Lumbar Spine Degenerative detection using ResNet-50 & VGG16

Sasidharan Kandamchirayil, Lekshmi (2024) Lumbar Spine Degenerative detection using ResNet-50 & VGG16. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (809kB) | Preview
Preview	PDF (Configuration Manual) Download (2MB) | Preview

Abstract

Degenerative diseases of the spine, foraminal stenoses, subarticular stenoses, lumbar canal stenoses impair the quality of life. MRI screening for such conditions is important and their classification plays a key role in subsequent management. This research presents a deep learning framework using ResNet-50 to automatically classify the severity of these disorders into three levels: normal/mild, moderate and severe, allocated to spinal levels L1/2 to L5/S1. The system deals with sagittal and axial MRI images through data preprocessing, augmentation, and TensorFlow/Keras model training. Technical evaluation measures such as accuracy, precision, recall, and F1-score speculate about how the model works when solving classification issues and the model received an accuracy of 91.4%. The automated method provides better diagnostic assistance to the radiologists in practices for the degenerative spine conditions with acceptable rates of accuracy and speed. This work contributes new strategies for the universal problem of scalable and efficient analysis of medical images using deep learning.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Trinh, Anh Duong UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science R Medicine > Healthcare Industry H Social Sciences > HM Sociology > Information Science > Communication > Medical Informatics
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	04 Sep 2025 14:07
Last Modified:	04 Sep 2025 14:07
URI:	https://norma.ncirl.ie/id/eprint/8795

Actions (login required)

View Item