Deep learning‐based segmentation of knee MRI for fully automatic subregional morphological assessment of cartilage tissues: Data from the Osteoarthritis Initiative

Abstract

Morphological changes in knee cartilage subregions are valuable imaging‐based biomarkers for understanding progression of osteoarthritis, and they are typically detected from magnetic resonance imaging (MRI). So far, accurate segmentation of cartilage has been done manually. Deep learning approaches show high promise in automating the task; however, they lack clinically relevant evaluation. We introduce a fully automatic method for segmentation and subregional assessment of articular cartilage, and evaluate its predictive power in context of radiographic osteoarthritis progression. Two data sets of 3D double‐echo steady‐state (DESS) MRI derived from the Osteoarthritis Initiative were used: first, n = 88; second, n = 600, 0‐/12‐/24‐month visits. Our method performed deep learning‐based segmentation of knee cartilage tissues, their subregional division via multi‐atlas registration, and extraction of subregional volume and thickness. The segmentation model was developed and assessed on the first data set. Subsequently, on the second data set, the morphological measurements from our and the prior methods were analyzed in correlation and agreement, and, eventually, by their discriminative power of radiographic osteoarthritis progression over 12 and 24 months, retrospectively. The segmentation model showed very high correlation (r > 0.934) and agreement (mean difference < 116 mm3) in volumetric measurements with the reference segmentations. Comparison of our and manual segmentation methods yielded r = 0.845–0.973 and mean differences = 262–501 mm3 for weight‐bearing cartilage volume, and r = 0.770–0.962 and mean differences = 0.513–1.138 mm for subregional cartilage thickness. With regard to osteoarthritis progression, our method found most of the significant associations identified using the manual segmentation method, for both 12‐ and 24‐month subregional cartilage changes. The method may be effectively applied in osteoarthritis progression studies to extract cartilage‐related imaging biomarkers.

Predicting total knee arthroplasty from ultrasonography using machine learning

“…Ultrasound in OA assessment may have multiple clinical advantages as it can be done immediately and, in many occasions, lowering the need for referral to radiography [ 13 ] or more costly modalities, such as MRI [ 34 ]. In contrast to x-rays it can detect soft tissue changes and meniscal extrusion, which have been reported to be present in the earliest stages of OA and predict later structural changes [ 35 , 36 ].…”

Automated detection of knee cystic lesions on magnetic resonance imaging using deep learning

“…Deep learning offers excellent performance for segmenting multi-tissue knee joints and detecting ACL, cartilage, or meniscus injuries (15)(16)(17). However, few papers have addressed cysts and effusions of the knee joint, which are associated with high morbidity and could also serve as biomarkers for degenerative disorders or acute injuries, like knee osteoarthritis and meniscus injuries.…”

“…Artificial intelligence and deep learning are increasingly utilized in the medical field both in medical imaging and biomedical analysis ( 11 , 12 ). The role of AI in medical imaging of knee joints has been described in many primary publications ( 13 ), with an emphasis on OA-related research, such as auto-segmentation of knee joint tissue ( 14 , 15 ), and auto-detection of cartilage lesions, meniscus injuries, and anterior cruciate ligament tears ( 16 – 19 ). The deep learning models for such detection demonstrated relatively superb accuracy, ranging between 70 and 100% across various studies, suggesting that such methods exhibit the potential to rival human-level performance in decision-making tasks related to the MRI-based diagnosis of knee injuries.…”

Predicting osteoarthritis onset and progression with 3D texture analysis of cartilage MRI DESS: 6‐Year data from osteoarthritis initiative

“…MR images were acquired using 3T Siemens clinical MR systems (Siemens Healthcare) according to the OAI knee MRI protocol 41 . In total, 642 DESS images of right knees were segmented using the automatic deep learning‐based method that was previously trained and validated against manual segmentations 43 . As an output, the software produced separate segmentation masks for femoral, tibial, and patellar cartilage tissues, as well as menisci.…”