CHAPTER ONE

INTRODUCTION

1.1     Background of the Study

Osteoarthritis (OA) is the most common form of arthritis, diagnosed by clinical joint symptoms and radiographic findings. It is a major cause of physical disability in the elderly. In the United States, 14 million people aged 25 years and older have symptomatic knee OA and more than half diagnosed will undergo primary total knee replacement (TKR) before death, with over 600000 TKRs performed each year. Clinical OA symptoms include joint pain, stiffness, and decreased range of motion. Radiographic OA is diagnosed by using a grading system such as the Kellgren Lawrence (KL) grade or OA Research Society International (OARSI) atlas on the basis of the assessment of osteophytes and joint space narrowing. The presence of definite osteophytes with possible joint space narrowing (ie, KL grade 2) defines the radiographic knee OA in KL system. In the OARSI atlas, the radiographic knee OA is defined by any one of the following three separate criteria: joint space narrowing grade 2 or greater, sum of osteophyte grades 2 or greater, or joint space narrowing grade 1 and osteophyte grade 1. However, radiographic knee OA grading systems have multiple versions with no uniform agreement (Kevin et al., 2020).

People who have knee osteoarthritis (OA) commonly report pain and physical limitation performing functional activities such as walking, transitioning from a chair and negotiating stairs (Fukutani et al., 2016). During these activities they also use less sagittal plane range of movement (knee flexion) during particular phases of activities (e.g., stance phase of walking) compared to people who do not have osteoarthritis (McCarthy et al., 2013).

Clinicians are interested in the relationship between specific kinematic measures and clinical outcomes in people with knee osteoarthritis. For example, a person may have difficulty descending stairs because they do not use available knee flexion movement during the stance phase (Tan et al., 2016).

Swelling, joint pain, and stiffness are the prominent symptoms among others, such as restrictions in movement including walking, stair climbing, and bending (Heidari, 2011). The symptoms worsen over time and elderly patients are affected more frequently than patients in other age groups. The presence of OA in the knee reduces activity in daily life and eventually leads to disability, which can incur high costs related to loss in productivity (Altman et al., 2011). It is estimated that functional impairment of the knee and the hip are the eleventh highest disability factors (Cross, 2014) contributing to considerable socio-economic burden with an estimated cost per patient per year of approximately 19,000 Euro (Puig-Junoy et al., 2015).

The estimated prevalence of disability due to arthritis is expected to reach 11.6 million individuals by the year 2020, which is greater than the estimated risk of disability attributable to cardiovascular diseases or any other medical condition (Jaynal et al., 2019).

Interventions such as exercise (Davis et al., 2018) and total knee replacement (Wang et al., 2019) have demonstrated the ability to improve knee flexion angle during walking in people who have knee osteoarthritis. Clinical guidelines recommend that the performance of painful and limited activities is monitored over the course of treatment (Dobson et al., 2019).

To reduce the influence of subjectivity in quantifying KOA severity from X-ray images, computer-aided diagnosis has been very helpful. To date the sample size of available images has been the main limiting factor to train an efficient model. The Osteoarthritis Initiative (OAI) and the Multi-center Osteoarthritis Study (MOST) mitigated this small sample size limitation by making thousands of patient’s data and X-ray images available. Recently, several researchers have used these resources to develop an automatic approach for quantifying KOA severity by analyzing X-ray images (Jaynal, 2019).

However, there are currently several limitations to clinicians being able to accurately quantify sagittal plane knee range of movement during functional activities in both clinical and free-living environments (e.g., patient’s home or work, or during recreation). Clinicians are unable to routinely access gold standard optoelectronic motion analysis systems (e.g., Vicon) due to cost and space requirements. Smartphone camera-based technology is more accessible to clinicians and has demonstrated validity and reliability for measuring sagittal plane knee angles (Milanese et al., 2014). Optoelectronic systems require the patient to be observed within a fixed volume to record useful clinical information, precluding their use in a free-living environment. A deep learning-based osteoarthritis predication is been proposed to aid it early diagnosis.

1.2     Statement of the Problem

Knee osteoarthritis (OA) has posed a lot of challenges in patients, due to late diagnosis. Clinicians are unable to routinely access gold standard optoelectronic motion analysis systems due to it cost and large space requirements and also the diagnosis process have been found to be time consuming and expensive also. Smartphone camera-based technology is more accessible to clinicians and has demonstrated validity and reliability for performing several medical imaging tasks, thus, a smart phone based deep learning approach to the prediction of knee osteoarthritis is been proposed to curtail the expensive and time-consuming process of optoelectronic motion analysis systems by training and deploying a model on a smart phone to automate the diagnosis process on mobile devices.

1.3     Aim and Objectives of the Study

The aim of this research is to develop a knee osteoarthritis prediction system, that will use a deep learning (DL) algorithm known as Convolutional Neural Network (CNN) to predict knee osteoarthritis through x-ray images of previous diagnostic conditions.

The objectives of the study are as follows to:

1. Development of a deep learning model for knee osteoarthritis prediction.
2. Reduce the time frame it takes medical professionals in the detection of knee osteoarthritis, by the deployment of the trained model on a mobile application. The camera of the mobile device will now be used to capture input images of knee osteoarthritis scans, while the trained model makes the prediction, which is faster than human doctors’ examination.
3. Evaluate the developed system for performance using the model training performance metrics such as the confusion Metrics and f1 scores.

1.4     Significance of the Study

The research and software will be of significant in the following ways;

1. Beneficial to medical imaging diagnosis
2. To knee osteoarthritis patients in terms of the cost of carrying out such diagnosis.
3. Contribute to the existing body of knowledge
4. Aid in the quest for further research in the area of deep learning and computer vision in the medical sector.

1.5     Scope of the Study

The system proposed in this research will only cover the prediction of knee osteoarthritis using radiographic X-ray images of past knee osteoarthritis cases. The system would take in input radiographic image scan and make prediction based on the model being trained and deployed using convolutional neural (CNN) algorithm.

1.6     Limitation of the Study

The researcher faced some limitations during the course of this research. Time factor was the major factor as the timeframe to submit the project was too short for extensive research. Lack of adequate finance to fund the researcher is also a constraint, as to outsources many data.

1.7     Definition of Terms

1. Artificial Intelligence (AI): This is intelligence demonstrated by machines, unlike the natural intelligence displayed by humans and animals, which involves consciousness and emotionality.
2. Convolutional Neural Network (CNN): This is a Deep Learning algorithm which can take in an input image, assign importance (learnable weights and biases) to various aspects/objects in the image and be able to differentiate one from the other.
3. Data Argumentation:   Is use in deep learning to regularize, normalize the information on data to be trained.
4. Data Notation:    This is the process of labeling data to make it usable for machine learning.
5. Dataset: A collection of related sets of information that is composed of separate elements but can be manipulated as a unit by a computer.
6. Deep Learning (DL): This is a subfield of machine learning concerned with algorithms inspired by the structure and function of the brain called artificial neural networks.
7. Machine Learning (ML): This is the study of computer algorithms that improve automatically through experience and by the use of data.
8. Mobile App:  A mobile app or mobile application is a computer program or software application designed to run on a mobile device such as a phone-tablet or watch.
9. Model: Is a description or analogy used to help visualize something that cannot be directly observed.
10. Multilayer Perceptron: This refers to fully connected networks, that is, each neuron in one layer is connected to all neurons in the next layer.
11. Neural Networks (NN): This is a series of algorithms that endeavors to recognize underlying relationships in a set of data through a process that mimics the way the human brain operates.
12. Perceptron: This refers to a linear classifier (binary) use in supervised learning. It helps to classify the given input data.
13. Knee Osteoarthritis (KOA): is the most common form of arthritis in the knee. It is a degenerative, “wear-and-tear" type of arthritis that occurs most often in people 50 years of age and older, although it may occur in younger people, too. In osteoarthritis, the cartilage in the knee joint gradually wears away.
14. Software:   It is a program and other operating information use by a computer system.
15. Tensor-Flow: is a free and open-source software library for dataflow and differentiable programming across a range of tasks. It is a symbolic math library, and is also used for machine learning applications such as neural networks.