Skip to main content

Eszter Kolozsi

Research student, School of Engineering

I am part of the  Medical Engineering Department, holding an MSc in Biomedical Engineering from the Technical University of Denmark (DTU) and BSc in Medical Physics from the West University of Timisoara. Currently working on a project entitled: Developing Radiomics as an Imaging Biomarker in High-Grade Glioma.

My areas of expertise are: Medical Physics; Radiation Physics; Medical Imaging;

My research interests include: Radiomics; MR imaging; X-ray imaging; CT imaging.

Research group: Sensors, Signals and Imaging

Research team: Cancer Imaging and Data Analytics (CIDA)


Developing Radiomics as an Imaging Biomarker in High-Grade Glioma.

The commonest primary brain tumour in adults is grade IV glioma, Glioblastoma Multiforme (GBM). Roughly 110 new cases are diagnosed each year through the South Wales Neuro Oncology Service. The current treatment options for these aggressive tumours are limited with poor overall prognosis. Diagnostic, prognostic, and predictive information from standard brain imaging remains restricted and a surgical biopsy is required to allow molecular genetic testing to provide these important biomarkers. Equally, biopsy or surgical resection are often unfeasible due to tumour location and the associated risk of morbidity from damage to surrounding normal brain tissue. Quantitative imaging is routinely used in the diagnostics and treatment response planning of HGG. However, the amount of quantication is limited and clinical decisions are still taken based on visual assessment. Nowadays, there is an increased interest in the combination of both quantication and visual assessment to provide a better understanding of the imaging data sets. In this way the description of characteristics, such as tumour shape and size or textural properties of the tumour will be quantified from the available images. This field of research is called Radiomics. 

In this study we will apply radiomics to MRI scans of patients with HGG using latest techniques and correlate imaging signatures with known genetic and pathological prognostic or predictive biomarkers to measure correlations. This will be achieved by establishing optimal MRI sequences for radiomic feature extraction; by establishing optimal radiomic features for extraction; and by correlating radiomic ndings with other clinical tumour features.

External profiles