RESULTS

Both AI (AUC of 0.77; 95% CI: 0.68 - 0.85) and RA (AUC: 0.76, 95% CI: 0.65 - 0.84) emphysema quantification could differentiate mild, moderate, and severe disease based on FEV1. There was a strong positive correlation between AI and RA (r = 0.72 - 0.80; p <0.001). The combination of emphysema and bronchial abnormality quantification from radiologists' and AI assessment could differentiate between different severities with AUC of 0.80 - 0.82 and 0.87, respectively.

CONCLUSION

The assessed AI-prototypes can predict the disease severity in patients with emphysema with the same predictive value as the radiologists.

CT-derived 3D-diaphragm motion in emphysema and IPF compared to normal subjects

Image registration-based local displacement analysis enables evaluation of respiratory motion between two computed tomography-captured lung volumes. The objective of this study was to compare diaphragm movement among emphysema, idiopathic pulmonary fibrosis (IPF) and normal subjects. 29 normal, 50 emphysema, and 51 IPF subjects were included. A mass preserving image registration technique was used to compute displacement vectors of local lung regions at an acinar scale. Movement of the diaphragm was assumed to be equivalent to movement of the basal lung within 5 mm from the diaphragm. Magnitudes and directions of displacement vectors were compared between the groups. Three-dimensional (3D) and apico-basal displacements were smaller in emphysema than normal subjects (P = 0.003, P = 0.002). Low lung attenuation area on expiration scan showed significant correlations with decreased 3D and apico-basal displacements (r = - 0.546, P < 0.0001; r = - 0.521, P < 0.0001) in emphysema patients. Dorsal-ventral displacement was smaller in IPF than normal subjects (P < 0.0001). The standard deviation of the displacement angle was greater in both emphysema and IPF patients than normal subjects (P < 0.0001). In conclusion, apico-basal movement of the diaphragm is reduced in emphysema while dorsal-ventral movement is reduced in IPF. Image registration technique to multi-volume CT scans provides insight into the pathophysiology of limited diaphragmatic motion in emphysema and IPF.