Multi-Modal Brain Tumor Segmentation with Attention Mechanisms

Mulani, Bilal Mustaq (2024) Multi-Modal Brain Tumor Segmentation with Attention Mechanisms. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (2MB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

The study of this research is basically aimed to improve and enhance the brain tumor segmentation in the scope MRI Scans which are integrated with advanced deep learning architectures which include Vanilla U-shaped Network, Residual U-shaped Network and Attention U-shaped Network. The research is specifically focusing on enhancing the segmentation accuracy while keeping and maintaining the computational efficiency, a critical requirement for clinical based applications. The segmentation of LGG dataset was used in the evaluation models where Dice Coefficient and Intersection over Union(IoU) along with precision and recall performance metrics which were employed in assessment of the performance. The output results show that Attention UNet model has outperformed well in comparison with other models very significantly, with the dice Coefficient value of 0.9123 and IoU value of 0.8612. The Attention U Network also showed highest precision at 0.9234 along with recall of 0.8912 by implying that it has the ability to precisely segment the complicated tumor regions. Observation of these performance metrics show and reveal that with integration of attention mechanism in the architecture and focusing on pertinent image areas is the intensified lead towards improvement of segmentation’s outcomes. The research shares and offers and optimized model for the application in the clinical cases and develops the program models which have been made openly available for studies. However it shreds and shows hints on further optimization needs when it comes to deal with complexity especially in low settings of computational resource.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Makki, Ahmed UNSPECIFIED
Uncontrolled Keywords:	Brain Tumor Segmentation; Magnetic Resonance Imaging (MRI); Deep Learning; UNet (U-Shaped Network); ResUNet (Residual UNet); Attention Universal Network (Attention UNet); Automated Segmentation; Machine Learning in Healthcare
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > Life sciences > Medical sciences > Pathology > Tumors > Cancer Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning Q Science > Life sciences > Medical sciences > Pathology > Tumors
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	20 Aug 2025 11:33
Last Modified:	20 Aug 2025 11:33
URI:	https://norma.ncirl.ie/id/eprint/8593

Actions (login required)

View Item