Annual Report 2020
Project Title: Analysis of Positron Emission Tomography data for tumour detection and delineation
Positron emission tomography (PET) is a type of nuclear medicine procedure that measures metabolic activity of the cells of body tissues. It has the potential to improve the outcome of cancer therapy because it allows the identification and characterization of tumours. PET imaging has been widely adopted as an important clinical modality in delineating the tumour target for radiation therapy, in quantitating tumour burden for therapy assessment and in detecting and quantitating recurrent or metastatic disease. In the past 20 years, a lot of PET auto-segmentation methods have been proposed on 2D slides. Some expand to 3D volumes which, however, are computationally expensive.
However, there is currently no consensus in the literature about methods which are optimal for clinical practice. In addition, algorithms for segmenting combinations of images from PET and other image modalities, such as CT, have appeared in the literature, but it is still challenging to use CT as a complementary tool for PET in automated image segmentations as the spatial resolution of the two modalities do not match.
To address those current concerns, in this project, we will first develop a statistical framework and algorithm for PET image analysis in 3D which will allow for automated or semi-automated tumour detection and delineation for a range of tumour types. On top of that, we will try to build a robust framework and algorithm that will allow us to seamlessly and objectively use complementary information from other modalities such as MRI, CT, DCECT for tumour detection monitoring and radiotherapy planning.
Awarded: Carnegie PhD Scholarship
Field: Mathematics & Statistics
University: University of Glasgow