Publication
Title
Tracheal stent prediction using statistical deformable models of tubular shapes
Author
Abstract
Tracheal stenosis is a narrowing of the trachea that impedes normal breathing. Tracheotomy is one solution, but subjects patients to intubation. An alternative technique employs tracheal stents, which are tubular structures that push the walls of the stenotic areas to their original location. They are implanted with endoscopes, therefore reducing the surgical risk to the patient. Stents can also be used in tracheal reconstruction to aid the recovery of reconstructed areas. Correct preoperative stent length and diameter specification is crucial to successful treatment, otherwise stents might not cover the stenotic area nor push the walls as required. The level of stenosis is usually measured from inside the trachea, either with endoscopes or with image processing techniques that, eg compute the distance from the centre line to the walls of the trachea. These methods are not suited for the prediction of stent sizes because they can not trivially estimate the healthy calibre of the trachea at the stenotic region. We propose an automatic method that enables the estimation of stent dimensions with statistical shape models of the trachea. An average trachea obtained from a training set of CT scans of healthy tracheas is placed in a CT image of a diseased person. The shape deforms according to the statistical model to match the walls of the trachea, except at stenotic areas. Since the deformed shape gives an estimation of the healthy trachea, it is possible to predict the size and diameter of the stent to be implanted in that specific subject.
Language
English
Source (journal)
Proceedings of the Society of Photo-optical Instrumentation Engineers / SPIE: International Society for Optical Engineering. - Bellingham, Wash.
Source (book)
Proceedings of SPIE Medical Imaging, San Diego, Calif., USA
Publication
San Diego, Calif. : SPIE , 2008
ISBN
978-0-8194-7098-0
DOI
10.1117/12.770237
Volume/pages
p. O9144,1-11
ISI
000256058600160
Full text (Publisher's DOI)
UAntwerpen
Faculty/Department
Research group
Publication type
Subject
Affiliation
Publications with a UAntwerp address
External links
Web of Science
Record
Identifier
Creation 08.10.2008
Last edited 23.08.2022
To cite this reference