Advancing Brain Tumor Detection: Hybrid Layered Model for Enhanced MRI Imaging Analysis

Marimuthu, Arun Murugan (2024) Advancing Brain Tumor Detection: Hybrid Layered Model for Enhanced MRI Imaging Analysis. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (814kB) | Preview
Preview	PDF (Configuration Manual) Download (605kB) | Preview

Abstract

Brain tumor pose a significant medical challenge, often resulting in organ dysfunction and potentially fatal outcomes. Magnetic Resonance Imaging (MRI) offers precise brain images with great resolution, making it an excellent tool for detecting tumors. Nevertheless, the process of manually identifying tumor-bearing regions by radiologists is prone to errors, especially when faced with the difficulties presented by CSF fluid and white matter. Deep learning models provide effective data segmentation and classification, which are crucial for planning treatment regimens and diagnosing cancer effectively. The CNN component efficiently extracts spatial features from MRI images, while the BiLSTM processes these features to recognize temporal relationships, which are crucial for distinguishing subtle differences in tumor characteristics. By integrating these two deep learning techniques, the model improves the accuracy and reliability of tumor classification. Extensive testing on a standard MRI dataset revealed that this approach achieved an impressive accuracy rate, underscoring the effectiveness of this hybrid model in overcoming the limitations of traditional CNN-based methods. This work suggests that the fusion of CNN and BiLSTM could lead to significant advancements in the field of brain tumor detection, offering a valuable tool for more accurate and reliable diagnostics in clinical practice. The integration of these models into clinical workflows holds the potential for groundbreaking advancements in brain tumor identification. This integration will enable quicker and more precise detection and classification, ultimately optimising treatment methods.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Raza Abidi, SM UNSPECIFIED
Uncontrolled Keywords:	MRI; Brain Tumor; Transfer learning; Image Pre-processing; ResNet V2
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 > Life sciences > Medical sciences 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 10:22
Last Modified:	20 Aug 2025 10:22
URI:	https://norma.ncirl.ie/id/eprint/8586

Actions (login required)

View Item