Lightweight Deep Learning Framework for Brain Tumour Classification

Salvi, Rohit Narayan (2023) Lightweight Deep Learning Framework for Brain Tumour Classification. Masters thesis, Dublin, National College of Ireland.

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

Abstract

The National Brain Tumour Society states that there are over 100 different forms of primary brain tumours, such as gliomas, meningiomas, pituitary and so on. Brain tumour diagnosis involves detecting the type of brain tumour and its severity. The challenge is to accurately identify and classify a brain tumour with limited computation. This research proposes a lightweight deep learning framework for brain tumour classification. The proposed framework combines a machine learning classification model and weight pruning optimization technique to detect brain tumours with limited computation. The classification model is implemented using Convolutional Neural Networks(CNN) and the magnitude-based weight pruning technique is used to optimise the classification model. A Brain Tumour Magnetic Resonance Imaging(BTMRI) dataset of 7023 MRIs representing 4 distinct classes of brain tumours namely– glioma, meningioma, pituitary and no tumour is utilised to analyse and evaluate a proposed framework. The results of the proposed framework are presented in this paper based on accuracy, sensitivity, specificity and loss function. Results of the proposed framework show an accuracy of 87.26% and loss 0f 0.39 after 25 epochs. The proposed framework is 4.65% more accurate and has 15% lower loss than the state-of-the-art CNN for multiclass brain tumour classification. This research shows promise for aiding patients in getting an early view of their tumour type.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Stynes, Paul UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science R Medicine > RC Internal medicine > RC0254 Neoplasms. Tumors. Oncology (including Cancer) Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Tamara Malone
Date Deposited:	02 Jan 2025 13:30
Last Modified:	02 Jan 2025 13:30
URI:	https://norma.ncirl.ie/id/eprint/7262

Actions (login required)

View Item