Texture analysis of multimodal magnetic resonance images in support of diagnostic classification of childhood brain tumours

Tantisatirapong, Suchada (2015). Texture analysis of multimodal magnetic resonance images in support of diagnostic classification of childhood brain tumours. University of Birmingham. Ph.D.

Preview

Tantisatirapong15PhD.pdf
PDF - Redacted Version
Download (7MB)

Abstract

Primary brain tumours are recognised as the most common form of solid tumours in children, with pilocytic astrocytoma, medulloblastoma and ependymoma being found most frequently. Despite their high mortality rate, early detection can be facilitated through the use of Magnetic Resonance Imaging (MRI), which is the preferred scanning technique for paediatric patients. MRI offers a variety of imaging sequences through structural and functional imaging, as well as providing complementary tissue information. However visual examination of MR images provides limited ability to characterise distinct histological types of brain tumours. In order to improve diagnostic classification, we explore the use of a computer-aided system based on texture analysis (TA) methods. TA has been applied on conventional MRI but has been less commonly studied on diffusion MRI of brain-related pathology. Furthermore, the combination of textural features derived from both imaging approaches has not yet been widely studied. In this thesis, the aim of the research is to investigate TA based on multi-centre multimodal MRI, in order to provide more comprehensive information and develop an automated processing framework for the classification of childhood brain tumours.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):