Spectral texture segmentation of Magnetic Resonance Imaging (MRI) brain images for glioma brain tumour detection / Rosniza Roslan

In 2009, statistics showed that five percent of Malaysians has been diagnosed with brain tumours with glioma being the most common type. Radiologist commonly used MRI image sequences to detect glioma clinically by examining the abnormalities on T1-Weighted, T2-Weighted and Fluid Attenuated Inversion...

Full description

Saved in:
Bibliographic Details
Main Author: Roslan, Rosniza
Format: Thesis
Language:English
Published: 2013
Subjects:
Online Access:https://ir.uitm.edu.my/id/eprint/16072/1/TM_ROSNIZA%20ROSLAN%20CS%2013_5.PDF
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In 2009, statistics showed that five percent of Malaysians has been diagnosed with brain tumours with glioma being the most common type. Radiologist commonly used MRI image sequences to detect glioma clinically by examining the abnormalities on T1-Weighted, T2-Weighted and Fluid Attenuated Inversion Recovery (FLAIR) images. However, when the tumour cannot be detected visually, they will inject a contrast agent of gadolinium to enhance the image modality. However, this process delays acquisition of results at a higher cost and imposes side effects to the patients. Therefore, this thesis proposes utilizing spectral texture features of the MRI images in detecting the tumour in all three sequences of Tl, T2-Weighted and FLAIR images. There are four phases involved in this research which are data collection, preprocessing (i.e. skull stripping), processing (i.e. texture feature extraction and segmentation) and post-processing (i.e. test and evaluation). For data collection, a total of 126 MRI images of adults ranging from 18 to 60 years old are obtained from Hospital Sungai Buloh and Hospital Tengku Ampuan Rahimah Klang in Selangor. Ninety MRI image sequences of Tl-Weighted, T2-Weighted and FLAIR are used for skull-stripping experiments and results showed that mathematical morphology method outperformed region growing at an accuracy rate of 96%. A new double thresholding algorithm and a fully automated multiple seed points selection algorithm that works on all three MRI image sequences are also proposed. For texture feature extraction, we tested three features that are inverse Fast Fourier Transform (IFFT), texture energy and transformed IFFT. Experiments conducted on 64 MRI images, of all sequences showed that texture energy is the best texture feature to be used in glioma segmentation. Fuzzy C-Means clustering algorithm is then used to segment texture energy features from 126 MRI brain images of all sequences. Results are then qualitatively evaluated by an expert radiologist and it showed that the glioma brain tumour is been detected. Therefore, texture energy features and Fuzzy C-Means clustering method are identified as promising methods of glioma brain tumour detection at an accuracy rate of 76%, 86% and 79% is detected as an abnormal of Tl- Weighted, T2-Weighted and FLAIR images. The final chapter concludes with limitations and recommendations for further improvements of glioma tumour detection.