Emergence of Deep Learning in Knee Osteoarthritis Diagnosis.

Pauline Shan Qing Yeoh,Khairunnisa Hasikin,Yee Kai Tee,Khin Wee Lai,Siew Li Goh,Samiappan Dhanalakshmi,Yan Chai Hum,Bai Yuan Ding

doi:10.1155/2021/4931437

Abstract

Osteoarthritis (OA), especially knee OA, is the most common form of arthritis, causing significant disability in patients worldwide. Manual diagnosis, segmentation, and annotations of knee joints remain as the popular method to diagnose OA in clinical practices, although they are tedious and greatly subject to user variation. Therefore, to overcome the limitations of the commonly used method as above, numerous deep learning approaches, especially the convolutional neural network (CNN), have been developed to improve the clinical workflow efficiency. Medical imaging processes, especially those that produce 3-dimensional (3D) images such as MRI, possess ability to reveal hidden structures in a volumetric view. Acknowledging that changes in a knee joint is a 3D complexity, 3D CNN has been employed to analyse the joint problem for a more accurate diagnosis in the recent years. In this review, we provide a broad overview on the current 2D and 3D CNN approaches in the OA research field. We reviewed 74 studies related to classification and segmentation of knee osteoarthritis from the Web of Science database and discussed the various state-of-the-art deep learning approaches proposed. We highlighted the potential and possibility of 3D CNN in the knee osteoarthritis field. We concluded by discussing the possible challenges faced as well as the potential advancements in adopting 3D CNNs in this field.

Highlights

Osteoarthritis (OA) is one of the most prevalent degenerative musculoskeletal diseases. is disease is affecting almost 5% of the global population [1]. e knee is the most common joint affected by OA, which is characterized by irreversible degeneration of the articular cartilage at the ends of the bones such as femoral, tibial, and patella cartilages [2]
We provided updates on the application of various convolutional neural network (CNN) approaches in segmentation and classification models
It is obvious that most studies utilize plain radiography as their architecture input for classifying OA

Summary

Introduction

Osteoarthritis (OA) is one of the most prevalent degenerative musculoskeletal diseases. is disease is affecting almost 5% of the global population [1]. e knee is the most common joint affected by OA, which is characterized by irreversible degeneration of the articular cartilage at the ends of the bones such as femoral, tibial, and patella cartilages [2]. Current radiographic grading scales for OA rely primarily on Kellgren–Lawrence grading which examines the changes shown on X-ray plain radiography images This approach causes delay in OA diagnosis because the bony changes only appear in advanced conditions. E use of deep learning, especially with convolutional neural networks, is prevalent as it has shown validated results as compared to human practitioners’ manual methods or classical methods [8, 12]. Deep learning methods such as CNN learn complex features by extracting visual features automatically using combinations of series of transformations in the model architecture [11, 14]. To ensure generalization of the model, the data are typically categorized into three sets: training set for hyperparameter optimization, validation set for overfitting control, and test set [16]

Nonimaging-Based Deep Learning

Application of 2D Deep Learning in Knee Osteoarthritis Assessment

Application of 3D Deep Learning in Knee Osteoarthritis Assessment

Discussion

Findings