ABSTRACT
BACKGROUND: Chest x-ray is widely utilized for the evaluation of pulmonary conditions due to its technical simplicity, cost-effectiveness, and portability. However, as a two-dimensional (2-D) imaging modality, chest x-ray images depict limited anatomical details and are challenging to interpret. PURPOSE: To validate the feasibility of reconstructing three-dimensional (3-D) lungs from a single 2-D chest x-ray image via Vision Transformer (ViT). METHODS: We created a cohort of 2525 paired chest x-ray images (scout images) and computed tomography (CT) acquired on different subjects and we randomly partitioned them as follows: (1) 1800 - training set, (2) 200 - validation set, and (3) 525 - testing set. The 3-D lung volumes segmented from the chest CT scans were used as the ground truth for supervised learning. We developed a novel model termed XRayWizard that employed ViT blocks to encode the 2-D chest x-ray image. The aim is to capture global information and establish long-range relationships, thereby improving the performance of 3-D reconstruction. Additionally, a pooling layer at the end of each transformer block was introduced to extract feature information. To produce smoother and more realistic 3-D models, a set of patch discriminators was incorporated. We also devised a novel method to incorporate subject demographics as an auxiliary input to further improve the accuracy of 3-D lung reconstruction. Dice coefficient and mean volume error were used as performance metrics as the agreement between the computerized results and the ground truth. RESULTS: In the absence of subject demographics, the mean Dice coefficient for the generated 3-D lung volumes achieved a value of 0.738 ± 0.091. When subject demographics were included as an auxiliary input, the mean Dice coefficient significantly improved to 0.769 ± 0.089 (p < 0.001), and the volume prediction error was reduced from 23.5 ± 2.7%. to 15.7 ± 2.9%. CONCLUSION: Our experiment demonstrated the feasibility of reconstructing 3-D lung volumes from 2-D chest x-ray images, and the inclusion of subject demographics as additional inputs can significantly improve the accuracy of 3-D lung volume reconstruction.
Subject(s)
Lung , Thorax , Humans , X-Rays , Lung/diagnostic imaging , Tomography, X-Ray Computed/methods , Image Processing, Computer-Assisted/methodsABSTRACT
PURPOSE: To clarify the causal relationship between factors contributing to the postoperative survival of patients with esophageal cancer. METHODS: A cohort of 195 patients who underwent surgery for esophageal cancer between 2008 and 2021 was used in the study. All patients had preoperative chest computed tomography (CT) and positron emission tomography-CT (PET-CT) scans prior to receiving any treatment. From these images, high throughput and quantitative radiomic features, tumor features, and various body composition features were automatically extracted. Causal relationships among these image features, patient demographics, and other clinicopathological variables were analyzed and visualized using a novel score-based directed graph called "Grouped Greedy Equivalence Search" (GGES) while taking prior knowledge into consideration. After supplementing and screening the causal variables, the intervention do-calculus adjustment (IDA) scores were calculated to determine the degree of impact of each variable on survival. Based on this IDA score, a GGES prediction formula was generated. Ten-fold cross-validation was used to assess the performance of the models. The prediction results were evaluated using the R-Squared Score (R2 score). RESULTS: The final causal graphical model was formed by two PET-based image variables, ten body composition variables, four pathological variables, four demographic variables, two tumor variables, and one radiological variable (Percentile 10). Intramuscular fat mass was found to have the most impact on overall survival month. Percentile 10 and overall TNM (T: tumor, N: nodes, M: metastasis) stage were identified as direct causes of overall survival (month). The GGES casual model outperformed GES in regression prediction (R2 = 0.251) (p < 0.05) and was able to avoid unreasonable causality that may contradict common sense. CONCLUSION: The GGES causal model can provide a reliable and straightforward representation of the intricate causal relationships among the variables that impact the postoperative survival of patients with esophageal cancer.
Subject(s)
Esophageal Neoplasms , Positron Emission Tomography Computed Tomography , Humans , Positron Emission Tomography Computed Tomography/methods , Fluorodeoxyglucose F18 , Esophageal Neoplasms/diagnostic imaging , Esophageal Neoplasms/surgery , Positron-Emission Tomography , Tomography, X-Ray Computed , Retrospective StudiesABSTRACT
We present a novel computer algorithm to automatically detect and segment pulmonary embolisms (PEs) on computed tomography pulmonary angiography (CTPA). This algorithm is based on deep learning but does not require manual outlines of the PE regions. Given a CTPA scan, both intra- and extra-pulmonary arteries were firstly segmented. The arteries were then partitioned into several parts based on size (radius). Adaptive thresholding and constrained morphological operations were used to identify suspicious PE regions within each part. The confidence of a suspicious region to be PE was scored based on its contrast in the arteries. This approach was applied to the publicly available RSNA Pulmonary Embolism CT Dataset (RSNA-PE) to identify three-dimensional (3-D) PE negative and positive image patches, which were used to train a 3-D Recurrent Residual U-Net (R2-Unet) to automatically segment PE. The feasibility of this computer algorithm was validated on an independent test set consisting of 91 CTPA scans acquired from a different medical institute, where the PE regions were manually located and outlined by a thoracic radiologist (>18 years' experience). An R2-Unet model was also trained and validated on the manual outlines using a 5-fold cross-validation method. The CNN model trained on the high-confident PE regions showed a Dice coefficient of 0.676±0.168 and a false positive rate of 1.86 per CT scan, while the CNN model trained on the manual outlines demonstrated a Dice coefficient of 0.647±0.192 and a false positive rate of 4.20 per CT scan. The former model performed significantly better than the latter model (p<0.01). The promising performance of the developed PE detection and segmentation algorithm suggests the feasibility of training a deep learning network without dedicating significant efforts to manual annotations of the PE regions on CTPA scans.
Subject(s)
Deep Learning , Pulmonary Embolism , Humans , Pulmonary Embolism/diagnostic imaging , Tomography, X-Ray Computed/methods , Pulmonary Artery/diagnostic imaging , AngiographyABSTRACT
INTRODUCTION: To evaluate the ability of artificial intelligence (AI) software to quantify proptosis for identifying patients who need surgical drainage. METHODS: We pursued a retrospective study including 56 subjects with a clinical diagnosis of subperiosteal orbital abscess (SPOA) secondary to sinusitis at a tertiary pediatric hospital from 2002 to 2016. AI computer software was developed to perform 3D visualization and quantitative assessment of proptosis from computed tomography (CT) images acquired at the time of hospital admission. The AI software automatically computed linear and volume metrics of proptosis to provide more practice-consistent and informative measures. Two experienced physicians independently measured proptosis using the interzygomatic line method on axial CT images. The AI software and physician proptosis assessments were evaluated for association with eventual treatment procedures as standalone markers and in combination with the standard predictors. RESULTS: To treat the SPOA, 31 of 56 (55%) children underwent surgical intervention, including 18 early surgeries (performed within 24 h of admission), and 25 (45%) were managed medically. The physician measurements of proptosis were strongly correlated (Spearman r = 0.89, 95% CI 0.82-0.93) with 95% limits of agreement of ± 1.8 mm. The AI linear measurement was on average 1.2 mm larger (p = 0.007) and only moderately correlated with the average physicians' measurements (r = 0.53, 95% CI 0.31-0.69). Increased proptosis of both AI volumetric and linear measurements were moderately predictive of surgery (AUCs of 0.79, 95% CI 0.68-0.91, and 0.78, 95% CI 0.65-0.90, respectively) with the average physician measurement being poorly to fairly predictive (AUC of 0.70, 95% CI 0.56-0.84). The AI proptosis measures were also significantly greater in the early as compared to the late surgery groups (p = 0.02, and p = 0.04, respectively). The surgical and medical groups showed a substantial difference in the abscess volume (p < 0.001). CONCLUSION: AI proptosis measures significantly differed from physician assessments and showed a good overall ability to predict the eventual treatment. The volumetric AI proptosis measurement significantly improved the ability to predict the likelihood of surgery compared to abscess volume alone. Further studies are needed to better characterize and incorporate the AI proptosis measurements for assisting in clinical decision-making.
ABSTRACT
Purpose: Lung transplantation is the standard treatment for end-stage lung diseases. A crucial factor affecting its success is size matching between the donor's lungs and the recipient's thorax. Computed tomography (CT) scans can accurately determine recipient's lung size, but donor's lung size is often unknown due to the absence of medical images. We aim to predict donor's right/left/total lung volume, thoracic cavity, and heart volume from only subject demographics to improve the accuracy of size matching. Approach: A cohort of 4610 subjects with chest CT scans and basic demographics (i.e., age, gender, race, smoking status, smoking history, weight, and height) was used in this study. The right and left lungs, thoracic cavity, and heart depicted on chest CT scans were automatically segmented using U-Net, and their volumes were computed. Eight machine learning models [i.e., random forest, multivariate linear regression, support vector machine, extreme gradient boosting (XGBoost), multilayer perceptron (MLP), decision tree, k -nearest neighbors, and Bayesian regression) were developed and used to predict the volume measures from subject demographics. The 10-fold cross-validation method was used to evaluate the performances of the prediction models. R -squared ( R 2 ), mean absolute error (MAE), and mean absolute percentage error (MAPE) were used as performance metrics. Results: The MLP model demonstrated the best performance for predicting the thoracic cavity volume ( R 2 : 0.628, MAE: 0.736 L, MAPE: 10.9%), right lung volume ( R 2 : 0.501, MAE: 0.383 L, MAPE: 13.9%), and left lung volume ( R 2 : 0.507, MAE: 0.365 L, MAPE: 15.2%), and the XGBoost model demonstrated the best performance for predicting the total lung volume ( R 2 : 0.514, MAE: 0.728 L, MAPE: 14.0%) and heart volume ( R 2 : 0.430, MAE: 0.075 L, MAPE: 13.9%). Conclusions: Our results demonstrate the feasibility of predicting lung, heart, and thoracic cavity volumes from subject demographics with superior performance compared with available studies in predicting lung volumes.
ABSTRACT
OBJECTIVES: To evaluate the impact of body composition derived from computed tomography (CT) scans on postoperative lung cancer recurrence. METHODS: We created a retrospective cohort of 363 lung cancer patients who underwent lung resections and had verified recurrence, death, or at least 5-year follow-up without either event. Five key body tissues and ten tumor features were automatically segmented and quantified based on preoperative whole-body CT scans (acquired as part of a PET-CT scan) and chest CT scans, respectively. Time-to-event analysis accounting for the competing event of death was performed to analyze the impact of body composition, tumor features, clinical information, and pathological features on lung cancer recurrence after surgery. The hazard ratio (HR) of normalized factors was used to assess individual significance univariately and in the combined models. The 5-fold cross-validated time-dependent receiver operating characteristics analysis, with an emphasis on the area under the 3-year ROC curve (AUC), was used to characterize the ability to predict lung cancer recurrence. RESULTS: Body tissues that showed a standalone potential to predict lung cancer recurrence include visceral adipose tissue (VAT) volume (HR = 0.88, p = 0.047), subcutaneous adipose tissue (SAT) density (HR = 1.14, p = 0.034), inter-muscle adipose tissue (IMAT) volume (HR = 0.83, p = 0.002), muscle density (HR = 1.27, p < 0.001), and total fat volume (HR = 0.89, p = 0.050). The CT-derived muscular and tumor features significantly contributed to a model including clinicopathological factors, resulting in an AUC of 0.78 (95% CI: 0.75-0.83) to predict recurrence at 3 years. CONCLUSIONS: Body composition features (e.g., muscle density, or muscle and inter-muscle adipose tissue volumes) can improve the prediction of recurrence when combined with clinicopathological factors.
Subject(s)
Lung Neoplasms , Humans , Lung Neoplasms/pathology , Retrospective Studies , Positron Emission Tomography Computed Tomography , Neoplasm Recurrence, Local , Lung/pathology , Body Composition/physiology , Tomography, X-Ray Computed/methodsABSTRACT
BACKGROUND: Body composition can be accurately quantified based on computed tomography (CT) and typically reflects an individual's overall health status. However, there is a dearth of research examining the relationship between body composition and survival following esophagectomy. METHODS: We created a cohort consisting of 183 patients who underwent esophagectomy for esophageal cancer without neoadjuvant therapy. The cohort included preoperative PET-CT scans, along with pathologic and clinical data, which were collected prospectively. Radiomic, tumor, PET, and body composition features were automatically extracted from the images. Cox regression models were utilized to identify variables associated with survival. Logistic regression and machine learning models were developed to predict one-, three-, and five-year survival rates. Model performance was evaluated based on the area under the receiver operating characteristics curve (ROC/AUC). To test for the statistical significance of the impact of body composition on survival, body composition features were excluded for the best-performing models, and the DeLong test was used. RESULTS: The one-year survival model contained 10 variables, including three body composition variables (bone mass, bone density, and visceral adipose tissue (VAT) density), and demonstrated an AUC of 0.817 (95% CI: 0.738-0.897). The three-year survival model incorporated 14 variables, including three body composition variables (intermuscular adipose tissue (IMAT) volume, IMAT mass, and bone mass), with an AUC of 0.693 (95% CI: 0.594-0.792). For the five-year survival model, 10 variables were included, of which two were body composition variables (intramuscular adipose tissue (IMAT) volume and visceral adipose tissue (VAT) mass), with an AUC of 0.861 (95% CI: 0.783-0.938). The one- and five-year survival models exhibited significantly inferior performance when body composition features were not incorporated. CONCLUSIONS: Body composition features derived from preoperative CT scans should be considered when predicting survival following esophagectomy.
ABSTRACT
OBJECTIVE: To develop and validate a novel deep learning architecture to classify retinal vein occlusion (RVO) on color fundus photographs (CFPs) and reveal the image features contributing to the classification. METHODS: The neural understanding network (NUN) is formed by two components: (1) convolutional neural network (CNN)-based feature extraction and (2) graph neural networks (GNN)-based feature understanding. The CNN-based image features were transformed into a graph representation to encode and visualize long-range feature interactions to identify the image regions that significantly contributed to the classification decision. A total of 7062 CFPs were classified into three categories: (1) no vein occlusion ("normal"), (2) central RVO, and (3) branch RVO. The area under the receiver operative characteristic (ROC) curve (AUC) was used as the metric to assess the performance of the trained classification models. RESULTS: The AUC, accuracy, sensitivity, and specificity for NUN to classify CFPs as normal, central occlusion, or branch occlusion were 0.975 (± 0.003), 0.911 (± 0.007), 0.983 (± 0.010), and 0.803 (± 0.005), respectively, which outperformed available classical CNN models. CONCLUSION: The NUN architecture can provide a better classification performance and a straightforward visualization of the results compared to CNNs.
Subject(s)
Nuns , Retinal Vein Occlusion , Humans , Retinal Vein Occlusion/diagnostic imaging , Neural Networks, Computer , Fundus Oculi , Diagnostic Techniques, OphthalmologicalABSTRACT
Rationale: Extrapulmonary manifestations of asthma, including fatty infiltration in tissues, may reflect systemic inflammation and influence lung function and disease severity. Objectives: To determine if skeletal muscle adiposity predicts lung function trajectory in asthma. Methods: Adult SARP III (Severe Asthma Research Program III) participants with baseline computed tomography imaging and longitudinal postbronchodilator FEV1% predicted (median follow-up 5 years [1,132 person-years]) were evaluated. The mean of left and right paraspinous muscle density (PSMD) at the 12th thoracic vertebral body was calculated (Hounsfield units [HU]). Lower PSMD reflects higher muscle adiposity. We derived PSMD reference ranges from healthy control subjects without asthma. A linear multivariable mixed-effects model was constructed to evaluate associations of baseline PSMD and lung function trajectory stratified by sex. Measurements and Main Results: Participants included 219 with asthma (67% women; mean [SD] body mass index, 32.3 [8.8] kg/m2) and 37 control subjects (51% women; mean [SD] body mass index, 26.3 [4.7] kg/m2). Participants with asthma had lower adjusted PSMD than control subjects (42.2 vs. 55.8 HU; P < 0.001). In adjusted models, PSMD predicted lung function trajectory in women with asthma (ß = -0.47 Δ slope per 10-HU decrease; P = 0.03) but not men (ß = 0.11 Δ slope per 10-HU decrease; P = 0.77). The highest PSMD tertile predicted a 2.9% improvement whereas the lowest tertile predicted a 1.8% decline in FEV1% predicted among women with asthma over 5 years. Conclusions: Participants with asthma have lower PSMD, reflecting greater muscle fat infiltration. Baseline PSMD predicted lung function decline among women with asthma but not men. These data support an important role of metabolic dysfunction in lung function decline.
Subject(s)
Asthma , Lung , Adult , Humans , Female , Male , Adiposity , Forced Expiratory Volume , Obesity , Muscle, Skeletal/diagnostic imagingABSTRACT
INTRODUCTION: Muscle loss is an important extrapulmonary manifestation of COPD. Dual energy X-ray absorptiometry (DXA) is the method of choice for body composition measurement but is not widely used for muscle mass evaluation. The pectoralis muscle area (PMA) is quantifiable by CT and predicts cross-sectional COPD-related morbidity. There are no studies that compare PMA with DXA measures or that evaluate longitudinal relationships between PMA and lung disease progression. METHODS: Participants from our longitudinal tobacco-exposed cohort had baseline and 6-year chest CT (n=259) and DXA (n=164) data. Emphysema was quantified by CT density histogram parenchymal scoring using the 15th percentile technique. Fat-free mass index (FFMI) and appendicular skeletal mass index (ASMI) were calculated from DXA measurements. Linear regression model relationships were reported using standardised coefficient (ß) with 95% CI. RESULTS: PMA was more strongly associated with DXA measures than with body mass index (BMI) in both cross-sectional (FFMI: ß=0.76 (95% CI 0.65 to 0.86), p<0.001; ASMI: ß=0.76 (95% CI 0.66 to 0.86), p<0.001; BMI: ß=0.36 (95% CI 0.25 to 0.47), p<0.001) and longitudinal (ΔFFMI: ß=0.43 (95% CI 0.28 to 0.57), p<0.001; ΔASMI: ß=0.42 (95% CI 0.27 to 0.57), p<0.001; ΔBMI: ß=0.34 (95% CI 0.22 to 0.46), p<0.001) models. Six-year change in PMA was associated with 6-year change in emphysema (ß=0.39 (95% CI 0.23 to 0.56), p<0.001) but not with 6-year change in airflow obstruction. CONCLUSIONS: PMA is an accessible measure of muscle mass and may serve as a useful clinical surrogate for assessing skeletal muscle loss in smokers. Decreased PMA correlated with emphysema progression but not lung function decline, suggesting a difference in the pathophysiology driving emphysema, airflow obstruction and comorbidity risk.
Subject(s)
Emphysema , Pulmonary Emphysema , Humans , Pectoralis Muscles , Nicotiana , Absorptiometry, Photon , Cross-Sectional Studies , Pulmonary Emphysema/diagnostic imaging , Pulmonary Emphysema/etiology , Muscle, Skeletal/diagnostic imaging , Tomography, X-Ray Computed/methodsABSTRACT
BACKGROUND: Estimating whole-body composition from limited region-computed tomography (CT) scans has many potential applications in clinical medicine; however, it is challenging. PURPOSE: To investigate if whole-body composition based on several tissue types (visceral adipose tissue [VAT], subcutaneous adipose tissue [SAT], intermuscular adipose tissue [IMAT], skeletal muscle [SM], and bone) can be reliably estimated from a chest CT scan only. METHODS: A cohort of 97 lung cancer subjects who underwent both chest CT scans and whole-body positron emission tomography-CT scans at our institution were collected. We used our in-house software to automatically segment and quantify VAT, SAT, IMAT, SM, and bone on the CT images. The field-of-views of the chest CT scans and the whole-body CT scans were standardized, namely, from vertebra T1 to L1 and from C1 to the bottom of the pelvis, respectively. Multivariate linear regression was used to develop the computer models for estimating the volumes of whole-body tissues from chest CT scans. Subject demographics (e.g., gender and age) and lung volume were included in the modeling analysis. Ten-fold cross-validation was used to validate the performance of the prediction models. Mean absolute difference (MAD) and R-squared (R2 ) were used as the performance metrics to assess the model performance. RESULTS: The R2 values when estimating volumes of whole-body SAT, VAT, IMAT, total fat, SM, and bone from the regular chest CT scans were 0.901, 0.929, 0.900, 0.933, 0.928, and 0.918, respectively. The corresponding MADs (percentage difference) were 1.44 ± 1.21 L (12.21% ± 11.70%), 0.63 ± 0.49 L (29.68% ± 61.99%), 0.12 ± 0.09 L (16.20% ± 18.42%), 1.65 ± 1.40 L (10.43% ± 10.79%), 0.71 ± 0.68 L (5.14% ± 4.75%), and 0.17 ± 0.15 L (4.32% ± 3.38%), respectively. CONCLUSION: Our algorithm shows promise in its ability to estimate whole-body compositions from chest CT scans. Body composition measures based on chest CT scans are more accurate than those based on vertebra third lumbar.
Subject(s)
Tomography, X-Ray Computed , Tomography , Humans , Body CompositionABSTRACT
Objective: To develop and validate a novel convolutional neural network (CNN) termed "Super U-Net" for medical image segmentation. Methods: Super U-Net integrates a dynamic receptive field module and a fusion upsampling module into the classical U-Net architecture. The model was developed and tested to segment retinal vessels, gastrointestinal (GI) polyps, skin lesions on several image types (i.e., fundus images, endoscopic images, dermoscopic images). We also trained and tested the traditional U-Net architecture, seven U-Net variants, and two non-U-Net segmentation architectures. K-fold cross-validation was used to evaluate performance. The performance metrics included Dice similarity coefficient (DSC), accuracy, positive predictive value (PPV), and sensitivity. Results: Super U-Net achieved average DSCs of 0.808±0.0210, 0.752±0.019, 0.804±0.239, and 0.877±0.135 for segmenting retinal vessels, pediatric retinal vessels, GI polyps, and skin lesions, respectively. The Super U-net consistently outperformed U-Net, seven U-Net variants, and two non-U-Net segmentation architectures (p < 0.05). Conclusion: Dynamic receptive fields and fusion upsampling can significantly improve image segmentation performance.
ABSTRACT
Background Clustering key clinical characteristics of participants in the Severe Asthma Research Program (SARP), a large, multicenter prospective observational study of patients with asthma and healthy controls, has led to the identification of novel asthma phenotypes. Purpose To determine whether quantitative CT (qCT) could help distinguish between clinical asthma phenotypes. Materials and Methods A retrospective cross-sectional analysis was conducted with the use of qCT images (maximal bronchodilation at total lung capacity [TLC], or inspiration, and functional residual capacity [FRC], or expiration) from the cluster phenotypes of SARP participants (cluster 1: minimal disease; cluster 2: mild, reversible; cluster 3: obese asthma; cluster 4: severe, reversible; cluster 5: severe, irreversible) enrolled between September 2001 and December 2015. Airway morphometry was performed along standard paths (RB1, RB4, RB10, LB1, and LB10). Corresponding voxels from TLC and FRC images were mapped with use of deformable image registration to characterize disease probability maps (DPMs) of functional small airway disease (fSAD), voxel-level volume changes (Jacobian), and isotropy (anisotropic deformation index [ADI]). The association between cluster assignment and qCT measures was evaluated using linear mixed models. Results A total of 455 participants were evaluated with cluster assignments and CT (mean age ± SD, 42.1 years ± 14.7; 270 women). Airway morphometry had limited ability to help discern between clusters. DPM fSAD was highest in cluster 5 (cluster 1 in SARP III: 19.0% ± 20.6; cluster 2: 18.9% ± 13.3; cluster 3: 24.9% ± 13.1; cluster 4: 24.1% ± 8.4; cluster 5: 38.8% ± 14.4; P < .001). Lower whole-lung Jacobian and ADI values were associated with greater cluster severity. Compared to cluster 1, cluster 5 lung expansion was 31% smaller (Jacobian in SARP III cohort: 2.31 ± 0.6 vs 1.61 ± 0.3, respectively, P < .001) and 34% more isotropic (ADI in SARP III cohort: 0.40 ± 0.1 vs 0.61 ± 0.2, P < .001). Within-lung Jacobian and ADI SDs decreased as severity worsened (Jacobian SD in SARP III cohort: 0.90 ± 0.4 for cluster 1; 0.79 ± 0.3 for cluster 2; 0.62 ± 0.2 for cluster 3; 0.63 ± 0.2 for cluster 4; and 0.41 ± 0.2 for cluster 5; P < .001). Conclusion Quantitative CT assessments of the degree and intraindividual regional variability of lung expansion distinguished between well-established clinical phenotypes among participants with asthma from the Severe Asthma Research Program study. © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Verschakelen in this issue.
Subject(s)
Asthma , Asthma/diagnostic imaging , Cross-Sectional Studies , Female , Humans , Lung/diagnostic imaging , Phenotype , Pulmonary Disease, Chronic Obstructive , Retrospective Studies , Tomography, X-Ray Computed/methodsABSTRACT
We present a novel integrative computerized solution to automatically identify and differentiate pulmonary arteries and veins depicted on chest computed tomography (CT) without iodinated contrast agents. We first identified the central extrapulmonary arteries and veins using a convolutional neural network (CNN) model. Then, a computational differential geometry method was used to automatically identify the tubular-like structures in the lungs with high densities, which we believe are the intrapulmonary vessels. Beginning with the extrapulmonary arteries and veins, we progressively traced the intrapulmonary vessels by following their skeletons and differentiated them into arteries and veins. Instead of manually labeling the numerous arteries and veins in the lungs for machine learning, this integrative strategy limits the manual effort only to the large extrapulmonary vessels. We used a dataset consisting of 120 chest CT scans acquired on different subjects using various protocols to develop, train, and test the algorithms. Our experiments on an independent test set (nâ¯=â¯15) showed promising performance. The computer algorithm achieved a sensitivity of â¼98% in labeling the pulmonary artery and vein branches when compared with a human expert's results, demonstrating the feasibility of our computerized solution in pulmonary artery/vein labeling.
Subject(s)
Pulmonary Artery , Tomography, X-Ray Computed , Algorithms , Humans , Neural Networks, Computer , Pulmonary Artery/diagnostic imaging , Thorax , Tomography, X-Ray Computed/methodsABSTRACT
PURPOSE: To investigate the relationship between macrovasculature features and the standardized uptake value (SUV) of positron emission tomography (PET), which is a surrogate for the metabolic activity of a lung tumor. METHODS: We retrospectively analyzed a cohort of 90 lung cancer patients who had both chest CT and PET-CT examinations before receiving cancer treatment. The SUVs in the medical reports were used. We quantified three macrovasculature features depicted on CT images (i.e., vessel number, vessel volume, and vessel tortuosity) and several tumor features (i.e., volume, maximum diameter, mean diameter, surface area, and density). Tumor size (e.g., volume) was used as a covariate to adjust for possible confounding factors. Backward stepwise multiple regression analysis was performed to develop a model for predicting PET SUV from the relevant image features. The Bonferroni correction was used for multiple comparisons. RESULTS: PET SUV was positively correlated with vessel volume (R = 0.44, p < 0.001) and vessel number (R = 0.44, p < 0.001) but not with vessel tortuosity (R = 0.124, p > 0.05). After adjusting for tumor size, PET SUV was significantly correlated with vessel tortuosity (R = 0.299, p = 0.004) and vessel number (R = 0.224, p = 0.035), but only marginally correlated with vessel volume (R = 0.187, p = 0.079). The multiple regression model showed a performance with an R-Squared of 0.391 and an adjusted R-Squared of 0.355 (p < 0.001). CONCLUSIONS: Our investigations demonstrate the potential relationship between macrovasculature and PET SUV and suggest the possibility of inferring the metabolic activity of a lung tumor from chest CT images.
Subject(s)
Lung Neoplasms , Positron Emission Tomography Computed Tomography , Fluorodeoxyglucose F18 , Humans , Lung Neoplasms/diagnostic imaging , Positron-Emission Tomography , Retrospective StudiesABSTRACT
BACKGROUND: Sarcoidosis is a multisystem granulomatous disease of unknown origin with a variable and often unpredictable course and pattern of organ involvement. In this study we sought to identify specific bronchoalveolar lavage (BAL) cell gene expression patterns indicative of distinct disease phenotypic traits. METHODS: RNA sequencing by Ion Torrent Proton was performed on BAL cells obtained from 215 well-characterised patients with pulmonary sarcoidosis enrolled in the multicentre Genomic Research in Alpha-1 Antitrypsin Deficiency and Sarcoidosis (GRADS) study. Weighted gene co-expression network analysis and nonparametric statistics were used to analyse genome-wide BAL transcriptome. Validation of results was performed using a microarray expression dataset of an independent sarcoidosis cohort (Freiburg, Germany; n=50). RESULTS: Our supervised analysis found associations between distinct transcriptional programmes and major pulmonary phenotypic manifestations of sarcoidosis including T-helper type 1 (Th1) and Th17 pathways associated with hilar lymphadenopathy, transforming growth factor-ß1 (TGFB1) and mechanistic target of rapamycin (MTOR) signalling with parenchymal involvement, and interleukin (IL)-7 and IL-2 with airway involvement. Our unsupervised analysis revealed gene modules that uncovered four potential sarcoidosis endotypes including hilar lymphadenopathy with increased acute T-cell immune response; extraocular organ involvement with PI3K activation pathways; chronic and multiorgan disease with increased immune response pathways; and multiorgan involvement, with increased IL-1 and IL-18 immune and inflammatory responses. We validated the occurrence of these endotypes using gene expression, pulmonary function tests and cell differentials from Freiburg. CONCLUSION: Taken together, our results identify BAL gene expression programmes that characterise major pulmonary sarcoidosis phenotypes and suggest the presence of distinct disease molecular endotypes.
Subject(s)
Sarcoidosis, Pulmonary , Sarcoidosis , Bronchoalveolar Lavage , Bronchoalveolar Lavage Fluid , Humans , Sarcoidosis, Pulmonary/genetics , TranscriptomeABSTRACT
PURPOSE: The potential to compute volume metrics of emphysema from planar scout images was investigated in this study. The successful implementation of this concept will have a wide impact in different fields, and specifically, maximize the diagnostic potential of the planar medical images. METHODS: We investigated our premise using a well-characterized chronic obstructive pulmonary disease (COPD) cohort. In this cohort, planar scout images from computed tomography (CT) scans were used to compute lung volume and percentage of emphysema. Lung volume and percentage of emphysema were quantified on the volumetric CT images and used as the "ground truth" for developing the models to compute the variables from the corresponding scout images. We trained two classical convolutional neural networks (CNNs), including VGG19 and InceptionV3, to compute lung volume and the percentage of emphysema from the scout images. The scout images (n = 1,446) were split into three subgroups: (1) training (n = 1,235), (2) internal validation (n = 99), and (3) independent test (n = 112) at the subject level in a ratio of 8:1:1. The mean absolute difference (MAD) and R-square (R2) were the performance metrics to evaluate the prediction performance of the developed models. RESULTS: The lung volumes and percentages of emphysema computed from a single planar scout image were significantly linear correlated with the measures quantified using volumetric CT images (VGG19: R2 = 0.934 for lung volume and R2 = 0.751 for emphysema percentage, and InceptionV3: R2 = 0.977 for lung volume and R2 = 0.775 for emphysema percentage). The mean absolute differences (MADs) for lung volume and percentage of emphysema were 0.302 ± 0.247L and 2.89 ± 2.58%, respectively, for VGG19, and 0.366 ± 0.287L and 3.19 ± 2.14, respectively, for InceptionV3. CONCLUSIONS: Our promising results demonstrated the feasibility of inferring volume metrics from planar images using CNNs.
Subject(s)
Emphysema , Pulmonary Disease, Chronic Obstructive , Pulmonary Emphysema , Emphysema/diagnostic imaging , Humans , Lung/diagnostic imaging , Lung Volume Measurements , Pilot Projects , Pulmonary Emphysema/diagnostic imagingABSTRACT
BACKGROUND: Smokers manifest varied phenotypes of pulmonary impairment. RESEARCH QUESTION: Which pulmonary phenotypes are associated with coronary artery disease (CAD) in smokers? STUDY DESIGN AND METHODS: We analyzed data from the University of Pittsburgh COPD Specialized Center for Clinically Oriented Research (SCCOR) cohort (n = 481) and the Genetic Epidemiology of COPD (COPDGene) cohort (n = 2,580). Participants were current and former smokers with > 10 pack-years of tobacco exposure. Data from the two cohorts were analyzed separately because of methodologic differences. Lung hyperinflation was assessed by plethysmography in the SCCOR cohort and by inspiratory and expiratory CT scan lung volumes in the COPDGene cohort. Subclinical CAD was assessed as the coronary artery calcium score, whereas clinical CAD was defined as a self-reported history of CAD or myocardial infarction (MI). Analyses were performed in all smokers and then repeated in those with airflow obstruction (FEV1 to FVC ratio, < 0.70). RESULTS: Pulmonary phenotypes, including airflow limitation, emphysema, lung hyperinflation, diffusion capacity, and radiographic measures of airway remodeling, showed weak to moderate correlations (r < 0.7) with each other. In multivariate models adjusted for pulmonary phenotypes and CAD risk factors, lung hyperinflation was the only phenotype associated with calcium score, history of clinical CAD, or history of MI (per 0.2 higher expiratory and inspiratory CT scan lung volume; coronary calcium: OR, 1.2; 95% CI, 1.1-1.5; P = .02; clinical CAD: OR, 1.6; 95% CI, 1.1-2.3; P = .01; and MI in COPDGene: OR, 1.7; 95% CI, 1.0-2.8; P = .05). FEV1 and emphysema were associated with increased risk of CAD (P < .05) in models adjusted for CAD risk factors; however, these associations were attenuated on adjusting for lung hyperinflation. Results were the same in those with airflow obstruction and were present in both cohorts. INTERPRETATION: Lung hyperinflation is associated strongly with clinical and subclinical CAD in smokers, including those with airflow obstruction. After lung hyperinflation was accounted for, FEV1 and emphysema no longer were associated with CAD. Subsequent studies should consider measuring lung hyperinflation and examining its mechanistic role in CAD in current and former smokers.
Subject(s)
Airway Obstruction , Coronary Artery Disease , Coronary Vessels/diagnostic imaging , Lung , Pulmonary Emphysema , Smoking/epidemiology , Airway Obstruction/diagnosis , Airway Obstruction/physiopathology , Airway Remodeling , Asymptomatic Diseases/epidemiology , Biological Variation, Population , Coronary Artery Disease/diagnosis , Coronary Artery Disease/epidemiology , Female , Humans , Lung/diagnostic imaging , Lung/pathology , Lung/physiopathology , Male , Middle Aged , Organ Size , Plethysmography/methods , Pulmonary Emphysema/diagnostic imaging , Pulmonary Emphysema/physiopathology , Respiratory Function Tests/methods , Risk Factors , Tomography, X-Ray Computed/methods , United States/epidemiologyABSTRACT
OBJECTIVES: To develop and validate a deep learning algorithm to automatically detect and segment an orbital abscess depicted on computed tomography (CT). METHODS: We retrospectively collected orbital CT scans acquired on 67 pediatric subjects with a confirmed orbital abscess in the setting of infectious orbital cellulitis. A context-aware convolutional neural network (CA-CNN) was developed and trained to automatically segment orbital abscess. To reduce the requirement for a large dataset, transfer learning was used by leveraging a pre-trained model for CT-based lung segmentation. An ophthalmologist manually delineated orbital abscesses depicted on the CT images. The classical U-Net and the CA-CNN models with and without transfer learning were trained and tested on the collected dataset using the 10-fold cross-validation method. Dice coefficient, Jaccard index, and Hausdorff distance were used as performance metrics to assess the agreement between the computerized and manual segmentations. RESULTS: The context-aware U-Net with transfer learning achieved an average Dice coefficient and Jaccard index of 0.78 ± 0.12 and 0.65 ± 0.13, which were consistently higher than the classical U-Net or the context-aware U-Net without transfer learning (P < 0.01). The average differences of the abscess between the computerized results and the experts in terms of volume and Hausdorff distance were 0.10 ± 0.11 mL and 1.94 ± 1.21 mm, respectively. The context-aware U-Net detected all orbital abscess without false positives. CONCLUSIONS: The deep learning solution demonstrated promising performance in detecting and segmenting orbital abscesses on CT images in strong agreement with a human observer.
Subject(s)
Deep Learning , Orbital Cellulitis , Abscess/diagnostic imaging , Child , Humans , Retrospective Studies , Tomography, X-Ray ComputedABSTRACT
INTRODUCTION: Our objective was to assess the utility of using lesion-mapping software to calculate precise orbital volumes to predict patients who would benefit from early surgical intervention. METHODS: We retrospectively reviewed patients diagnosed with subperiosteal orbital abscess (SPOA) secondary to sinusitis at a tertiary pediatric hospital from 2005 to 2016. Diagnoses were confirmed by CT scans. Lesion-mapping software was used to measure SPOA volume using initial CT images. Data collected included patient demographics, length of hospital stay, and subsequent medical or surgical treatment. RESULTS: Thirty-three (52%) patients ultimately underwent surgical intervention, while 30 (48%) were managed medically. Between the surgical and medical groups, there were no differences in gender, age, or comorbidities. The surgical group had larger abscess volumes than the medically managed group (0.94 mL vs. 0.31 mL, p < 0.01). Overall, increased SPOA volume was associated with increased age (Pearson's coefficient = 0.374, p ≤ 0.01) and increased total days of intravenous (IV) antibiotic administration (Pearson's coefficient = 0.260, p = 0.039). Patients who underwent surgery on the day of admission had 25% shorter hospital stay than patients who had delayed surgery (p < 0.01). Our calculated sensitivity-optimized SPOA volume cutoff of 0.231 mL yielded sensitivity of 90.9% and specificity of 70.0%. CONCLUSIONS: This is the first study to use lesion-mapping software for precise calculation of SPOA volumes, which can help refine indications for early surgical intervention and help decrease length of hospital stay.