Tissue Quantification in Veterinary Radiology: Region-of-Interest Methodology and Feature Variability

PhD defence by Panida Pongvittayanon

Assessment Committee

Associate Professor Caroline Marcussen, Department of Veterinary Clinical Sciences, University of Copenhagen (Chairperson)
Associate Professor Aage Kristian Olsen Alstrup, Aarhus University
Associate Professor Gabriel Manso-Díaz, Complutense University of Madrid (UCM)

Supervisors

Professor Fintan McEvoy
Associate Professor Anna Müller

Department

Department of Veterinary Clinical Sciences

Place

Digital defence, please follow the link: https://ucph-ku.zoom.us/j/6861352848?pwd=Frjp4D1RsGw0tXI6iVbabaxa54LHio.1&omn=64948280986
MeetingID, if relevant: 686 135 2848
Password, if relevant: 696514

As the defence will be held digitally, we kindly ask guests who do not have an active role in the defence to mute their microphones and keep their cameras turned off throughout the defence. Questions from the audience should be submitted via the chat function.

Email address to gain access to the thesis: dlf490@ku.dk.
You will either receive a copy of the thesis or be informed where you can read a physical copy.
Recipients of copies of the thesis are not allowed to share or distribute it due to copyright compliance.

Short description of the thesis

Perception is not reality. In medical imaging analysis, what we see, the area we select as a Region of Interest (ROI), and the values extracted from the ROI may be influenced by many sources of variation. This thesis provides insight into quantitative veterinary radiology, mainly focusing on ROI methodology and sources of variability in image features. Through three main projects, this thesis presents pixel-wise agreement and multivariate data analysis for ROI-methodology comparisons and evaluation of confounding factors on image features, while also presenting novel automatic segmentation approaches for predictive model improvement.