Detect and Analyze the Performance of Lumbar Spinal Stenosis Detection from MRI Images by Using Semantic Segmentation Technique

Abstract

Lower back pain is mainly caused by some complications present in the lumbar spine. In
general, human beings face a lot of problems with lower back pain and very few people figure out the exact
cause, and most of them unable to find the exact cause behind the pain. As we know that the diagnosis of a
medical record is very complex and plays a crucial role to the medical persons in order to treat the patients
who suffer with low back pain. In general, the medical practitioner tries to study the abnormality present in
the medical records like MRI or Scan images in a manual manner under direct eye contact, which is a very
complicated task to figure out the minute abnormalities which is present inside the report. This motivated
me to design this proposed application using machine learning (ML) models in the medical field for disease
prediction and to guide the medical experts about the patient’s current situation. In this present work, we try
to identify the most important physical parameters which are required to figure out the spinal abnormalities
which are collected physically from spine patients. Here we propose a novel method to predict and trace the
lumbar spinal stenosis through semantic segmentation and delineation of magnetic resonance imaging
(MRI) scans of the lumbar spine. By conducting various experiments on the spine dataset which contain
nearly 575 MRI studies of patients who are having symptomatic back pain. Our theoretical and
experimental results clearly state that proposed method produces a very good performance as compared
with primitive region-based metrics.