Logo des Repositoriums
Zur Startseite
  • English
  • Deutsch
Anmelden
  1. Startseite
  2. SuUB
  3. Dissertationen
  4. Morphological Volumetry : Theory, Concepts, and Application to Quantitative Medical Imaging
 
Zitierlink URN
https://nbn-resolving.de/urn:nbn:de:gbv:46-diss000012525

Morphological Volumetry : Theory, Concepts, and Application to Quantitative Medical Imaging

Veröffentlichungsdatum
2005-03-11
Autoren
Hahn, Horst Karl  
Betreuer
Peitgen, Heinz-Otto  
Gutachter
Reiber, Johan H. C.  
Zusammenfassung
This thesis brings quantitative methods on complex medical image data closer to clinical routine, and bridges the gap between manual and automated processing. Efficient and effective image segmentation methods are described, and applied to currently insufficiently solved problems that are relevant for medical diagnosis and therapy monitoring. The methods are designed to be suitable for software assistants that meet the requirements imposed by a quantitative analysis in clinical imaging: Applicability regarding the required hardware and acquisition protocols, as well as an effective and intuitive user control, robustness to variations of image quality and also of anatomy and pathology, precision of all derived quantitative measures, sensitivity with respect to relevant changes of the examined structures, and efficiency in order to be applicable to very large data sets on standard hardware. The first part of this thesis describes novel methods that build upon and extend the morphological watershed transform that is known as the key concept for image segmentation in Mathematical Morphology, with applications to brain segmentation in magnetic resonance imaging as well as bone removal in computed tomography and CT angiography. The second part focuses on quantitative image analysis in general, as well as the volumetry of specific neuroanatomical structures in particular. Morphological Volumetry is proposed as a framework for the reliable volumetric quantification of complex three-dimensional structures found in tomographic medical images, combining the extended watershed transform and automated histogram analysis. In cooperation with clinical and industrial partners, the proposed methods were successfully evaluated on a great variety of patient and volunteer images, as well as on software phantoms. The methods have shown potential to improve medical diagnosis and quantitative therapy monitoring and to enhance objectivity in medical image analysis in various instances.
Schlagwörter
Quantitative Medical Imaging

; 

Neuroradiology

; 

Magnetic Resonance Imaging

; 

Volumetry

; 

Image Analysis

; 

Mathematical Morphology

; 

Interactive Systems
Institution
Universität Bremen  
Fachbereich
Fachbereich 03: Mathematik/Informatik (FB 03)  
Dokumenttyp
Dissertation
Zweitveröffentlichung
Nein
Sprache
Englisch
Dateien
Lade...
Vorschaubild
Name

E-Diss1252_Hahn.pdf

Size

8.35 MB

Format

Adobe PDF

Checksum

(MD5):ee9592cdd5bc05def43fbea2d003e631

Built with DSpace-CRIS software - Extension maintained and optimized by 4Science

  • Datenschutzbestimmungen
  • Endnutzervereinbarung
  • Feedback schicken