Early Detection of Mesothelioma through Automatic Identification of Pleural Thickenings

Beschreibung

This dissertation describes an image analysis system which automatically detects pleural thickenings and measures their characteristic values from thoracic spiral CT images without any user interaction. The final end of the computer-aided diagnosis has been accomplished through the integration of the semi-automatic follow-up assessment into the system to observe a possible growth rate of a pleural thickening. The implementation of semi-automatic spatiotemporal matching enables the system for the first time to assign each thickening from a baseline data-set to its match from the follow-up scan. As a result, a list of all matched follow-up thickenings is provided to the physician, accompanied by its representative visualization. Moreover, the physician still has the possibility to interact manually giving a correct amendment. In comparison to the manual investigation, the new automatic pre-matching expedites the finding process and allows the physician to diagnose more data-sets on the same expense of time. Based on the volume calculations of the thickenings, differences in size of each thickening between the temporally consecutive data-sets are calculated and documented, providing the physician a quantitatively measured growth rate. While the research and development of the system algorithms have been scientifically posed to achieve the goal of this work, clinical partners were always involved by providing their indispensable suggestions from the clinical point of view as well as on-going validation in each step. The results of this work are promising and indicate that the developed system is able to assist physicians in the tedious task of finding and quantifying pleural thickenings in CT data.

Autor*in

Kraisorn Chaisaowong

Themen in »Early Detection of Mesothelioma through Automatic Identification of Pleural Thickenings«

Früherkennung Mustererkennung Pleuramesotheliom Asbest Automatisches Bildverarbeitungssystem Computergestützte Diagnostik Medizinische Bildverarbeitung Pleurale Verdickungen Spiral-CT-Aufnahmen Verlaufskontrolle

Stimmen zu »Early Detection of Mesothelioma through Automatic Identification of Pleural Thickenings«

Details

Link teilen