A Transfer Learning Approach for Lumbar Spine Disc State Classification

Ali Al-Kubaisi,Nasser N Khamiss

doi:10.3390/electronics11010085

Ali Al-Kubaisi, Nasser N Khamiss

Open Access

https://doi.org/10.3390/electronics11010085

Copy DOI

Abstract

Recently, deep learning algorithms have become one of the most popular methods and forms of algorithms used in the medical imaging analysis process. Deep learning tools provide accuracy and speed in the process of diagnosing and classifying lumbar spine problems. Disk herniation and spinal stenosis are two of the most common lower back diseases. The process of diagnosing pain in the lower back can be considered costly in terms of time and available expertise. In this paper, we used multiple approaches to overcome the problem of lack of training data in disc state classification and to enhance the performance of disc state classification tasks. To achieve this goal, transfer learning from different datasets and a proposed region of interest (ROI) technique were implemented. It has been demonstrated that using transfer learning from the same domain as the target dataset may increase performance dramatically. Applying the ROI method improved the disc state classification results in VGG19 2%, ResNet50 16%, MobileNetV2 5%, and VGG16 2%. The results improved VGG16 4% and in VGG19 6%, compared with the transfer from ImageNet. Moreover, it has been stated that the closer the data to be classified is to the data that the system trained on, the better the achieved results will be.

Highlights

Since the 1970s, researchers have built systems to analyze medical images and diagnose diseases based on images uploaded to computers
This paper aims to solve the problem of lack of training data in disc state classification of the lumbar spine, improve the performance of disc state classification tasks and to determine if the kind of images used for transfer learning has an impact on performance
It is known that Convolutional Neural Network (CNN) models require large amounts of data for training

Summary

Introduction

Since the 1970s, researchers have built systems to analyze medical images and diagnose diseases based on images uploaded to computers. The reason for the spread of medical images and the interest of researchers in this analysis is due to a large number of diseases and their spread in the world, especially lower back diseases. The causes of this pain can be due to spinal deformity, herniated disc, osteoporosis, or muscle strain as a result of modern lifestyle through office work. The intervertebral discs help stabilize the spine and act as shock absorbers, in addition to protecting the bones from friction and interference. These discs are filled with a gel-like fluid and, if they dry out, it is an indication of some problem [7]

Objectives

Methods

Results

Conclusion