Pulmonary lobe segmentation in CT examinations using implicit surface fitting

Quantitative assessment of lung structure changes in low-intensity smokers: a retrospective study in a Chinese male cohort

Background

With an increasing number of smokers who consume fewer cigarettes, it is crucial to understand the lung structure changes of low-intensity smoking. This study aimed to investigate the lung structure changes in low-intensity smokers in a Chinese male cohort.

Methods

Chest computed tomography (CT) examinations of 465 asymptomatic healthy male participants were divided into non-smoking (n=256), light-smoking (n=84), intermediate-smoking (n=85), and heavy-smoking (n=40) groups. Low-intensity smokers (fewer than 10 cigarettes per day) were included (n=32), and a new group of non-smokers was generated using propensity score matching according to age. Quantitative CT parameters, including the volume of the intrapulmonary vessel (IPVV), the volume of the lung, mean lung density (MLD), the low-attenuation areas below -910 Hounsfield units (LAA-910), and the volume ratio of intrapulmonary vessel to the lung for the total lung and each lobe were measured. Quantitative CT parameters were compared among the four smoking groups and also between the low-intensity smokers and non-smokers. Binary logistic regression was used to determine the independent quantitative CT measurements of smoking intensity.

Results

Compared with that in non-smokers, the IPVV and the MLD of the total lung and five lobes was significantly higher in light smokers (P<0.05); meanwhile, the LAA-910 of the total lung and five lobes of the light and intermediate smokers were significantly lower (P<0.05). The IPVV of the total lung and five lobes was significantly higher in the low-intensity smoking group (P<0.05). The IPVV of the total lung was the independent factor for discriminating between the non-smokers and light smokers (odds ratio =1.040; 95% confidence interval: 1.027-1.053) and between the non-smokers and low-intensity smokers (odds ratio =1.034; 95% confidence interval: 1.013-1.055).

Conclusions

CT-quantified measurements of the IPVVs and MLD increased in light and intermediate smokers. The IPVV of the total lung was selected as the independent factor between non-smokers and light smokers and between non-smokers and low-intensity smokers.

Predicting Postoperative Lung Cancer Recurrence and Survival Using Cox Proportional Hazards Regression and Machine Learning

BACKGROUND

Surgical resection remains the standard treatment for early-stage lung cancer. However, the recurrence rate after surgery is unacceptably high, ranging from 30% to 50%. Despite extensive efforts, accurately predicting the likelihood and timing of recurrence remains a significant challenge. This study aims to predict postoperative recurrence by identifying novel image biomarkers from preoperative chest CT scans.

METHODS

A cohort of 309 patients was selected from 512 non-small-cell lung cancer patients who underwent lung resection. Cox proportional hazards regression analysis was employed to identify risk factors associated with recurrence and was compared with machine learning (ML) methods for predictive performance. The goal is to improve the ability to predict the risk and time of recurrence in seemingly "cured" patients, enabling personalized surveillance strategies to minimize lung cancer recurrence.

RESULTS

The Cox hazards analyses identified surgical procedure, TNM staging, lymph node involvement, body composition, and tumor characteristics as significant determinants of recurrence risk, both for local/regional and distant recurrence, as well as recurrence-free survival (RFS) and overall survival (OS) (p < 0.05). ML models and Cox models exhibited comparable predictive performance, with an area under the receiver operative characteristic (ROC) curve (AUC) ranging from 0.75 to 0.77.

CONCLUSIONS

These promising findings demonstrate the feasibility of predicting postoperative lung cancer recurrence and survival time using preoperative chest CT scans. However, further validation using larger, multisite cohort is necessary to ensure robustness and facilitate integration into clinical practice for improved cancer management.

CT-Derived Features as Predictors of Clot Burden and Resolution

Objectives: To evaluate the prognostic utility of CT-imaging-derived biomarkers in distinguishing acute pulmonary embolism (PE) resolution and its progression to chronic PE, as well as their association with clot burden. Materials and Methods: We utilized a cohort of 45 patients (19 male (42.2%)) and 96 corresponding CT scans with exertional dyspnea following an acute PE. These patients were referred for invasive cardiopulmonary exercise testing (CPET) at the University of Pittsburgh Medical Center from 2018 to 2022, for whom we have ground truth classification of chronic PE, as well as CT-derived features related to body composition, cardiopulmonary vasculature, and PE clot burden using artificial intelligence (AI) algorithms. We applied Lasso regularization to select parameters, followed by (1) Ordinary Least Squares (OLS) regressions to analyze the relationship between clot burden and the selected parameters and (2) logistic regressions to differentiate between chronic and resolved patients. Results: Several body composition and cardiopulmonary factors showed statistically significant association with clot burden. A multivariate model based on cardiopulmonary features demonstrated superior performance in predicting PE resolution (AUC: 0.83, 95% CI: 0.71-0.95), indicating significant associations between airway ratio (negative correlation), aorta diameter, and heart volume (positive correlation) with PE resolution. Other multivariate models integrating demographic features showed comparable performance, while models solely based on body composition and baseline clot burden demonstrated inferior performance. Conclusions: Our analysis suggests that cardiopulmonary and demographic features hold prognostic value for predicting PE resolution, whereas body composition and baseline clot burden do not. Clinical Relevance: Our identified prognostic factors may facilitate the follow-up procedures for patients diagnosed with acute PE.

Lightweight Techniques to Improve Generalization and Robustness of U-Net Based Networks for Pulmonary Lobe Segmentation

Lung lobe segmentation in chest CT is relevant to a wide range of clinical applications. However, existing segmentation pipelines often exhibit vulnerabilities and performance degradations when applied to external datasets. This is usually attributed to the size of the available dataset or model. We show that it is possible to enhance generalizability without huge resources by carefully curating the dataset and combining machine learning with medical expertise. Multiple machine learning techniques (self-supervision (SSL), attention (A), and data augmentation (DA)) are used to train a fast and fully-automated lung lobe segmentation model based on 2D U-Net. Our study involved evaluating these techniques on a diverse dataset collected under the RACOON project, encompassing 100 CT chest scans from patients with bacterial, viral, or SARS-CoV2 infections. We compare our model to a baseline U-Net trained on the same dataset. Our approach significantly improved segmentation accuracy (Dice score of 92.8% vs. 82.3%, p < 0.001). Moreover, our model achieved state-of-the-art performance (Dice score of 92.8% vs. 90.8% for the literature's state-of-the-art, p = 0.102) with reduced training examples (69 vs. 231 CT Scans). Among the techniques, data augmentation with expert knowledge displayed the most significant impact, enhancing the Dice score by +0.056. Notably, these enhancements are not limited to lobe segmentation but can be seamlessly integrated into various medical imaging segmentation tasks, demonstrating their versatility and potential for broader applications.

ARR-GCN: Anatomy-Relation Reasoning Graph Convolutional Network for Automatic Fine-Grained Segmentation of Organ's Surgical Anatomy

Anatomical resection (AR) based on anatomical sub-regions is a promising method of precise surgical resection, which has been proven to improve long-term survival by reducing local recurrence. The fine-grained segmentation of an organ's surgical anatomy (FGS-OSA), i.e., segmenting an organ into multiple anatomic regions, is critical for localizing tumors in AR surgical planning. However, automatically obtaining FGS-OSA results in computer-aided methods faces the challenges of appearance ambiguities among sub-regions (i.e., inter-sub-region appearance ambiguities) caused by similar HU distributions in different sub-regions of an organ's surgical anatomy, invisible boundaries, and similarities between anatomical landmarks and other anatomical information. In this paper, we propose a novel fine-grained segmentation framework termed the "anatomic relation reasoning graph convolutional network" (ARR-GCN), which incorporates prior anatomic relations into the framework learning. In ARR-GCN, a graph is constructed based on the sub-regions to model the class and their relations. Further, to obtain discriminative initial node representations of graph space, a sub-region center module is designed. Most importantly, to explicitly learn the anatomic relations, the prior anatomic-relations among the sub-regions are encoded in the form of an adjacency matrix and embedded into the intermediate node representations to guide framework learning. The ARR-GCN was validated on two FGS-OSA tasks: i) liver segments segmentation, and ii) lung lobes segmentation. Experimental results on both tasks outperformed other state-of-the-art segmentation methods and yielded promising performances by ARR-GCN for suppressing ambiguities among sub-regions.

Quantification of pulmonary vessel volumes on low-dose computed tomography in a healthy male Chinese population: the effects of aging and smoking

BACKGROUND

This study sought to determine pulmonary vascular volumes (PVVs) on low-dose computed tomography (LDCT) in a healthy male Chinese population and analyze the effects of aging and smoking on PVVs.

METHODS

A total of 1,320 healthy male participants (comprising 720 non-smokers, 445 smokers, and 155 ex-smokers) who underwent LDCT were retrospectively included in this study. Their demographic data and smoking status data were collected. An automatic integration segmentation approach for LDCT was used to segment pulmonary vessels semi-automatically. The PVVs of the whole lung, left lung, and right lung on LDCT were calculated, and correlations between PVVs and age and smoking status were then compared.

RESULTS

The inter-rater correlation coefficient of the whole lung, left lung, and right lung PVVs was 0.98 [95% confidence interval (CI): 0.95-0.99], 0.97 (95% CI: 0.93-0.98), and 0.97 (95% CI: 0.94-0.99), respectively. The intra-class correlation coefficient of the whole lung left lung, and right lung PVVs was 0.98 (95% CI: 0.95-0.99), 0.96 (95% CI: 0.95-0.99), and 0.96 (95% CI: 0.92-0.98), respectively. In non-smokers, PVVs decreased with age. The PVVs of heavy smokers were higher than those of light smokers, ex-smokers, and non-smokers. The PVVs of ex-smokers were comparable to those of light smokers.

CONCLUSIONS

The PVVs measured on LDCT tended to decrease with age in healthy male non-smokers gradually. Compared to non-smokers, the PVVs of smokers increased, even with the normal lung function.

Sex-related Differences in Airway Dimensions: A Study Based on Quantitative Computed Tomography among Chinese Population

Sex-dependent radiation injury may be related to the differences in physiological characteristics between the sexes. This study aimed to better understand variations in airway dimensions among male and female Chinese non-smokers. This study included 970 adults and 45 children who underwent chest CT. All participants were non-smokers, without current or former chronic pulmonary disease, and all underwent CT examination. The CT images were quantitatively assessed, providing airway dimensions. The differences in inner diameter, wall thickness, wall area (WA), and WA% for each airway were compared between male and female patients. Sex is an important influencing factor in airway morphological parameters. These parameters are different between men and women: men have a larger airway diameter (P < 0.05) and smaller wall area (WA%, P < 0.05) compared with women. Younger women (<35 years) have a greater diameter and smaller WA% compared with older women (P < 0.05). Sex-related differences in airway morphology were not observed in pediatric participants. Significant differences were found in quantitative CT measures of WA% and an internal diameter among non-smokers of varying sex. The differences found in this study might explain, in part, sex-dependency of radiation injury and a possible radiological protection scheme.

Potential Value of Expiratory CT in Quantitative Assessment of Pulmonary Vessels in COPD

Objective: To investigate the associations between intrapulmonary vascular volume (IPVV) depicted on inspiratory and expiratory CT scans and disease severity in COPD patients, and to determine which CT parameters can be used to predict IPVV.

Methods: We retrospectively collected 89 CT examinations acquired on COPD patients from an available database. All subjects underwent both inspiratory and expiratory CT scans. We quantified the IPVV, airway wall thickness (WT), the percentage of the airway wall area (WA%), and the extent of emphysema (LAA%₋₉₅₀) using an available pulmonary image analysis tool. The underlying relationship between IPVV and COPD severity, which was defined as mild COPD (GOLD stage I and II) and severe COPD (GOLD stage III and IV), was analyzed using the Student's t-test (or Mann-Whitney U-test). The correlations of IPVV with pulmonary function tests (PFTs), LAA%₋₉₅₀, and airway parameters for the third to sixth generation bronchus were analyzed using the Pearson or Spearman's rank correlation coefficients and multiple stepwise regression.

Results: In the subgroup with only inspiratory examinations, the correlation coefficients between IPVV and PFT measures were −0.215 ~ −0.292 (p < 0.05), the correlation coefficients between IPVV and WT₃₋₆ were 0.233 ~ 0.557 (p < 0.05), and the correlation coefficient between IPVV and LAA%₋₉₅₀ were 0.238 ~ 0.409 (p < 0.05). In the subgroup with only expiratory scan, the correlation coefficients between IPVV and PFT measures were −0.238 ~ −0.360 (p < 0.05), the correlation coefficients between IPVV and WT₃₋₆ were 0.260 ~ 0.566 (p < 0.05), and the correlation coefficient between IPVV and LAA%₋₉₅₀ were 0.241 ~ 0.362 (p < 0.05). The multiple stepwise regression analyses demonstrated that WT were independently associated with IPVV (P < 0.05).

Conclusion: The expiratory CT scans can provide a more accurate assessment of COPD than the inspiratory CT scans, and the airway wall thickness maybe an independent predictor of pulmonary vascular alteration in patients with COPD.

Development of a Machine Learning-Assisted Model for the Early Detection of Severe COVID-19 Cases Combining Blood Test and Quantitative Computed Tomography Parameters

<sec> <title>Purpose:</title> This study aimed to identify severe Coronavirus Disease 2019 (COVID-19) cases combining blood test results and imaging parameters based on a machine learning classifier at the initial admission. </sec> <sec> <title>Materials and methods:</title> Ninety-five non-severe and 22 severe laboratory-confirmed COVID-19 cases treated between January 23, 2020 and March 25, 2020 were examined in this retrospective trial. Blood test results and chest computed tomography (CT) images were obtained at the initial admission. The lesions on CT images were segmented using an artificial intelligent (AI) tool. Then, quantitative CT (QCT) parameters, including the volume, percentage, ground glass opacity (GGO) percentage and heterogeneity of the lesions were calculated. Correlations of blood test results and QCT parameters were analyzed by the Pearson test first. Then, discriminative features for detecting severe cases were selected by both the independent samples t test and least absolute shrinkage and selection operator (LASSO) regression. Next, support vector machine (SVM), Gaussian naïve Bayes (GNB), Knearest neighbor (KNN), decision tree (DT), random forest (RF) and multi-layer perceptron-neural net (MLP-NN) algorithms were used as classifiers, and their accuracies were assessed by 10-fold-cross-validation. </sec> <sec> <title>Results:</title> Blood test indexes and CT parameters were moderately to medially correlated. Of all selected features, lesion percentage contributed mostly to the classification of the two groups, followed by lesion volume, patient age, lymphocyte count, neutrophil count, GGO percentage and tumor heterogeneity. RF-assisted identification had the highest accuracy of 91.38%, followed by GNB (87.83%), KNN (87.93%), SVM (86.21%), MLP-NN (85.34%) and DT (84.48%). </sec> <sec> <title>Conclusions:</title> The RF-assisted model combining blood test and QCT parameters is helpful in the identification of severe COVID-19 cases. </sec>

The Emerging Role of Quantification of Imaging for Assessing the Severity and Disease Activity of Emphysema, Airway Disease, and Interstitial Lung Disease

There has been an explosion of use for quantitative image analysis in the setting of lung disease due to advances in acquisition protocols and postprocessing technology, including machine and deep learning. Despite the plethora of published papers, it is important to understand which approach has clinical validation and can be used in clinical practice. This paper provides an introduction to quantitative image analysis techniques being used in the investigation of lung disease and focusses on the techniques that have a reasonable clinical validation for being used in clinical trials and patient care.

CT image segmentation for inflamed and fibrotic lungs using a multi-resolution convolutional neural network

The purpose of this study was to develop a fully-automated segmentation algorithm, robust to various density enhancing lung abnormalities, to facilitate rapid quantitative analysis of computed tomography images. A polymorphic training approach is proposed, in which both specifically labeled left and right lungs of humans with COPD, and nonspecifically labeled lungs of animals with acute lung injury, were incorporated into training a single neural network. The resulting network is intended for predicting left and right lung regions in humans with or without diffuse opacification and consolidation. Performance of the proposed lung segmentation algorithm was extensively evaluated on CT scans of subjects with COPD, confirmed COVID-19, lung cancer, and IPF, despite no labeled training data of the latter three diseases. Lobar segmentations were obtained using the left and right lung segmentation as input to the LobeNet algorithm. Regional lobar analysis was performed using hierarchical clustering to identify radiographic subtypes of COVID-19. The proposed lung segmentation algorithm was quantitatively evaluated using semi-automated and manually-corrected segmentations in 87 COVID-19 CT images, achieving an average symmetric surface distance of [Formula: see text] mm and Dice coefficient of [Formula: see text]. Hierarchical clustering identified four radiographical phenotypes of COVID-19 based on lobar fractions of consolidated and poorly aerated tissue. Lower left and lower right lobes were consistently more afflicted with poor aeration and consolidation. However, the most severe cases demonstrated involvement of all lobes. The polymorphic training approach was able to accurately segment COVID-19 cases with diffuse consolidation without requiring COVID-19 cases for training.

Consensus Recommendations on the Use of 18F-FDG PET/CT in Lung Disease

PET with 18F-FDG has been increasingly applied, predominantly in the research setting, to study drug effects and pulmonary biology and to monitor disease progression and treatment outcomes in lung diseases that interfere with gas exchange through alterations of the pulmonary parenchyma, airways, or vasculature. To date, however, there are no widely accepted standard acquisition protocols or imaging data analysis methods for pulmonary 18F-FDG PET/CT in these diseases, resulting in disparate approaches. Hence, comparison of data across the literature is challenging. To help harmonize the acquisition and analysis and promote reproducibility, we collated details of acquisition protocols and analysis methods from 7 PET centers. From this information and our discussions, we reached the consensus recommendations given here on patient preparation, choice of dynamic versus static imaging, image reconstruction, and image analysis reporting.

A Shape Approximation for Medical Imaging Data

This study proposes a shape approximation approach to portray the regions of interest (ROI) from medical imaging data. An effective algorithm to achieve an optimal approximation is proposed based on the framework of Particle Swarm Optimization. The convergence of the proposed algorithm is derived under mild assumptions on the selected family of shape equations. The issue of detecting Parkinson’s disease (PD) based on the Tc-99m TRODAT-1 brain SPECT/CT images of 634 subjects, with 305 female and an average age of 68.3 years old from Kaohsiung Chang Gung Memorial Hospital, Taiwan, is employed to demonstrate the proposed procedure by fitting optimal ellipse and cashew-shaped equations in the 2D and 3D spaces, respectively. According to the visual interpretation of 3 experienced board-certified nuclear medicine physicians, 256 subjects are determined to be abnormal, 77 subjects are potentially abnormal, 174 are normal, and 127 are nearly normal. The coefficients of the ellipse and cashew-shaped equations, together with some well-known features of PD existing in the literature, are employed to learn PD classifiers under various machine learning approaches. A repeated hold-out with 100 rounds of 5-fold cross-validation and stratified sampling scheme is adopted to investigate the classification performances of different machine learning methods and different sets of features. The empirical results reveal that our method obtains 0.88 ± 0.04 classification accuracy, 0.87 ± 0.06 sensitivity, and 0.88 ± 0.08 specificity for test data when including the coefficients of the ellipse and cashew-shaped equations. Our findings indicate that more constructive and useful features can be extracted from proper mathematical representations of the 2D and 3D shapes for a specific ROI in medical imaging data, which shows their potential for improving the accuracy of automated PD identification.

Optimal threshold in low-dose CT quantification of emphysema

OBJECTIVE

Low-dose CT is now widely used in the screening of lung cancer and the detection of pulmonary nodules. There has also been increasing interest in using Low-dose CT for evaluating emphysema. In conventional dose CT, the threshold of -950HU is a common threshold for density-based emphysema quantification for worldwide population. However, the optimal threshold for assessing emphysema at low-dose CT has not been determined. The purpose of this study is to determine the optimal threshold for low-dose CT quantification of emphysema for Chinese population.

MATERIALS AND METHODS

In this study, 548 low-dose chest CT examinations acquired from different subjects (119 none, 49 mild, 163 moderate, 152 severe, and 65 very severe obstruction) are collected. At the level of the entire lung and individual lobes, the extent of emphysema was quantified by the percentage of the low attenuation area (LAA%) at a wide range of thresholds from -850HU to -1000HU. Both Pearson and Spearman's rank correlation coefficients were used to assess the correlations between 1) LAA% and pulmonary functions and 2) LAA% and the five-category classification. The statistical significance of the difference between correlation coefficients were evaluated using Steiger'Z test.

RESULTS

LAA% had a good correlation with both pulmonary function (|r| = 0.1-0.600, p < 0.001) and the five-category classification (r = 0.163-0.602, p < 0.001) in both the entire lung and individual lobes under different thresholds. The highest correlation coefficient is obtained at -940HU instead of -950HU.

CONCLUSION

Low-dose CT can be used for quantitative assessment of emphysema, and the threshold of -940HU is a suitable threshold for quantifying emphysema in low-dose CT images for Chinese population.

Lung involvement in patients with coronavirus disease-19 (COVID-19): a retrospective study based on quantitative CT findings

Purpose

To explore lung involvement in patients with coronavirus disease-19 (COVID-19) using quantitative computed tomography (QCT).

Methods

A total of 52 patients with COVID-19 who were admitted to three hospitals in China from January 23, 2020 to February 1, 2020 were retrospectively analyzed using QCT. The accuracy of QCT segmentation was assessed. The relationship between the time from symptom onset to initial CT and QCT parameters acquired on the initial CT were explored.

Results

First, the ability of QCT to detect and segment lesions was investigated and it was unveiled that results of segmentation of the majority of cases (42/52) were satisfactory and for 8 out of 52 patients, the images depicted lesions with miss-segmentation; besides, 2 out of 52 cases had negative finding on chest CT achieved by both radiologists and QCT. QCT-related parameters showed to have a relationship with the time from symptom onset to initial CT. In the early-stage (0-3 days), the percentage of lung involvement was 4%, with a mean density of - 462 ± 99 HU. The peak density of lesions appeared at the range of - 500 to - 700 HU on density histogram. In the intermediate-stage (4-6 days), the mean percentage of lung involvement noticeably increased compared with that in early stage (7%, p < 0.05). In late stage (7-14 days), the percentage of lung involvement decreased to 5%. The mean density of lesions was the highest (- 430 ± 80), and heterogeneity density distribution showed a dual-peak on density histogram.

Conclusion

COVID-19 can be promptly detected by QCT. In addition, the QCT-related parameters can highly facilitate assessment of pulmonary involvement.

Quantitative computed tomography analysis for stratifying the severity of Coronavirus Disease 2019

To examine the feasibility of using a computer tool for stratifying the severity of Coronavirus Disease 2019 (COVID-19) based on computed tomography (CT) images. We retrospectively examined 44 confirmed COVID-19 cases. All cases were evaluated separately by radiologists (visually) and through an in-house computer software. The degree of lesions was visually scored by the radiologist, as follows, for each of the 5 lung lobes: 0, no lesion present; 1, <1/3 involvement; 2, >1/3 and < 2/3 involvement; and 3, >2/3 involvement. Lesion density was assessed based on the proportion of ground-glass opacity (GGO), consolidation and fibrosis of the lesions. The parameters obtained using the computer tool included lung volume (mL), lesion volume (mL), lesion percentage (%), and mean lesion density (HU) of the whole lung, right lung, left lung, and each lobe. The scores obtained by the radiologists and quantitative results generated by the computer software were tested for correlation. A Chi-square test was used to test the consistency of radiologist- and computer-derived lesion percentage in the right/left lung, upper/lower lobe, and each of the 5 lobes. The results showed a strong to moderate correlation between lesion percentage scores obtained by radiologists and the computer software (r ranged from 0.7679 to 0.8373, P < 0.05), and a moderate correlation between the proportion of GGO and mean lesion density (r = −0.5894, P < 0.05), and proportion of consolidation and mean lesion density (r = 0.6282, P < 0.05). Computer-aided quantification showed a statistical significant higher lesion percentage for lower lobes than that assessed by the radiologists (χ² = 8.160, P = 0.004). Our experiments demonstrated that the computer tool could reliably and accurately assess the severity and distribution of pneumonia on CT scans.

•

Lesion percentage calculated by computer and radiologist are highly correlated.

•

Lesion density quantifiedby computer was correlated with the visually scored proportion of ground glass opacity and consolidation.

•

The computer tool could reliably and accurately assess the distribution of pneumonia.

Fully Automated Lung Lobe Segmentation in Volumetric Chest CT with 3D U-Net: Validation with Intra- and Extra-Datasets

Lung lobe segmentation in chest CT has been used for the analysis of lung functions and surgical planning. However, accurate lobe segmentation is difficult as 80% of patients have incomplete and/or fake fissures. Furthermore, lung diseases such as chronic obstructive pulmonary disease (COPD) can increase the difficulty of differentiating the lobar fissures. Lobar fissures have similar intensities to those of the vessels and airway wall, which could lead to segmentation error in automated segmentation. In this study, a fully automated lung lobe segmentation method with 3D U-Net was developed and validated with internal and external datasets. The volumetric chest CT scans of 196 normal and mild-to-moderate COPD patients from three centers were obtained. Each scan was segmented using a conventional image processing method and manually corrected by an expert thoracic radiologist to create gold standards. The lobe regions in the CT images were then segmented using a 3D U-Net architecture with a deep convolutional neural network (CNN) using separate training, validation, and test datasets. In addition, 40 independent external CT images were used to evaluate the model. The segmentation results for both the conventional and deep learning methods were compared quantitatively to the gold standards using four accuracy metrics including the Dice similarity coefficient (DSC), Jaccard similarity coefficient (JSC), mean surface distance (MSD), and Hausdorff surface distance (HSD). In internal validation, the segmentation method achieved high accuracy for the DSC, JSC, MSD, and HSD (0.97 ± 0.02, 0.94 ± 0.03, 0.69 ± 0.36, and 17.12 ± 11.07, respectively). In external validation, high accuracy was also obtained for the DSC, JSC, MSD, and HSD (0.96 ± 0.02, 0.92 ± 0.04, 1.31 ± 0.56, and 27.89 ± 7.50, respectively). This method took 6.49 ± 1.19 s and 8.61 ± 1.08 s for lobe segmentation of the left and right lungs, respectively. Although various automatic lung lobe segmentation methods have been developed, it is difficult to develop a robust segmentation method. However, the deep learning-based 3D U-Net method showed reasonable segmentation accuracy and computational time. In addition, this method could be adapted and applied to severe lung diseases in a clinical workflow.

Evaluation of dynamic lung changes during coronavirus disease 2019 (COVID-19) by quantitative computed tomography

OBJECTIVES

This study aims to trace the dynamic lung changes of coronavirus disease 2019 (COVID-19) using computed tomography (CT) images by a quantitative method.

METHODS

In this retrospective study, 28 confirmed COVID-19 cases with 145 CT scans are collected. The lesions are detected automatically and the parameters including lesion volume (LeV/mL), lesion percentage to lung volume (LeV%), mean lesion density (MLeD/HU), low attenuation area lower than - 400HU (LAA-400%), and lesion weight (LM/mL*HU) are computed for quantification. The dynamic changes of lungs are traced from the day of initial symptoms to the day of discharge. The lesion distribution among the five lobes and the dynamic changes in each lobe are also analyzed.

RESULTS

LeV%, MLeD, and LM reach peaks on days 9, 6 and 8, followed by a decrease trend in the next two weeks. LAA-400% (mostly the ground glass opacity) declines to the lowest on days 4-5, and then increases. The lesion is mostly seen in the bilateral lower lobes, followed by the left upper lobe, right upper lobe and right middle lobe (p < 0.05). The right middle lobe is the earliest one (on days 6-7), while the right lower lobe is the latest one (on days 9-10) that reaches to peak among the five lobes.

CONCLUSIONS

Severity of COVID-19 increases from the day of initial symptoms, reaches to the peak around on day 8, and then decreases. Lesion is more commonly seen in the bilateral lower lobes.

Risk stratification of acute pulmonary embolism based on the clot volume and right ventricular dysfunction on CT pulmonary angiography

OBJECTIVE

To test the feasibility of the clot volume and right ventricular dysfunction for risk stratification of acute pulmonary embolism (APE) patients.

METHODS

CT pulmonary angiography (CTPA) images of 158 APE patients were collected. After excluding 38 (24.1%) patients due to unsatisfactory quality, 120 APE patients (61 males and 59 females) were divided into high-risk (n = 37) and non-high-risk (n = 83) groups. Clot burden was measured by an automated programme (clot volume) and by two semi-quantitative systems (Qanadli and Mastora scores). The ratios of the right ventricular diameter to left ventricular diameter (RVd/LVd) and area (RVa/LVa) were obtained. The correlations amongst the above parameters were analysed. Receiver operating characteristic (ROC) curves were calculated to determine the efficacy of high-risk APE. Multivariate analyses were used to identify the independent predictors.

RESULTS

Strong positive correlations were found between the clot volume and both Qanadli score (r²  = 0.696, P < 0.001) and Mastora score (r²  = 0.728, P < 0.001), and moderate correlations were found between the clot volume and both RVd/LVd (r²  = 0.392, P < 0.001) and RVa/LVa (r²  = 0.389, P < 0.001). The clot volume contributed the highest efficacy (AUC = 0.992) for the identification of high-risk cases, followed by Mastora score (0.968), Qanadli score (0.952), RVa/LVa (0.900) and RVd/LVd (0.892). The clot volume and RVd/LVd were two independent factors of high-risk APE.

CONCLUSIONS

The clot volume is correlated with semi-quantitative clot burden scores and CT measured cardiac parameters. The clot volume and RVd/LVd were two independent factors of high-risk APE patients.

Role of Quantitative Computed Tomographic Scan Analysis in Lung Volume Reduction for Emphysema

Recent advances in bronchoscopic lung volume reduction (BLVR) offer new therapeutic alternatives for patients with emphysema and hyperinflation. Endobronchial valves and coils are 2 potential BLVR techniques which have been shown to improve pulmonary function and the quality of life in patients with emphysema. Current patient selection for LVR procedures relies on 3 main inclusion criteria: low attenuation area (in %), also known as emphysema score, heterogeneity score, and fissure integrity score. Volumetric analysis in combination with densitometric analysis of the affected lung lobe or segment with quantitative CT to determine emphysema severity play an important role in treatment planning and post-operative assessment. Due to the variations in lung anatomy, manual corrections are often required to ensure successful and accurate lobe segmentation for pathological and post-treatment CT scan analysis. The advanced development and utilisation of quantitative CT do not simply represent regional changes in pulmonary function but aids in analysis for better patient selection with severe emphysema who are most likely to benefit from BLVR.

Determination of vascular alteration in smokers by quantitative computed tomography measurements

A new method of quantitative computed tomography (CT) measurements of pulmonary vessels are applicable to morphological studies and may be helpful in defining the progression of emphysema in smokers. However, limited data are available on the relationship between the smoking status and pulmonary vessels alteration established in longitudinal observations. Therefore, we investigated the change of pulmonary vessels on CTs in a longitudinal cohort of smokers.Chest CTs were available for 287 current smokers, 439 non-smokers, and 80 former smokers who quit smoking at least 2 years after the baseline CT. CT images obtained at the baseline and 1 year later were assessed by a new quantitative CT measurement method, computing the total number of pulmonary vessels (TNV), mean lung density (MLD), and the percentage of low-attenuation areas at a threshold of -950 (density attenuation area [LAA]%950). Analysis of variance (ANOVA) and the independent sample t test were used to estimate the influence of the baseline parameters. The t paired test was employed to evaluate the change between the baseline and follow-up results.The current smokers related to have higher whole-lung MLD, as well as less and lower TNV values than the non-smokers (P <.05). But no significant differences in LAA%950 were found between smokers and non-smokers. After one year, the increase in LAA%950 was more rapid in the current (additional 0.3% per year, P <. 05-.01) than in the former smokers (additional 0.2% per year, P = .3). Additionally, the decline in TNV was faster in the current (additional -1.3 per year, P <.05-.01) than that in the former smokers (additional -0.2 per year, P = .6). Current smoke, pack-years, weight, and lung volume independently predicted TNV at baseline (P <.001) in multivariate analysis.The findings of this study reveal that the decline in the pulmonary vessels in smokers can be measured and related to their smoking status.

Morphological analysis of blood vessels near lung tumors using 3-D quantitative CT

BACKGROUND

Improved visualization of lung cancer-associated vessels is vital.

OBJECTIVE

To evaluate the efficacy of 3-D quantitative CT in lung cancer-associated pulmonary vessel assessment.

METHODS

Vascular CT changes were assessed visually and using FACT-Digital lung TM software (n = 162 patients, 178 controls). The total number of pulmonary vessels (TNV) and mean lumen area of pulmonary vessels (MAV) vertical to cross-sections of fifth/sixth-generation bronchioles were measured.

RESULTS

Visual investigation revealed fewer ipsilateral pulmonary vascular abnormalities in lung cancer (151/162) than did quantitative CT (162/162), and required more time (3.2±1.5 vs. 2.5±1.3 min) (P < 0.05). CT measurements revealed that the TNV vertical to the fifth-generation bronchial cross-section of the ipsilateral, contralateral, and control groups was 14.58±4.75, 9.58±3.74, and 10.22±4.07 and the MAV in these groups was 99.70±26.20, 58.76±29.29, and 57.76±18.32, respectively. The TNV vertical to the sixth-generation bronchial cross-section of the ipsilateral, contralateral, and control groups was 16.64±5.14, 11.59±4.06, and 11.75±4.16 and the MAV was 110.22±31.47, 67.62±30.41, and 60.24±16.18, respectively. The TNV and MAV in ipsilateral lung cancer tissues exceeded those in the contralateral side and control group tissues (P < 0.001).

CONCLUSIONS

Automated 3-D quantitative CT could successfully characterize pulmonary vessels and their lung cancer-associated changes.

Segmentation of pulmonary vascular tree by incorporating vessel enhancement filter and variational region-growing

BACKGROUND

Automatic segmentation of pulmonary vascular tree in the thoracic computed tomography (CT) image is a promising but challenging task with great clinical potential values. It is difficult to segment the whole vascular tree in reasonable time and acceptable accuracy.

OBJECTIVE

To develop a novel pulmonary vessel segmentation approach by incorporating vessel enhancement filters and the anisotropic diffusion filter with the variational region growing.

METHODS

First, the airway wall from the lung lobes is eliminated from CT images by using multi-scale morphological operations. Second, a Hessian-based multi-scale vesselness filter and medialness filter are applied to detect and enhance the potential vessel. Third, an anisotropic diffusion filter is used to remove noise and enhance the tube-like structures in CT images. Last, the vascular tree is segmented by applying variational region growing algorithm.

RESULTS

Applying to the CT images collected from the entire dataset of VESSEL12 challenge, we achieved an average sensitivity of 92.9%, specificity of 91.6% and the area under the ROC curve of AUC = 0.972.

CONCLUSIONS

This study demonstrated feasibility of segmenting the pulmonary vessel effectively by incorporating vessel enhancement filters and the anisotropic diffusion filter with the variational region growing algorithm. Our method cannot only segment both large and peripheral vessels, but also distinguish the vessels from the adjacent tissues, especially the airway walls.

Application of automated bronchial 3D-CT measurement in pulmonary contusion complicated with acute respiratory distress syndrome

BACKGROUNDQuantitative measurement of bronchial morphological changes in pulmonary contusion with acute respiratory distress syndrome (ARDS) has important clinical implications.OBJECTIVETo investigate the morphological changes in bronchus before and after treatment in patients with pulmonary contusion combined with ARDS using an automated bronchial three-dimensional computed tomography (3D-CT) measurement method.METHODSThe study involves a dataset of CT images of 62 patients diagnosed with pulmonary contusion combined with ARDS. The volume of pulmonary contusion lesions was calculated as a percentage of the total lung volume using the automated 3D-CT method. The bronchial luminal cross-sectional area, wall cross-sectional area, the maximum and average wall thickness, the maximum and average luminal densities, intraluminal and extraluminal diameters, and circumferences of generations 2-4 bronchi before and after treatment were measured. Furthermore, the corresponding differences were analyzed statistically.RESULTSThe luminal cross-sectional area, wall cross-sectional area, intraluminal and extraluminal diameters, and circumferences of generations 2-4 bronchi were all significantly lower before treatment than after treatment (P < 0.05). However, the maximum and average wall thicknesses were both significantly higher before treatment than after treatment (P < 0.05). No significant difference was found in the maximum and average luminal densities before and after treatment (P > 0.05). The percentage of the pulmonary contusion lesion volume to the total lung volume correlated positively with the thoracic trauma severity score (r = 0.74, P < 0.01).CONCLUSIONSQuantitative bronchial CT image analysis enables to detect and assess bronchial morphological changes in patients diagnosed with pulmonary contusion combined with ARDS.

FissureNet: A Deep Learning Approach For Pulmonary Fissure Detection in CT Images

Pulmonary fissure detection in computed tomography (CT) is a critical component for automatic lobar segmentation. The majority of fissure detection methods use feature descriptors that are hand-crafted, low-level, and have local spatial extent. The design of such feature detectors is typically targeted toward normal fissure anatomy, yielding low sensitivity to weak, and abnormal fissures that are common in clinical data sets. Furthermore, local features commonly suffer from low specificity, as the complex textures in the lung can be indistinguishable from the fissure when the global context is not considered. We propose a supervised discriminative learning framework for simultaneous feature extraction and classification. The proposed framework, called FissureNet, is a coarse-to-fine cascade of two convolutional neural networks. The coarse-to-fine strategy alleviates the challenges associated with training a network to segment a thin structure that represents a small fraction of the image voxels. FissureNet was evaluated on a cohort of 3706 subjects with inspiration and expiration 3DCT scans from the COPDGene clinical trial and a cohort of 20 subjects with 4DCT scans from a lung cancer clinical trial. On both data sets, FissureNet showed superior performance compared with a deep learning approach using the U-Net architecture and a Hessian-based fissure detection method in terms of area under the precision-recall curve (PR-AUC). The overall PR-AUC for FissureNet, U-Net, and Hessian on the COPDGene (lung cancer) data set was 0.980 (0.966), 0.963 (0.937), and 0.158 (0.182), respectively. On a subset of 30 COPDGene scans, FissureNet was compared with a recently proposed advanced fissure detection method called derivative of sticks (DoS) and showed superior performance with a PR-AUC of 0.991 compared with 0.668 for DoS.

Imaging Features of Chronic Bronchitis with Preserved Ratio and Impaired Spirometry (PRISm)

PURPOSE

The purpose of the study was to investigate the quantitative chest tomographic features of chronic bronchitis with preserved ratio and impaired spirometry (PRISm), including airway wall area, emphysema index, and lung capacity.

METHODS

An observational, cross-sectional study of 343 patients at the Ninth Hospital of Xi'an Affiliated Hospital of Xi'an Jiaotong University between October 2014 and September 2017. The patients were divided into three groups: 77 cases of chronic bronchitis with normal lung function (forced expiratory volume in one second/forced vital capacity) (FEV1/FVC > 70%, FEV1%pred > 80%), 80 cases of chronic bronchitis with PRISm (FEV1/FVC > 70%, FEV1%pred < 80%), and 186 cases of the early chronic obstructive pulmonary disease (COPD) (FEV1/FVC < 70%, FEV1%pred > 50%, that is, Global Initiative for Chronic Obstructive Lung Disease (GOLD) grade 1 + 2). We compared and analyzed the differences in imaging between the chronic bronchitis with PRISm and the other two groups.

RESULTS

Compared with the early COPD group, the PRISm group revealed significant differences in airway wall area, emphysema index, and lung capacity (P < 0.05). Compared with the chronic bronchitis with normal lung function group, the PRISm group showed increased WA%_LUL5, decreased lung capacity, and higher mean lung density.

CONCLUSION

In terms of airway wall area and emphysema index, patients with chronic bronchitis with PRISm were essentially no different than those with chronic bronchitis without abnormal spirometry, whereas for symptoms, they are more like GOLD 1 and 2 patients. Our findings show that it is not yet clear whether it constitutes an intermediate stage of chronic bronchitis with normal lung function that progression to early COPD.

Pulmonary lobe separation in expiration chest CT scans based on subject-specific priors derived from inspiration scans

Segmentation of pulmonary lobes in inspiration and expiration chest CT scan pairs is an important prerequisite for lobe-based quantitative disease assessment. Conventional methods process each CT scan independently, resulting typically in lower segmentation performance at expiration compared to inspiration. To address this issue, we present an approach, which utilizes CT scans at both respiratory states. It consists of two main parts: a base method that processes a single CT scan and an extended method that utilizes the segmentation result obtained on the inspiration scan as a subject-specific prior for segmentation of the expiration scan. We evaluated the methods on a diverse set of 40 CT scan pairs. In addition, we compare the performance of our method to a registration-based approach. On inspiration scans, the base method achieved an average distance error of 0.59, 0.64, and 0.91 mm for the left oblique, right oblique, and right horizontal fissures, respectively, when compared with expert-based reference tracings. On expiration scans, the base method's errors were 1.54, 3.24, and 3.34 mm, respectively. In comparison, utilizing proposed subject-specific priors for segmentation of expiration scans allowed decreasing average distance errors to 0.82, 0.79, and 1.04 mm, which represents a significant improvement ([Formula: see text]) compared with all other methods investigated.

Limitations of an automated embolism segmentation method in clinical practice

PURPOSE

Automated pulmonary embolism (PE) segmentation is frequently used as a preprocessing step in the quantitative analysis of pulmonary embolism. Objective of this study is to analyze the potential limitation in automated PE segmentation using clinical cases.

METHODS

A database of 304 computer tomography pulmonary angiography (CTPA) examinations was collected and confirmed to be PE. After processing using an automated scheme, two radiologists classified these cases into four groups of A, B, C and D, which represent 4 different segmentation results namely, (1) entire pulmonary artery identified without motivation artifacts, (2) entire pulmonary artery identified with motivation artifacts, (3) part of the pulmonary artery identified, and (4) none of the pulmonary artery identified. Then, the possible failed reasons in PE segmentation were analyzed and determined based on the image characterization of the diseases and the applied CTPA scanning protocols.

RESULTS

In the study, 143 (47.0%., 30 (9.9%., 110 (36.2%. and 21 (6.9%. examinations were classified into groups A, B, C and D, respectively. Group C and D included the cases with failed segmentation. Fifteen failure reasons, including intrapulmonary abnormalities, extra-pulmonary abnormalities, diffuse pulmonary diseases, enlarged heart, absolute occluded vessels, embolism attached to artery wall, delayed scan time, skewed location, low scan dose, obvious artifact of superior vena cava, previous chest surgery, congenital deformities of the chest, incorrect positioning, missed images and other unknown reasons, were determined with corresponding case percentages ranging from 0.3%.o 9.2%.

CONCLUSIONS

Automated segmentation failures were caused by specific lung diseases, anatomy varieties, improper scan time, improper scan dose, manual errors or other unknown reasons. Realization of those limitations is crucial for developing robust automated schemes to handle these issues in a single pass when a large number of CTPA examinations need to be analyzed.

Pulmonary Lobe Segmentation With Probabilistic Segmentation of the Fissures and a Groupwise Fissure Prior

A fully automated, unsupervised lobe segmentation algorithm is presented based on a probabilistic segmentation of the fissures and the simultaneous construction of a populationmodel of the fissures. A two-class probabilistic segmentation segments the lung into candidate fissure voxels and the surrounding parenchyma. This was combined with anatomical information and a groupwise fissure prior to drive non-parametric surface fitting to obtain the final segmentation. The performance of our fissure segmentation was validated on 30 patients from the chronic obstructive pulmonary disease COPDGene cohort, achieving a high median F₁ -score of 0.90 and showed general insensitivity to filter parameters. We evaluated our lobe segmentation algorithm on the Lobe and Lung Analysis 2011 dataset, which contains 55 cases at varying levels of pathology. We achieved the highest score of 0.884 of the automated algorithms. Our method was further tested quantitatively and qualitatively on 80 patients from the COPDgene study at varying levels of functional impairment. Accurate segmentation of the lobes is shown at various degrees of fissure incompleteness for 96% of all cases. We also show the utility of including a groupwise prior in segmenting the lobes in regions of grossly incomplete fissures.

Fast interactive segmentation of the pulmonary lobes from thoracic computed tomography data

Automated lung lobe segmentation methods often fail for challenging and clinically relevant cases with incomplete fissures or substantial amounts of pathology. We present a fast and intuitive method to interactively correct a given lung lobe segmentation or to quickly create a lobe segmentation from scratch based on a lung mask. A given lobar boundary is converted into a mesh by principal component analysis of 3D lobar boundary markers to obtain a plane where nodes correspond to the position of the markers. An observer can modify the mesh by drawing on 2D slices in arbitrary orientations. After each drawing, the mesh is immediately adapted in a 3D region around the user interaction. For evaluation we participated in the international lung lobe segmentation challenge LObe and lung analysis 2011 (LOLA11). Two observers applied the method to correct a given lung lobe segmentation obtained by a fully automatic method for all 55 CT scans of LOLA11. On average observer 1/2 required 8  ±  4/25  ±  12 interactions per case and took 1:30  ±  0:34/3:19  ±  1:29 min. The average distances to the reference segmentation were improved from an initial 2.68  ±  14.71 mm to 0.89  ±  1.63/0.74  ±  1.51 mm. In addition, one observer applied the proposed method to create a segmentation from scratch. This took 3:44  ±  0:58 minutes on average per case, applying an average of 20  ±  3 interactions to reach an average distance to the reference of 0.77  ±  1.14 mm. Thus, both the interactive corrections and the creation of a segmentation from scratch were feasible in a short time with excellent results and only little interaction. Since the mesh adaptation is independent of image features, the method can successfully handle patients with severe pathologies, provided that the human operator is capable of correctly indicating the lobar boundaries.

Large-scale image region documentation for fully automated image biomarker algorithm development and evaluation

With the advent of fully automated image analysis and modern machine learning methods, there is a need for very large image datasets having documented segmentations for both computer algorithm training and evaluation. This paper presents a method and implementation for facilitating such datasets that addresses the critical issue of size scaling for algorithm validation and evaluation; current evaluation methods that are usually used in academic studies do not scale to large datasets. This method includes protocols for the documentation of many regions in very large image datasets; the documentation may be incrementally updated by new image data and by improved algorithm outcomes. This method has been used for 5 years in the context of chest health biomarkers from low-dose chest CT images that are now being used with increasing frequency in lung cancer screening practice. The lung scans are segmented into over 100 different anatomical regions, and the method has been applied to a dataset of over 20,000 chest CT images. Using this framework, the computer algorithms have been developed to achieve over 90% acceptable image segmentation on the complete dataset.

A three-dimensional approach for identifying small pulmonary vessels in smokers

BACKGROUND

This study aims to develop a computerized scheme that utilizes a differential geometric approach to identify pulmonary vessels and then evaluate the performance of the scheme on the CT images of heavy smokers.

METHODS

The scheme consists of two primary steps to segment entire lung vascular tree and identify the number of pulmonary vessels in a cross section. The scheme performance including accuracy, consistency, and efficiency was assessed using 102 chest CT scans. Further assessment was performed on the relationship between pulmonary vessels and the extent of emphysema as well as pulmonary artery alteration.

RESULTS

The mean number of vessels in the cross section at the 5th generation was 17.84±4.74 and 17.23±4.85 assessed by computerized scheme and radiologists, respectively, which are significantly different (t = 2.12, p = 0.055). The results were consistent with those obtained by using a semi-automatic tool (r = 0.75, p = 0.01). In addition, in the 5th generation, the mean number of vessels was inversely related to the percentage of the low attenuation area (r = -0.704, p = 0.000), the mean lumen area of pulmonary vessel was inversely related to the mean value of main pulmonary artery diameter (r = -0.617, p = 0.000). The computational time of segmenting vessels was 6.50±0.02 seconds, which is much less than the average 8 minutes of the time spent by radiologists using the semi-automatic tool.

CONCLUSION

Applying the computerized scheme yields reasonable performance on the segmentation of pulmonary vessels. The alteration of pulmonary vessels may reflect the presence of pulmonary hypertension, as well as the extent of emphysema.

Computed tomography quantification of pulmonary vessels in chronic obstructive pulmonary disease as identified by 3D automated approach

The aim of this study was to investigate the vascular alteration of the whole lung and individual lobes in patients with COPD, and assess the association between pulmonary vessels and the extent and distribution of emphysema as well as pulmonary function by a 3-dimensional automated approach.A total of 83 computed tomography images from COPD patients were analyzed. Automated computerized approach was used to measure the total number of vessels at the fifth generation. The extent of emphysema (%LAA-950) in the whole lung and individual lobes were also calculated automatically. The association between the vascular number and the extent and distribution of emphysema, as well as the pulmonary function were assessed.Both the vascular number of fifth generation in the upper lobe and in the lower lobe were significantly negatively correlated with %LAA-950 (P < 0.05). Furthermore, there were significant, yet weak correlations between the vascular number and FEV1% predicted (R = 0.556, P = 0.039) and FEV1/FVC (R = 0.538, P = 0.047). In contrast, the vascular numbers were strongly correlated with DLco (R = 0.770, P = 0.003). Finally, the vascular number correlated closer with %LAA-950 of upper lobes than with %LAA-950 of lower lobes.Pulmonary vessel alteration can be measured; it is related to the extent of emphysema rather than the distribution of emphysema.

High-Accuracy Classification of Parkinson's Disease Through Shape Analysis and Surface Fitting in 123I-Ioflupane SPECT Imaging

Early and accurate identification of Parkinsonian syndromes (PS) involving presynaptic degeneration from nondegenerative variants such as scans without evidence of dopaminergic deficit (SWEDD) and tremor disorders is important for effective patient management as the course, therapy, and prognosis differ substantially between the two groups. In this study, we use single photon emission computed tomography (SPECT) images from healthy normal, early PD, and SWEDD subjects, as obtained from the Parkinson's Progression Markers Initiative (PPMI) database, and process them to compute shape- and surface-fitting-based features. We use these features to develop and compare various classification models that can discriminate between scans showing dopaminergic deficit, as in PD, from scans without the deficit, as in healthy normal or SWEDD. Along with it, we also compare these features with striatal binding ratio (SBR)-based features, which are well established and clinically used, by computing a feature-importance score using random forests technique. We observe that the support vector machine (SVM) classifier gives the best performance with an accuracy of 97.29%. These features also show higher importance than the SBR-based features. We infer from the study that shape analysis and surface fitting are useful and promising methods for extracting discriminatory features that can be used to develop diagnostic models that might have the potential to help clinicians in the diagnostic process.

Preoperative (3-dimensional) computed tomography lung reconstruction before anatomic segmentectomy or lobectomy for stage I non-small cell lung cancer

OBJECTIVES

Accurate cancer localization and negative resection margins are necessary for successful segmentectomy. In this study, we evaluate a newly developed software package that permits automated segmentation of the pulmonary parenchyma, allowing 3-dimensional assessment of tumor size, location, and estimates of surgical margins.

METHODS

A pilot study using a newly developed 3-dimensional computed tomography analytic software package was performed to retrospectively evaluate preoperative computed tomography images of patients who underwent segmentectomy (n = 36) or lobectomy (n = 15) for stage 1 non-small cell lung cancer. The software accomplishes an automated reconstruction of anatomic pulmonary segments of the lung based on bronchial arborization. Estimates of anticipated surgical margins and pulmonary segmental volume were made on the basis of 3-dimensional reconstruction.

RESULTS

Autosegmentation was achieved in 72.7% (32/44) of preoperative computed tomography images with slice thicknesses of 3 mm or less. Reasons for segmentation failure included local severe emphysema or pneumonitis, and lower computed tomography resolution. Tumor segmental localization was achieved in all autosegmented studies. The 3-dimensional computed tomography analysis provided a positive predictive value of 87% in predicting a marginal clearance greater than 1 cm and a 75% positive predictive value in predicting a margin to tumor diameter ratio greater than 1 in relation to the surgical pathology assessment.

CONCLUSIONS

This preoperative 3-dimensional computed tomography analysis of segmental anatomy can confirm the tumor location within an anatomic segment and aid in predicting surgical margins. This 3-dimensional computed tomography information may assist in the preoperative assessment regarding the suitability of segmentectomy for peripheral lung cancers.

Review of automatic pulmonary lobe segmentation methods from CT

The computational detection of pulmonary lobes from CT images is a challenging segmentation problem with important respiratory health care applications, including surgical planning and regional image analysis. Several authors have proposed automated algorithms and we present a methodological review. These algorithms share a number of common stages and we consider each stage in turn, comparing the methods applied by each author and discussing their relative strengths. No standard method has yet emerged and none of the published methods have been demonstrated across a full range of clinical pathologies and imaging protocols. We discuss how improved methods could be developed by combining different approaches, and we use this to propose a workflow for the development of new algorithms.

Impact of emphysema heterogeneity on pulmonary function

Objectives

To investigate the association between emphysema heterogeneity in spatial distribution, pulmonary function and disease severity.

Methods and Materials

We ascertained a dataset of anonymized Computed Tomography (CT) examinations acquired on 565 participants in a COPD study. Subjects with chronic bronchitis (CB) and/or bronchodilator response were excluded resulting in 190 cases without COPD and 160 cases with COPD. Low attenuations areas (LAAs) (≤950 Hounsfield Unit (HU)) were identified and quantified at the level of individual lobes. Emphysema heterogeneity was defined in a manner that ranged in value from −100% to 100%. The association between emphysema heterogeneity and pulmonary function measures (e.g., FEV1% predicted, RV/TLC, and DLco% predicted) adjusted for age, sex, and smoking history (pack-years) was assessed using multiple linear regression analysis.

Results

The majority (128/160) of the subjects with COPD had a heterogeneity greater than zero. After adjusting for age, gender, smoking history, and extent of emphysema, heterogeneity in depicted disease in upper lobe dominant cases was positively associated with pulmonary function measures, such as FEV1 Predicted (p<.001) and FEV1/FVC (p<.001), as well as disease severity (p<0.05). We found a negative association between HI% , RV/TLC (p<0.001), and DLco% (albeit not a statistically significant one, p = 0.06) in this group of patients.

Conclusion

Subjects with more homogeneous distribution of emphysema and/or lower lung dominant emphysema tend to have worse pulmonary function.

Towards local progression estimation of pulmonary emphysema using CT

PURPOSE

Whole lung densitometry on chest CT images is an accepted method for measuring tissue destruction in patients with pulmonary emphysema in clinical trials. Progression measurement is required for evaluation of change in health condition and the effect of drug treatment. Information about the location of emphysema progression within the lung may be important for the correct interpretation of drug efficacy, or for determining a treatment plan. The purpose of this study is therefore to develop and validate methods that enable the local measurement of lung density changes, which requires proper modeling of the effect of respiration on density.

METHODS

Four methods, all based on registration of baseline and follow-up chest CT scans, are compared. The first naïve method subtracts registered images. The second employs the so-called dry sponge model, where volume correction is performed using the determinant of the Jacobian of the transformation. The third and the fourth introduce a novel adaptation of the dry sponge model that circumvents its constant-mass assumption, which is shown to be invalid. The latter two methods require a third CT scan at a different inspiration level to estimate the patient-specific density-volume slope, where one method employs a global and the other a local slope. The methods were validated on CT scans of a phantom mimicking the lung, where mass and volume could be controlled. In addition, validation was performed on data of 21 patients with pulmonary emphysema.

RESULTS

The image registration method was optimized leaving a registration error below half the slice increment (median 1.0 mm). The phantom study showed that the locally adapted slope model most accurately measured local progression. The systematic error in estimating progression, as measured on the phantom data, was below 2 gr/l for a 70 ml (6%) volume difference, and 5 gr/l for a 210 ml (19%) difference, if volume correction was applied. On the patient data an underlying linearity assumption relating lung volume change with density change was shown to hold (fitR(2) = 0.94), and globalized versions of the local models are consistent with global results (R(2) of 0.865 and 0.882 for the two adapted slope models, respectively).

CONCLUSIONS

In conclusion, image matching and subsequent analysis of differences according to the proposed lung models (i) has good local registration accuracy on patient data, (ii) effectively eliminates a dependency on inspiration level at acquisition time, (iii) accurately predicts progression in phantom data, and (iv) is reasonably consistent with global results in patient data. It is therefore a potential future tool for assessing local emphysema progression in drug evaluation trials and in clinical practice.

Pulmonary lobe segmentation based on ridge surface sampling and shape model fitting

PURPOSE

Performing lobe-based quantitative analysis of the lung in computed tomography (CT) scans can assist in efforts to better characterize complex diseases such as chronic obstructive pulmonary disease (COPD). While airways and vessels can help to indicate the location of lobe boundaries, segmentations of these structures are not always available, so methods to define the lobes in the absence of these structures are desirable.

METHODS

The authors present a fully automatic lung lobe segmentation algorithm that is effective in volumetric inspiratory and expiratory computed tomography (CT) datasets. The authors rely on ridge surface image features indicating fissure locations and a novel approach to modeling shape variation in the surfaces defining the lobe boundaries. The authors employ a particle system that efficiently samples ridge surfaces in the image domain and provides a set of candidate fissure locations based on the Hessian matrix. Following this, lobe boundary shape models generated from principal component analysis (PCA) are fit to the particles data to discriminate between fissure and nonfissure candidates. The resulting set of particle points are used to fit thin plate spline (TPS) interpolating surfaces to form the final boundaries between the lung lobes.

RESULTS

The authors tested algorithm performance on 50 inspiratory and 50 expiratory CT scans taken from the COPDGene study. Results indicate that the authors' algorithm performs comparably to pulmonologist-generated lung lobe segmentations and can produce good results in cases with accessory fissures, incomplete fissures, advanced emphysema, and low dose acquisition protocols. Dice scores indicate that only 29 out of 500 (5.85%) lobes showed Dice scores lower than 0.9. Two different approaches for evaluating lobe boundary surface discrepancies were applied and indicate that algorithm boundary identification is most accurate in the vicinity of fissures detectable on CT.

CONCLUSIONS

The proposed algorithm is effective for lung lobe segmentation in absence of auxiliary structures such as vessels and airways. The most challenging cases are those with mostly incomplete, absent, or near-absent fissures and in cases with poorly revealed fissures due to high image noise. However, the authors observe good performance even in the majority of these cases.

A hybrid approach to segmentation of diseased lung lobes

Complete segmentation of diseased lung lobes by automatically identifying fissure surfaces is a nontrivial task, due to incomplete, disrupted, and deformed fissures. In this paper, we present a novel algorithm employing a hybrid two-dimensional/three-dimensional approach for segmenting diseased lung lobes. Our approach models complete fissure surfaces from partial fissures found in individual computed tomography (CT) images. Evaluated using 24 patients' lungs with a variety of different diseases, our algorithm produced root-mean square errors of 2.21 ± 1.21, 2.51 ± 1.36, and 2.38 ± 1.27 mm for segmenting the left oblique fissure (LOF), right oblique fissure (ROF) and right horizontal fissure (RHF), respectively. The average accuracies for segmenting the LOF, ROF, and RHF are 86.59%, 84.80%, and 82.62%, using our ±3-mm percentile measure. These results indicate the feasibility of developing an automatic algorithm for complete segmentation of diseased lung lobes.

Pulmonary fissure integrity and collateral ventilation in COPD patients

Purpose

To investigate whether the integrity (completeness) of pulmonary fissures affects pulmonary function in patients with chronic obstructive pulmonary disease (COPD).

Materials and Methods

A dataset consisting of 573 CT exams acquired on different subjects was collected from a COPD study. According to the global initiative for chronic obstructive lung disease (GOLD) criteria, these subjects (examinations) were classified into five different subgroups, namely non-COPD (222 subjects), GOLD-I (83 subjects), GOLD-II (141 subjects), GOLD-III (63 subjects), and GOLD-IV (64 subjects), in terms of disease severity. An available computer tool was used to aid in an objective and efficient quantification of fissure integrity. The correlations between fissure integrity, and pulmonary functions (e.g., FEV1, and FEV1/FVC) and COPD severity were assessed using Pearson and Spearman's correlation coefficients, respectively.

Results

For the five sub-groups ranging from non-COPD to GOLD-IV, the average integrities of the right oblique fissure (ROF) were 81.8%, 82.4%, 81.8%, 82.8%, and 80.2%, respectively; the average integrities of the right horizontal fissure (RHF) were 62.6%, 61.8%, 62.1%, 62.2%, and 62.3%, respectively; the average integrities of the left oblique fissure (LOF) were 82.0%, 83.2%, 81.7%, 82.0%, and 78.4%, respectively; and the average integrities of all fissures in the entire lung were 78.0%, 78.6%, 78.1%, 78.5%, and 76.4%, respectively. Their Pearson correlation coefficients with FEV1 and FE1/FVC range from 0.027 to 0.248 with p values larger than 0.05. Their Spearman correlation coefficients with COPD severity except GOLD-IV range from −0.013 to −0.073 with p values larger than 0.08.

Conclusion

There is no significant difference in fissure integrity for patients with different levels of disease severity, suggesting that the development of COPD does not change the completeness of pulmonary fissures and incomplete fissures alone may not contribute to the collateral ventilation.

Automated segmentation of pulmonary structures in thoracic computed tomography scans: a review

Computed tomography (CT) is the modality of choice for imaging the lungs in vivo. Sub-millimeter isotropic images of the lungs can be obtained within seconds, allowing the detection of small lesions and detailed analysis of disease processes. The high resolution of thoracic CT and the high prevalence of lung diseases require a high degree of automation in the analysis pipeline. The automated segmentation of pulmonary structures in thoracic CT has been an important research topic for over a decade now. This systematic review provides an overview of current literature. We discuss segmentation methods for the lungs, the pulmonary vasculature, the airways, including airway tree construction and airway wall segmentation, the fissures, the lobes and the pulmonary segments. For each topic, the current state of the art is summarized, and topics for future research are identified.

Bidirectional elastic image registration using B-spline affine transformation

A registration scheme termed as B-spline affine transformation (BSAT) is presented in this study to elastically align two images. We define an affine transformation instead of the traditional translation at each control point. Mathematically, BSAT is a generalized form of the affine transformation and the traditional B-spline transformation (BST). In order to improve the performance of the iterative closest point (ICP) method in registering two homologous shapes but with large deformation, a bidirectional instead of the traditional unidirectional objective/cost function is proposed. In implementation, the objective function is formulated as a sparse linear equation problem, and a sub-division strategy is used to achieve a reasonable efficiency in registration. The performance of the developed scheme was assessed using both two-dimensional (2D) synthesized dataset and three-dimensional (3D) volumetric computed tomography (CT) data. Our experiments showed that the proposed B-spline affine model could obtain reasonable registration accuracy.

Pulmonary lobar volumetry using novel volumetric computer-aided diagnosis and computed tomography

OBJECTIVES

To compare the accuracy of pulmonary lobar volumetry using the conventional number of segments method and novel volumetric computer-aided diagnosis using 3D computed tomography images.

METHODS

We acquired 50 consecutive preoperative 3D computed tomography examinations for lung tumours reconstructed at 1-mm slice thicknesses. We calculated the lobar volume and the emphysematous lobar volume < -950 HU of each lobe using (i) the slice-by-slice method (reference standard), (ii) number of segments method, and (iii) semi-automatic and (iv) automatic computer-aided diagnosis. We determined Pearson correlation coefficients between the reference standard and the three other methods for lobar volumes and emphysematous lobar volumes. We also compared the relative errors among the three measurement methods.

RESULTS

Both semi-automatic and automatic computer-aided diagnosis results were more strongly correlated with the reference standard than the number of segments method. The correlation coefficients for automatic computer-aided diagnosis were slightly lower than those for semi-automatic computer-aided diagnosis because there was one outlier among 50 cases (2%) in the right upper lobe and two outliers among 50 cases (4%) in the other lobes. The number of segments method relative error was significantly greater than those for semi-automatic and automatic computer-aided diagnosis (P < 0.001). The computational time for automatic computer-aided diagnosis was 1/2 to 2/3 than that of semi-automatic computer-aided diagnosis.

CONCLUSIONS

A novel lobar volumetry computer-aided diagnosis system could more precisely measure lobar volumes than the conventional number of segments method. Because semi-automatic computer-aided diagnosis and automatic computer-aided diagnosis were complementary, in clinical use, it would be more practical to first measure volumes by automatic computer-aided diagnosis, and then use semi-automatic measurements if automatic computer-aided diagnosis failed.

Assessment of lung volume collapsibility in chronic obstructive lung disease patients using CT

OBJECTIVE

To investigate the collapsibility of the lung and individual lobes in patients with COPD during inspiration/expiration and assess the association of whole lung and lobar volume changes with pulmonary function tests (PFTs) and disease severity.

METHODS

PFT measures used were RV/TLC%, FEV1% predicted, FVC, FEV1/FVC%, DLco% predicted and GOLD category. A total of 360 paired inspiratory and expiratory CT examinations acquired in 180 subjects were analysed. Automated computerised algorithms were used to compute individual lobe and total lung volumes. Lung volume collapsibility was assessed quantitatively using the simple difference between CT computed inspiration (I) and expiration (E) volumes (I-E), and a relative measure of volume changes, (I-E)/I.

RESULTS

Mean absolute collapsibility (I-E) decreased in all lung lobes with increasing disease severity defined by GOLD classification. Relative collapsibility (I-E)/I showed a similar trend. Upper lobes had lower volume collapsibility across all GOLD categories and lower lobes collectively had the largest volume collapsibility. Whole lung and left lower lobe collapsibility measures tended to have the highest correlations with PFT measures. Collapsibility of lung lobes and whole lung was also negatively correlated with the degree of air trapping between expiration and inspiration, as measured by mean lung density. All measured associations were statistically significant (P < 0.01).

CONCLUSION

Severity of COPD appears associated with increased collapsibility in the upper lobes, but change (decline) in collapsibility is faster in the lower lobes.

KEY POINTS

• Inspiratory and expiratory computed tomography allows assessment of lung collapsibility • Lobe volume collapsibility is significantly correlated with measures of lung function. • As COPD severity increases, collapsibility of individual lung lobes decreases. • Upper lobes exhibit more severe disease, while lower lobes decline faster.