A review on segmentation of knee articular cartilage: from conventional methods towards deep learning

Abstract

In this paper, we review the state-of-the-art approaches for knee articular cartilage segmentation from conventional techniques to deep learning (DL) based techniques. Knee articular cartilage segmentation on magnetic resonance (MR) images is of great importance in early diagnosis of osteoarthritis (OA). Besides, segmentation allows estimating the articular cartilage loss rate which is utilised in clinical practice for assessing the disease progression and morphological changes. It has been traditionally applied in quantifying longitudinal knee OA progression pattern to detect and assess the articular cartilage thickness and volume. Topics covered include various image processing algorithms and major features of different segmentation techniques, feature computations and the performance evaluation metrics. This paper is intended to provide researchers with a broad overview of the currently existing methods in the field, as well as to highlight the shortcomings and potential considerations in the application at clinical practice. The survey showed that state-of-the-art techniques based on DL outperform the other segmentation methods. The analysis of the existing methods reveals that integration of DL-based algorithms with other traditional model-based approaches has achieved the best results (mean Dice similarity coefficient (DSC) between 85.8% and 90%).

Interscan measurement error of knee cartilage thickness and projected cartilage area ratio at 9 regions and 45 subregions by fully automatic three-dimensional MRI analysis

“…We have addressed this problem by the recent development of software for automatic segmentation of cartilage using deep neural networks [2]. The accuracy of this software for knee MRI was comparable to that reported previously by Liu et al [3] and Norman et al [4] in Dice similarity coefficient (DSC) [5] between manual segmentation and automatic segmentation. These three software had a common feature in that they used convolutional neural networks, but ours was unique in that it used a 3D convolutional neural network [2].…”

Subchondral bone length in knee osteoarthritis: A deep learning derived imaging measure and its association with radiographic and clinical outcomes

“…SBL depicts both tibiofemoral articular cartilage morphology and bone shape, as it reflects the cartilage-covered articular area translated to length by excluding osteophytes and denuded areas. As accurate manual segmentation of MRI studies can be highly labor-intensive, there have been several efforts to develop computational methods for automating the segmentation and the measurement of knee structures with specific focus on the cartilage and the meniscus [24][25][26][27][28][29][30][31], and other structures of the knee joint [25,[32][33][34]. We performed additional analyses to show that SBL values at different locations of the knee vary in magnitude as a function of JSN grade in both the femur and the tibia.…”

Emergence of Deep Learning in Knee Osteoarthritis Diagnosis

“…Different architectures of deep learning have been applied in different types of medical images from imaging modalities such as radiography, ultrasound, computed tomography, and MRI to diagnose knee OA. Among all the deep learning architectures, CNN architecture has gained a large amount of research interest, particularly in knee OA segmentations and diagnosis [ 26 , 28 , 29 ]. One of the main advantages of CNN is that they are easier to train and have fewer parameters compared to other architectures [ 30 ].…”

“…CNN, basically the U-Net architecture, is popularly used in knee OA for automated segmentation of the cartilage, menisci, bone, or total knee joint anatomy [ 31 , 32 ]. Segmentation of the anatomical structures is important in the clinical practice to evaluate the disease progression and morphological changes where the recent breakthrough of this field is segmenting the cartilage from magnetic resonance (MR) images [ 28 , 33 ].…”

“…In the study of OA pathophysiology, there are a variety of imaging modalities available in the healthcare and research sector where the choice depends on the specific role of the modality [ 28 ]. Previously, plain radiographs were the “gold standard” used for initial radiographic evaluation to diagnose or assess the severity of knee OA.…”