Militzer A (2014)
Publication Language: English
Publication Type: Thesis
Publication year: 2014
Publisher: Friedrich-Alexander-Universität Erlangen-Nürnberg
City/Town: Erlangen
Pages Range: 119
URI: https://opus4.kobv.de/opus4-fau/frontdoor/index/index/docId/6044
Over the past decades, huge progress has been made in treatment of cancer, decreasing fatality rates despite a growing number of cases. Technical achievements had a big share in this development.
With modern image acquisition techniques, most types of tumors can be made visible.
Automatic processing of these images to support diagnosis and therapy, on the other hand, is still very basic. Marking lesions for volume measurements, intervention planning or tracking over time requires a lot of manual interaction, which is both tedious and error prone.
The work at hand therefore aims at providing tools for the automatic segmentation of
liver lesions. A system is presented that receives a contrast enhanced CT image of the liver as input and, after several preprocessing steps, decides for each image voxel inside the liver whether it belongs to a tumor or not. That way, tumors are not only detected in the image but also precisely delineated in three dimensions. For the decision step, which is the main target of this thesis, we adopted the recently proposed Probabilistic Boosting Tree. In an offline learning phase, this classifier is trained using a number of example images. After training, it can process new and previously unseen images.
Such automatic segmentation systems are particularly valuable when it comes to monitoring tumors of a patient over a longer period of time. Therefore, we propose a method for
learning a prior model to improve segmentation accuracy for such follow-up examinations.
It is learned from a number of series of CT images, where each series contains images of
one patient. Two different ways of incorporating the model into the segmentation system are investigated. When acquiring an image of a patient, the system can use the model to calculate a patient specific lesion prior from images of the same patient acquired earlier
and thus guide the segmentation in the current image.
The validity of this approach is shown in a set of experiments on clinical images. When comparing the points of 90% sensitivity in these experiments, incorporating the prior improved the precision of the segmentation from 82.7% to 91.9%. This corresponds to a
reduction of the number of false positive voxels per true positive voxel by 57.8%.
Finally, we address the issue of long processing times of classification based segmentation systems. During training, the Probabilistic Boosting Tree builds up a hierarchy of AdaBoost classifiers. In order to speed up classification during application phase, we modify this hierarchy so that simpler and thus faster AdaBoost classifiers are used in higher levels. To this end, we introduce a cost term into AdaBoost training that trades off discriminative
power and computational complexity during feature selection. That way the
optimization process can be guided to build less complex classifiers for higher levels of the tree and more complex and thus stronger ones for deeper levels. Results of an experimental evaluation on clinical images are presented, which show that this mechanism can reduce
the overall cost during application phase by up to 76% without degrading classification accuracy.
It is also shown that this mechanism could be used to optimize arbitrary secondary
conditions during AdaBoost training.
APA:
Militzer, A. (2014). Boosting Methods for Automatic Segmentation of Focal Liver Lesions (Dissertation).
MLA:
Militzer, Arne. Boosting Methods for Automatic Segmentation of Focal Liver Lesions. Dissertation, Erlangen: Friedrich-Alexander-Universität Erlangen-Nürnberg, 2014.
BibTeX: Download