Optimal graph search based segmentation of airway tree double surfaces across bifurcations

A Novel Predictive Method Incorporating Parameters of Main Pulmonary Artery Bifurcation for Short-Term Prognosis in Non-high-risk Acute Pulmonary Embolism Patients

The aim of this study was to build a formula to predict short-term prognosis using main pulmonary artery (MPA) parameters reconstructed from computed tomographic pulmonary angiography in non-high-risk acute pulmonary embolism (PE) patients. After reconstructing the MPA and its centerline, the MPA, the right and left pulmonary artery inlet, and the MPA outlet plane were differentiated to measure the cross-sectional area (CSA), the maximal diameter and the hydraulic diameter. The MPA bifurcation area, volume and angle were measured. MPA dilation was defined as >29 mm at the transverse section plane. The patients were randomly divided into a training set and a validation set. A least absolute shrinkage and selection operator (LASSO) logistic regression algorithm was used to build a predictive formula. The performances of the predictive formula from LASSO were tested by the area under the receiver operating characteristic curve (AUC) and precision-recall (PR) curve with 10-fold cross-validation. The clinical utility was assessed by decision curve analysis (DCA). In total, 296 patients were enrolled and randomly divided (50:50) into a training set and a validation set. The LASSO predictive formula (lambda.1SE) was as follows: 0.92 × MPA bifurcation area + 0.50 × MPA outlet hydraulic diameter + 0.10 × MPA outlet CSA. The AUCs of the predictive formula were 0.860 (95% CI: 0.795-0.912) and 0.943 (95% CI: 0.892-0.975) in the training set and validation set, respectively. The LASSO predictive formula had a higher average area under the PR curve than MPA dilation (0.71 vs. 0.23 in the training set and 0.55 vs. 0.23 in the validation set) and added a net benefit in clinical utility by DCA. Integration of MPA outlet CSA, hydraulic diameter, and bifurcation area with the LASSO predictive formula as a novel weighting method facilitated the prediction of poor short-term prognosis within 30 days after hospital admission in non-high-risk acute PE patients.

Generative-based airway and vessel morphology quantification on chest CT images

Accurately and precisely characterizing the morphology of small pulmonary structures from Computed Tomography (CT) images, such as airways and vessels, is becoming of great importance for diagnosis of pulmonary diseases. The smaller conducting airways are the major site of increased airflow resistance in chronic obstructive pulmonary disease (COPD), while accurately sizing vessels can help identify arterial and venous changes in lung regions that may determine future disorders. However, traditional methods are often limited due to image resolution and artifacts. We propose a Convolutional Neural Regressor (CNR) that provides cross-sectional measurement of airway lumen, airway wall thickness, and vessel radius. CNR is trained with data created by a generative model of synthetic structures which is used in combination with Simulated and Unsupervised Generative Adversarial Network (SimGAN) to create simulated and refined airways and vessels with known ground-truth. For validation, we first use synthetically generated airways and vessels produced by the proposed generative model to compute the relative error and directly evaluate the accuracy of CNR in comparison with traditional methods. Then, in-vivo validation is performed by analyzing the association between the percentage of the predicted forced expiratory volume in one second (FEV1%) and the value of the Pi10 parameter, two well-known measures of lung function and airway disease, for airways. For vessels, we assess the correlation between our estimate of the small-vessel blood volume and the lungs' diffusing capacity for carbon monoxide (DLCO). The results demonstrate that Convolutional Neural Networks (CNNs) provide a promising direction for accurately measuring vessels and airways on chest CT images with physiological correlates.

Optimal surface segmentation with convex priors in irregularly sampled space

Optimal surface segmentation is a state-of-the-art method used for segmentation of multiple globally optimal surfaces in volumetric datasets. The method is widely used in numerous medical image segmentation applications. However, nodes in the graph based optimal surface segmentation method typically encode uniformly distributed orthogonal voxels of the volume. Thus the segmentation cannot attain an accuracy greater than a single unit voxel, i.e. the distance between two adjoining nodes in graph space. Segmentation accuracy higher than a unit voxel is achievable by exploiting partial volume information in the voxels which shall result in non-equidistant spacing between adjoining graph nodes. This paper reports a generalized graph based multiple surface segmentation method with convex priors which can optimally segment the target surfaces in an irregularly sampled space. The proposed method allows non-equidistant spacing between the adjoining graph nodes to achieve subvoxel segmentation accuracy by utilizing the partial volume information in the voxels. The partial volume information in the voxels is exploited by computing a displacement field from the original volume data to identify the subvoxel-accurate centers within each voxel resulting in non-equidistant spacing between the adjoining graph nodes. The smoothness of each surface modeled as a convex constraint governs the connectivity and regularity of the surface. We employ an edge-based graph representation to incorporate the necessary constraints and the globally optimal solution is obtained by computing a minimum s-t cut. The proposed method was validated on 10 intravascular multi-frame ultrasound image datasets for subvoxel segmentation accuracy. In all cases, the approach yielded highly accurate results. Our approach can be readily extended to higher-dimensional segmentations.

Multiple surface segmentation using convolution neural nets: application to retinal layer segmentation in OCT images

Automated segmentation of object boundaries or surfaces is crucial for quantitative image analysis in numerous biomedical applications. For example, retinal surfaces in optical coherence tomography (OCT) images play a vital role in the diagnosis and management of retinal diseases. Recently, graph based surface segmentation and contour modeling have been developed and optimized for various surface segmentation tasks. These methods require expertly designed, application specific transforms, including cost functions, constraints and model parameters. However, deep learning based methods are able to directly learn the model and features from training data. In this paper, we propose a convolutional neural network (CNN) based framework to segment multiple surfaces simultaneously. We demonstrate the application of the proposed method by training a single CNN to segment three retinal surfaces in two types of OCT images - normal retinas and retinas affected by intermediate age-related macular degeneration (AMD). The trained network directly infers the segmentations for each B-scan in one pass. The proposed method was validated on 50 retinal OCT volumes (3000 B-scans) including 25 normal and 25 intermediate AMD subjects. Our experiment demonstrated statistically significant improvement of segmentation accuracy compared to the optimal surface segmentation method with convex priors (OSCS) and two deep learning based UNET methods for both types of data. The average computation time for segmenting an entire OCT volume (consisting of 60 B-scans each) for the proposed method was 12.3 seconds, demonstrating low computation costs and higher performance compared to the graph based optimal surface segmentation and UNET based methods.

A Framework for Classification and Segmentation of Branch Retinal Artery Occlusion in SD-OCT

Branch retinal artery occlusion (BRAO) is an ocular emergency, which could lead to blindness. Quantitative analysis of the BRAO region in the retina is necessary for the assessment of the severity of retinal ischemia. In this paper, a fully automatic framework was proposed to segment BRAO regions based on 3D spectral-domain optical coherence tomography (SD-OCT) images. To the best of our knowledge, this is the first automatic 3D BRAO segmentation framework. First, the input 3D image is automatically classified into BRAO of acute phase and BRAO of chronic phase or normal retina using an AdaBoost classifier based on combining local structural, intensity, textural features with our new feature distribution analyzing strategy. Then, BRAO regions of acute phase and chronic phase are segmented separately. A thickness model is built to segment BRAO in the chronic phase. While for segmenting BRAO in the acute phase, a two-step segmentation strategy is performed: rough initialization and refine segmentation. The proposed method was tested on SD-OCT images of 35 patients (12 BRAO acute phase, 11 BRAO chronic phase, and 12 normal eyes) using the leave-one-out strategy. The classification accuracy for BRAO acute phase, BRAO chronic phase, and normal retina were 100%, 90.9%, and 91.7%, respectively. The overall true positive volume fraction (TPVF) and false positive volume fraction (FPVF) for the acute phase were 91.1% and 5.5% and for the chronic phase were 92.7% and 8.4%, respectively.

Airway tree reconstruction in expiration chest CT scans facilitated by information transfer from corresponding inspiration scans

PURPOSE

Analysis and comparison of airways imaged in pairs of inspiration and expiration lung CT scans provides important information for quantitative assessment of lung diseases like chronic obstructive pulmonary disease. However, airway tree reconstruction in expiration CT scans is a challenging problem. Typically, only a low number of airway branches are found in expiration scans, compared to inspiration scans. To detect more airways in expiration CT scans, the authors introduce a novel airway reconstruction approach and assess its performance.

METHODS

The method requires a pair of inspiration and expiration CT scans and utilizes information from the inspiration scan to facilitate reconstructing the airway tree in the expiration lung CT scan. First, an initial airway tree (high confidence) and airway candidates (limited confidence) are reconstructed in the expiration scan by utilizing a 3D graph-based reconstruction method. Then, the 3D airway tree is reconstructed in the inspiration scan. Second, correspondences between expiration and inspiration tree structures are identified by utilizing a novel hierarchical tree matching algorithm that utilizes a local CT image-based similarity criterion. Third, the tree information from the inspiration airway tree is used to select expiration candidates, resulting in the final expiration tree structure. The approach was evaluated on a diverse set of 40 scan pairs and compared to the baseline method, which utilizes only the expiration CT scan.

RESULTS

The proposed method produced a significant (p < 0.05) increase in airway tree length by 13.35 cm, on average, which represents an 11.21% increase relative to the baseline result using only the expiration CT scan. A detailed analysis of all additionally identified airways resulted in a true and false positive rate of 94.8% and 5.2%, respectively. The true positive rate was found to be significantly higher than the false positive rate (p < 0.05).

CONCLUSIONS

The proposed method allowed increasing the number of found airways in expiration scans significantly. In addition, the algorithm establishes correspondence between inspiration and expiration airway trees, which can facilitate label transfer between airway trees and quantitative assessment of change in airways. The approach can be adapted to facilitate airway reconstruction in several longitudinal lung CT scans by means of mutual information transfer.

A robust statistics driven volume-scalable active contour for segmenting anatomical structures in volumetric medical images with complex conditions

BACKGROUND

Accurate segmentation of anatomical structures in medical images is a critical step in the development of computer assisted intervention systems. However, complex image conditions, such as intensity inhomogeneity, noise and weak object boundary, often cause considerable difficulties in medical image segmentation. To cope with these difficulties, we propose a novel robust statistics driven volume-scalable active contour framework, to extract desired object boundary from magnetic resonance (MR) and computed tomography (CT) imagery in 3D.

METHODS

We define an energy functional in terms of the initial seeded labels and two fitting functions that are derived from object local robust statistics features. This energy is then incorporated into a level set scheme which drives the active contour evolving and converging at the desired position of the object boundary. Due to the local robust statistics and the volume scaling function in the energy fitting term, the object features in local volumes are learned adaptively to guide the motion of the contours, which thereby guarantees the capability of our method to cope with intensity inhomogeneity, noise and weak boundary. In addition, the initialization of active contour is simplified by select several seeds in the object and/or background to eliminate the sensitivity to initialization.

RESULTS

The proposed method was applied to extensive public available volumetric medical images with challenging image conditions. The segmentation results of various anatomical structures, such as white matter (WM), atrium, caudate nucleus and brain tumor, were evaluated quantitatively by comparing with the corresponding ground truths. It was found that the proposed method achieves consistent and coherent segmentation accuracy of 0.9246 ± 0.0068 for WM, 0.9043 ± 0.0131 for liver tumors, 0.8725 ± 0.0374 for caudate nucleus, 0.8802 ± 0.0595 for brain tumors, etc., measured by Dice similarity coefficients value for the overlap between the algorithm one and the ground truth. Further comparative experimental results showed desirable performances of the proposed method over several well-known segmentation methods in terms of accuracy and robustness.

CONCLUSION

We proposed an approach to accurate segment volumetric medical images with complex conditions. The accuracy of segmentation, robustness to noise and contour initialization were validated on the basis of extensive MR and CT volumes.

Carotid Artery Wall Segmentation in Multispectral MRI by Coupled Optimal Surface Graph Cuts

We present a new three-dimensional coupled optimal surface graph-cut algorithm to segment the wall of the carotid artery bifurcation from Magnetic Resonance (MR) images. The method combines the search for both inner and outer borders into a single graph cut and uses cost functions that integrate information from multiple sequences. Our approach requires manual localization of only three seed points indicating the start and end points of the segmentation in the internal, external, and common carotid artery. We performed a quantitative validation using images of 57 carotid arteries. Dice overlap of 0.86 ± 0.06 for the complete vessel and 0.89 ± 0.05 for the lumen compared to manual annotation were obtained. Reproducibility tests were performed in 60 scans acquired with an interval of 15 ± 9 days, showing good agreement between baseline and follow-up segmentations with intraclass correlations of 0.96 and 0.74 for the lumen and complete vessel volumes respectively.

Advanced image processing for optical coherence tomographic angiography of macular diseases

This article provides an overview of advanced image processing for three dimensional (3D) optical coherence tomographic (OCT) angiography of macular diseases, including age-related macular degeneration (AMD) and diabetic retinopathy (DR). A fast automated retinal layers segmentation algorithm using directional graph search was introduced to separates 3D flow data into different layers in the presence of pathologies. Intelligent manual correction methods are also systematically addressed which can be done rapidly on a single frame and then automatically propagated to full 3D volume with accuracy better than 1 pixel. Methods to visualize and analyze the abnormalities including retinal and choroidal neovascularization, retinal ischemia, and macular edema were presented to facilitate the clinical use of OCT angiography.

A hybrid method for airway segmentation and automated measurement of bronchial wall thickness on CT

Inflammatory and infectious lung diseases commonly involve bronchial airway structures and morphology, and these abnormalities are often analyzed non-invasively through high resolution computed tomography (CT) scans. Assessing airway wall surfaces and the lumen are of great importance for diagnosing pulmonary diseases. However, obtaining high accuracy from a complete 3-D airway tree structure can be quite challenging. The airway tree structure has spiculated shapes with multiple branches and bifurcation points as opposed to solid single organ or tumor segmentation tasks in other applications, hence, it is complex for manual segmentation as compared with other tasks. For computerized methods, a fundamental challenge in airway tree segmentation is the highly variable intensity levels in the lumen area, which often causes a segmentation method to leak into adjacent lung parenchyma through blurred airway walls or soft boundaries. Moreover, outer wall definition can be difficult due to similar intensities of the airway walls and nearby structures such as vessels. In this paper, we propose a computational framework to accurately quantify airways through (i) a novel hybrid approach for precise segmentation of the lumen, and (ii) two novel methods (a spatially constrained Markov random walk method (pseudo 3-D) and a relative fuzzy connectedness method (3-D)) to estimate the airway wall thickness. We evaluate the performance of our proposed methods in comparison with mostly used algorithms using human chest CT images. Our results demonstrate that, on publicly available data sets and using standard evaluation criteria, the proposed airway segmentation method is accurate and efficient as compared with the state-of-the-art methods, and the airway wall estimation algorithms identified the inner and outer airway surfaces more accurately than the most widely applied methods, namely full width at half maximum and phase congruency.

Graph-Based Airway Tree Reconstruction From Chest CT Scans: Evaluation of Different Features on Five Cohorts

We present a graph-based framework for airway tree reconstruction from computerized tomography (CT) scans and evaluate the performance of different feature categories and their combinations on five lung cohorts. The approach consists of two main processing steps. First, potential airway branch and connection candidates are identified and represented by a graph structure with weighted nodes and edges, respectively. Second, an optimization algorithm is utilized for generating an airway detection result by selecting a subset of airway branches and connections based on graph weights derived from image features. The performance of the algorithm with different feature categories and their combinations was assessed on a set of 50 lung CT scans from five different cohorts, including normal and diseased lungs. Results show trade-offs between feature categories/combinations in terms of correctly (true positive) and incorrectly (false positive) identified airways. Also, the performance of features in dependence of lung cohort was analyzed. Across all cohorts, a good trade-off with high true positive rate (TPR) and low false positive rate (FPR) was achieved by a combination of gray-value, local shape, and structural features. This combination enabled extracting 91.80% of reference airways (TPR) in combination with a low FPR of 1.00%. In addition, this variant was evaluated on the public EXACT'09 test set, and a comparison with other airway detection approaches is provided. One of the main advantages of the presented method is that it is robust against local disturbances/artifacts or other ambiguities that are frequently occurring in lung CT scans.

Subvoxel accurate graph search using non-Euclidean graph space

Graph search is attractive for the quantitative analysis of volumetric medical images, and especially for layered tissues, because it allows globally optimal solutions in low-order polynomial time. However, because nodes of graphs typically encode evenly distributed voxels of the volume with arcs connecting orthogonally sampled voxels in Euclidean space, segmentation cannot achieve greater precision than a single unit, i.e. the distance between two adjoining nodes, and partial volume effects are ignored. We generalize the graph to non-Euclidean space by allowing non-equidistant spacing between nodes, so that subvoxel accurate segmentation is achievable. Because the number of nodes and edges in the graph remains the same, running time and memory use are similar, while all the advantages of graph search, including global optimality and computational efficiency, are retained. A deformation field calculated from the volume data adaptively changes regional node density so that node density varies with the inverse of the expected cost. We validated our approach using optical coherence tomography (OCT) images of the retina and 3-D MR of the arterial wall, and achieved statistically significant increased accuracy. Our approach allows improved accuracy in volume data acquired with the same hardware, and also, preserved accuracy with lower resolution, more cost-effective, image acquisition equipment. The method is not limited to any specific imaging modality and readily extensible to higher dimensions.

LOGISMOS-B: layered optimal graph image segmentation of multiple objects and surfaces for the brain

Automated reconstruction of the cortical surface is one of the most challenging problems in the analysis of human brain magnetic resonance imaging (MRI). A desirable segmentation must be both spatially and topologically accurate, as well as robust and computationally efficient. We propose a novel algorithm, LOGISMOS-B, based on probabilistic tissue classification, generalized gradient vector flows and the LOGISMOS graph segmentation framework. Quantitative results on MRI datasets from both healthy subjects and multiple sclerosis patients using a total of 16,800 manually placed landmarks illustrate the excellent performance of our algorithm with respect to spatial accuracy. Remarkably, the average signed error was only 0.084 mm for the white matter and 0.008 mm for the gray matter, even in the presence of multiple sclerosis lesions. Statistical comparison shows that LOGISMOS-B produces a significantly more accurate cortical reconstruction than FreeSurfer, the current state-of-the-art approach (p << 0.001). Furthermore, LOGISMOS-B enjoys a run time that is less than a third of that of FreeSurfer, which is both substantial, considering the latter takes 10 h/subject on average, and a statistically significant speedup.

Graph-based IVUS segmentation with efficient computer-aided refinement

A new graph-based approach for segmentation of luminal and external elastic lamina (EEL) surface of coronary vessels in gated 20 MHz intravascular ultrasound (IVUS) image sequences (volumes) is presented. The approach consists of a fully automated segmentation stage ("new automated" or NA) and a user-guided computer-aided refinement ("new refinement" or NR) stage. Both approaches are based on the LOGISMOS approach for simultaneous dual-surface graph-based segmentation. This combination allows the user to efficiently combine general information about IVUS image appearance and case-specific IVUS morphology and therefore deal with frequently occurring issues like calcified plaque-causing signal shadowing-and imaging artifacts. The automated segmentation stage starts with pre-segmenting the lumen to automatically define the lumen centerline, which is used to transform the segmentation task into a LOGISMOS-family graph optimization problem. Following the automated segmentation, the user can inspect the result and correct local or regional segmentation inaccuracies by (iteratively) providing approximate clues regarding the location of the desired surface locations. This expert information is utilized to modify the previously calculated cost functions, locally re-optimizing the underlying modified graph without a need to start the new optimization from scratch. Validation of our method was performed on 41 gated 20 MHz IVUS data sets for which an expert-defined independent standard was available. Resulting from the automated stage of the approach (NA), the mean and standard deviation of the root mean square area errors for the luminal and external elastic lamina surfaces were 1.12 ±0.67 mm (2) and 2.35 ±1.61 mm (2) , respectively. Following the refinement stage (NR), the root mean square area errors significantly decreased to 0.82 ±0.44 mm (2) and 1.17 ±0.65 mm (2) for the same surfaces, respectively ( for both surfaces). The approach is delivering a previously unachievable speed of obtaining clinically relevant segmentations compared to the current approaches of automated segmentation followed by manual editing.