GPU-Based Airway Tree Segmentation and Centerline Extraction
Lung cancer is one of the deadliest and most common types of cancer in
Norway. Early and precise diagnosis is crucial for improving the survival
rate. Diagnosis is often done by extracting a tissue sample in the lung through
the mouth and throat. It is difficult to navigate to the tissue because of the
complexity of the airways inside the lung and the reduced visibility. Our goal
is to make a program that can automatically extract a map of the Airways
directly from X-ray Computer Tomography(CT) images of the patient. This
is a complex task and requires time consuming processing.
In this thesis we explore different methods for extracting the Airways from
CT images. We also investigate parallel processing and the usage of modern
graphic processing units for speeding up the computations. We rate several
methods in terms of reported performance and the possibility of parallel
processing. The best rated method is implemented in a parallel framework
called Open Computing Language.
The results shows that our implementation is able to extract large parts of
the Airway Tree, but struggles with the smaller airways and airways that
deviate from a perfect circular cross-section. Our implementation is able
to process a full CT scan using less than a minute with a modern graphic
processing units. The implementation is very general and is able to extract
other tubular structures as well. To show this we also run our implementation
on a Magnetic Resonance Angio dataset for finding blood vessels in the brain
and achieve good results.
We see a lot of potential in this method for extracting tubular structures. The
method struggles the most with noise handling and tubes that deviate from
a circular cross-sectional shape. We believe that this can be improved by
using another method than ridge traversal for the centerline extraction step.
Because this is a local greedy algorithm, it often terminates prematurely due
to noise and other image artifacts.
Place, publisher, year, edition, pages
Institutt for datateknikk og informasjonsvitenskap , 2012. , 101 p.
ntnudaim:6751, MTDT datateknikk, Intelligente systemer
IdentifiersURN: urn:nbn:no:ntnu:diva-18353Local ID: ntnudaim:6751OAI: oai:DiVA.org:ntnu-18353DiVA: diva2:565858
Lindseth, Frank, Førsteamanuensis IIElster, Anne C.