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

Abstract

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.

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

“…In cases with incomplete fissures, various methods have been developed that draw information from pulmonary anatomy and atlases. Lobe segmentation algorithms can be broadly categorised as either supervised [7]–[12] or unsupervised [13]–[16]. In our study, we extend the definition of supervised methods to encompass any algorithm that requires prior manual labelling to determine optimal fissure properties or the construction of anatomical atlases.…”

“…Segmentation algorithms can be further subdivided on the basis of the segmentation of auxiliary structures. Methods can be dependent [7], [10], [13]–[17] or independent [12], [18] of the information provided by the airway and vascular trees. Algorithms can also be classified based on their dependence on anatomical atlases [10]–[13], [18] or whether the method is uniquely performed in the patient-space [7], [14].…”

“…Such thresholds are often compromised when lung attenuation values themselves are variably influenced by a range of factors including the CT reconstruction algorithm, CT slice thickness and patient inspiratory effort. The issue of the removal of false-positive voxels and its dependence on prior knowledge is a further limitation of several algorithms [7], [12], [19], [20]. The likelihood function of Lassen et al [7] requires prior knowledge of fissure Hessian eigenvalues and may lead to an over-segmentation of fissure voxels that cannot be corrected through post-processing techniques.…”

Multiorgan structures detection using deep convolutional neural networks

“…We pose the problem in 2D instead of 3D since 2D detections have been proven sufficient in the search for image-based biomarkers in different clinical problems. 7,8 Besides, the inherent higher complexity of 3D architectures implies the need for more training data to achieve stable results.…”

Prevalence of incomplete interlobar fissures of the lung

“…Since the numbers of radiodiagnostic examinations are constantly growing and the search for incomplete fissures is time-consuming, automated computerized assessment [26][27][28] offers a solution. A publication comparing the results of radiologists and an automatic system found no statistically significant difference in assessment 19 ; in addition, the authors comparing CT and endobronchial measurement of collateral air flow [29][30][31] concluded that the results correlated.…”