Three-Dimensional Human Airway Segmentation Methods for Clinical Virtual Bronchoscopy

Abstract

Rationale and Objectives. The segmentation of airways from CT images is a critical first step for numerous virtual bronchoscopic (VB) applications. Automatic or semiautomatic methods are necessary, since manual segmentation is prohibitively time consuming. The methods must be robust and operate within a reasonable time frame to be useful for clinical VB use. The authors developed an integrated airway segmentation system and demonstrated its effectiveness on a series of human images. Materials and Methods.The authors' airway segmentation system draws on two segmentation algorithms: (a) an adaptive region-growing algorithm and (b) a new hybrid algorithm that uses both region growing and mathematical morphology. Images from an ongoing VB study were segmented by means of both the adaptive region-growing and the new hybrid methods. The segmentation volume, branch number estimate, and segmentation quality were determined for each case.Results. The results demonstrate the need for an integrated segmentation system, since no single method is superior for all clinically relevant cases. The region-growing algorithm is the fastest and provides acceptable segmentations for most VB applications, but the hybrid method provides superior airway edge localization, making it better suited for quantitative applications. In addition, the authors show that prefiltering the image data before airway segmentation increases the robustness of both regiongrowing and hybrid methods. Conclusion.The combination of these two algorithms with the prefiltering options allowed the successful segmentation of all test images. The times required for all segmentations were acceptable, and the results were suitable for the authors' VB application needs.Key Words. Bronchi, CT; bronchoscopy; computed tomography (CT), image processing; computed tomography (CT), threedimensional; trachea, CT. © AUR, 2002New multidetector spiral CT scanners can produce threedimensional (3D) volumetric images of the human airway tree consisting of hundreds of two-dimensional (2D) sections (1,2). A typical 3D image can include 400 or more 512 ϫ 512 0.6-mm sections. Such images provide an excellent basis for virtual bronchoscopy (VB) applications (3-15) and quantitative airway analysis (8,16 -20). New VB methods also allow for live guided nodule and lymph node biopsies (13,15). A critical first step in these VB applications is the segmentation of the airway tree. Manual interactive segmentation has been applied in some cases, but routine manual analysis is impractical for the large 3D images arising from the new scanners (21,22). A variety of semiautomatic airway segmentation techniques have been proposed, but none have been conclusively proved adequate for very large, high-resolution, 3D CT chest images (4,20 -30).

Heterogeneity of Specific Gas Volume Changes

“…5 Region of Interest Selection: A three-dimensional airway reconstruction was used to choose a region of interest (ROI) in each lung segment (see later) at the two different lung volumes ( Fig 1 ). An automatic algorithm using a region-growing approach, based on the iterative algorithm proposed by Kiraly et al, 16 was developed to segment airway trees. The airways segmentation began by selecting a seed point in the trachea.…”

From CT Scans to CFD Modelling – Fluid and Heat Transfer in a Realistic Human Nasal Cavity

“…In this case a threshold limit is applied through the scans, where the segmenting region defined by image pixels has a value higher than or equal to a defined threshold value. Many segmentation algorithms exist (Kiraly et al, 2002;Pal and Pal, 1993) which have been adopted by commercial software such as Mimics (Materialise, 2008) and Amira (VisageImaging, 2008). An alternative is Matlab's Image Processing module (MathWorks, 2007) which provides useful functions to develop algorithms found in the literature (Beare, 2006;Kundu, 1990) or to develop custom algortihms.…”

Intravascular virtual MR endoscopy evaluation of cerebral aneurysms

“…Studies have described clinical applications of virtual endoscopy to visualize the inside of the bronchi, paranasal sinus, stomach, bladder, small intestine, pancreatic duct, bile duct, and spine (2,6,8,10,11,16,19,21,23,24). Shigematsu et al (20) reported the use of virtual intravascular MR endoscopy for aneurysms in three cases.…”

“…Studies have described clinical applications of virtual endoscopy to visualize the inside of the tracheobronchial tree, paranasal sinus, stomach, bladder, small intestine, pancreatic duct, bile duct, and spine (2,(6)(7)(8)10,11,16,19,21,23,24). Virtual endoscopy used for preoperative planning in neurosurgery for endoscopic third ventriculostomy, nasal cavity and paranasal sinus surgery, and virtual endoscopy visualization of the sphenoid sinus and important adjacent structures is valuable for training and the preoperative planning of endonasal transsphenoidal surgery (15).…”