Cartilage Segmentation Research Articles

A microCT imaging protocol for reproducible and efficient quantitative morphometric analysis (QMA) of joint structures of the in situ mouse tibio-femoral joint.

Micro-computed tomography (microCT) offers a three-dimensional (3D), high-resolution technique for the visualisation and analysis of bone microstructure. Using contrast-enhanced microCT, this capability has been expanded in recent studies to include cartilage morphometry and whole joint measures, known together as quantitative morphometric analysis (QMA). However, one of the main challenges in quantitative analysis of joint images is sensitivity to joint pose and alignment, which may influence measures related to both joint space and joint biomechanics. Thus, this study proposes a novel microCT imaging protocol for reproducible and efficient QMA of in situ mouse tibio-femoral joint. This work consists of two parts: an in situ diffusion kinetics study for a known cationic iodinated contrast agent (CA4+) for QMA of the cartilage, and a joint positioning and image processing workflow for whole joint QMA. In the diffusion kinetics study, 8 mice were injected at both of their tibio-femoral joints with distinct CA4+ concentrations and diffusion times. The mice were scanned at different time points after injection, and evaluated using attenuation and cartilage QMA measures. Results show that cartilage segmentation and QMA could be performed for CA4+ solution at a concentration of 48mg/ml, and that reliable measurement and quantification of cartilage were achieved after 5min of diffusion following contrast agent injection. We established the joint positioning and image processing workflow by developing a novel positioning device to control joint pose during scanning, and a spherical harmonics-based image processing workflow to ensure consistent alignment during image processing. Both legs of seven mice were scanned 10 times, 5 prior to receiving CA4+ and 5 after, and evaluated using whole joint QMA parameters. Joint QMA evaluation of the workflow showed excellent reproducibility; intraclass correlation coefficients ranged from 0.794 to 0.930, confirming that the imaging protocol enables reproducible and efficient QMA of joint structures in preclinical models, and that contrast agent injection did not cause significant alteration to the measured parameters.

基于深度学习的膝关节软骨MR图像分割方法研究

基于深度学习的膝关节软骨MR图像分割方法研究

Individual Influence of Trochlear Dysplasia on Patellofemoral Kinematics after Isolated MPFL Reconstruction.

The influence of the MPFL graft in cases of patella instability with dysplastic trochlea is a controversial topic. The effect of the MPFL reconstruction as single therapy is under investigation, especially with severely dysplastic trochlea (Dejour types C and D). The purpose of this study was to evaluate the impact of trochlear dysplasia on patellar kinematics in patients suffering from low flexion patellar instability under weight-bearing conditions after isolated MPFL reconstruction. Thirteen patients were included in this study, among them were eight patients with mild dysplasia (Dejour type A and B) and five patients with severe dysplasia (Dejour type C and D). By performing a knee MRI with in situ loading, patella kinematics and the patellofemoral cartilage contact area could be measured under the activation of the quadriceps musculature in knee flexion angles of 0°, 15° and 30°. To mitigate MRI motion artefacts, prospective motion correction based on optical tracking was applied. Bone and cartilage segmentation were performed semi-automatically for further data analysis. Cartilage contact area (CCA) and patella tilt were the main outcome measures for this study. Pre- and post-surgery measures were compared for each group. Data visualized a trending lower patella tilt after MPFL graft installation in both groups and flexion angles of the knee. There were no significant changes in patella tilt at 0° (unloaded pre-surgery: 22.6 ± 15.2; post-surgery: 17.7 ± 14.3; p = 0.110) and unloaded 15° flexion (pre-surgery: 18.9 ± 12.7; post-surgery: 12.2 ± 13.0; p = 0.052) of the knee in patients with mild dysplasia, whereas in patients with severe dysplasia of the trochlea the results happened not to be significant in the same angles with loading of 5 kg (0° flexion pre-surgery: 34.4 ± 12.1; post-surgery: 31.2 ± 16.1; p = 0.5; 15° flexion pre-surgery: 33.3 ± 6.1; post-surgery: 23.4 ± 8.6; p = 0.068). CCA increased in every flexion angle and group, but significant increase was seen only between 0°-15° (unloaded and loaded) in mild dysplasia of the trochlea, where significant increase in Dejour type C and D group was seen with unloaded full extension of the knee (0° flexion) and 30° flexion (unloaded and loaded). This study proves a significant effect of the MPFL graft to cartilage contact area, as well as an improvement of the patella tilt in patients with mild dysplasia of the trochlea. Thus, the MPFL can be used as a single treatment for patient with Dejour type A and B dysplasia. However, in patients with severe dysplasia the MPFL graft alone does not significantly increase CCA.

KCB-Net: A 3D knee cartilage and bone segmentation network via sparse annotation.

Knee cartilage and bone segmentation is critical for physicians to analyze and diagnose articular damage and knee osteoarthritis (OA). Deep learning (DL) methods for medical image segmentation have largely outperformed traditional methods, but they often need large amounts of annotated data for model training, which is very costly and time-consuming for medical experts, especially on 3D images. In this paper, we report a new knee cartilage and bone segmentation framework, KCB-Net, for 3D MR images based on sparse annotation. KCB-Net selects a small subset of slices from 3D images for annotation, and seeks to bridge the performance gap between sparse annotation and full annotation. Specifically, it first identifies a subset of the most effective and representative slices with an unsupervised scheme; it then trains an ensemble model using the annotated slices; next, it self-trains the model using 3D images containing pseudo-labels generated by the ensemble method and improved by a bi-directional hierarchical earth mover's distance (bi-HEMD) algorithm; finally, it fine-tunes the segmentation results using the primal-dual Internal Point Method (IPM). Experiments on four 3D MR knee joint datasets (the SKI10 dataset, OAI ZIB dataset, Iowa dataset, and iMorphics dataset) show that our new framework outperforms state-of-the-art methods on full annotation, and yields high quality results for small annotation ratios even as low as 10%.

A novel multi-atlas segmentation approach under the semi-supervised learning framework: Application to knee cartilage segmentation

Background and objective: Multi-atlas based segmentation techniques, which rely on an atlas library comprised of training images labeled by an expert, have proven their effectiveness in multiple automatic segmentation applications. However, the usage of exhaustive patch libraries combined with the voxel-wise labeling incur a large computational cost in terms of memory requirements and execution times. Methods: To confront this shortcoming, we propose a novel two-stage multi-atlas approach designed under the Semi-Supervised Learning (SSL) framework. The main properties of our method are as follows: First, instead of the voxel-wise labeling approach, the labeling of target voxels is accomplished here by exploiting the spectral content of globally sampled datasets from the target image, along with their spatially correspondent data collected from the atlases. Following SSL, voxels classification is boosted by incorporating unlabeled data from the target image, in addition to the labeled ones from atlas library. Our scheme integrates constructively fruitful concepts, including sparse reconstructions of voxels from linear neighborhoods, HOG feature descriptors of patches/regions, and label propagation via sparse graph constructions. Segmentation of the target image is carried out in two stages: stage-1 focuses on the sampling and labeling of global data, while stage-2 undertakes the above tasks for the out-of-sample data. Finally, we propose different graph-based methods for the labeling of global data, while these methods are extended to deal with the out-of-sample voxels. Results: A thorough experimental investigation is conducted on 76 subjects provided by the publicly accessible Osteoarthritis Initiative (OAI) repository. Comparative results and statistical analysis demonstrate that the suggested methodology exhibits superior segmentation performance compared to the existing patch-based methods, across all evaluation metrics (DSC:88.89%, Precision: 89.86%, Recall: 88.12%), while at the same time it requires a considerably reduced computational load (>70% reduction on average execution time with respect to other patch-based). In addition, our approach is favorably compared against other non patch-based and deep learning methods in terms of performance accuracy (on the 3-class problem). A final experimentation on a 5-class setting of the problems demonstrates that our approach is capable of achieving performance comparable to existing state-of-the-art knee cartilage segmentation methods (DSC:88.22% and DSC:85.84% for femoral and tibial cartilage respectively).

Cartilage Topography Assessment With Local-Area Cartilage Segmentation for Knee Magnetic Resonance Imaging.

Local-area cartilage segmentation (LACS) software was developed to segment medial femur (MF) cartilage on magnetic resonance imaging (MRI). Our objectives were 1) to extend LACS to the lateral femur (LF), medial tibia (MT), and lateral tibia (LT), 2) to compare LACS to an established manual segmentation method, and 3) to visualize cartilage responsiveness over each cartilage plate. Osteoarthritis Initiative participants with symptomatic knee osteoarthritis (OA) were selected, including knees selected at random (n=40) and knees identified with loss of cartilage based on manual segmentation (Chondrometrics GmbH), an enriched sample of 126 knees. LACS was used to segment cartilage in the MF, LF, MT, and LT on sagittal 3D double-echo steady-state MRI scans at baseline and at 2-year follow-up. We compared LACS and Chondrometrics average thickness measures by estimating the correlation in each cartilage plate and estimating the standardized response mean (SRM) for 2-year cartilage change. We illustrated cartilage loss topographically with SRM heatmaps. The estimated correlation between LACS and Chondrometrics measures was r=0.91 (95% confidence interval [95% CI] 0.86, 0.94) for LF, r=0.93 (95% CI 0.89, 0.95) for MF, r=0.97 (95% CI 0.96, 0.98) for LT, and r=0.87 (95% CI 0.81, 0.91) for MT. Estimated SRMs for LACS and Chondrometrics measures were similar in the random sample, and SRM heatmaps identified subregions of LACS-measured cartilage loss. LACS cartilage thickness measurement in the MF and LF and tibia correlated well with established manual segmentation-based measurement, with similar responsiveness to change, among knees with symptomatic knee OA. LACS measurement of cartilage plate topography enables spatiotemporal analysis of cartilage loss in future knee OA studies.

Performance and Robustness of Regional Image Segmentation Driven by Selected Evolutionary and Genetic Algorithms: Study on MR Articular Cartilage Images.

The analysis and segmentation of articular cartilage magnetic resonance (MR) images belongs to one of the most commonly routine tasks in diagnostics of the musculoskeletal system of the knee area. Conventional regional segmentation methods, which are based either on the histogram partitioning (e.g., Otsu method) or clustering methods (e.g., K-means), have been frequently used for the task of regional segmentation. Such methods are well known as fast and well working in the environment, where cartilage image features are reliably recognizable. The well-known fact is that the performance of these methods is prone to the image noise and artefacts. In this context, regional segmentation strategies, driven by either genetic algorithms or selected evolutionary computing strategies, have the potential to overcome these traditional methods such as Otsu thresholding or K-means in the context of their performance. These optimization strategies consecutively generate a pyramid of a possible set of histogram thresholds, of which the quality is evaluated by using the fitness function based on Kapur’s entropy maximization to find the most optimal combination of thresholds for articular cartilage segmentation. On the other hand, such optimization strategies are often computationally demanding, which is a limitation of using such methods for a stack of MR images. In this study, we publish a comprehensive analysis of the optimization methods based on fuzzy soft segmentation, driven by artificial bee colony (ABC), particle swarm optimization (PSO), Darwinian particle swarm optimization (DPSO), and a genetic algorithm for an optimal thresholding selection against the routine segmentations Otsu and K-means for analysis and the features extraction of articular cartilage from MR images. This study objectively analyzes the performance of the segmentation strategies upon variable noise with dynamic intensities to report a segmentation’s robustness in various image conditions for a various number of segmentation classes (4, 7, and 10), cartilage features (area, perimeter, and skeleton) extraction preciseness against the routine segmentation strategies, and lastly the computing time, which represents an important factor of segmentation performance. We use the same settings on individual optimization strategies: 100 iterations and 50 population. This study suggests that the combination of fuzzy thresholding with an ABC algorithm gives the best performance in the comparison with other methods as from the view of the segmentation influence of additive dynamic noise influence, also for cartilage features extraction. On the other hand, using genetic algorithms for cartilage segmentation in some cases does not give a good performance. In most cases, the analyzed optimization strategies significantly overcome the routine segmentation methods except for the computing time, which is normally lower for the routine algorithms. We also publish statistical tests of significance, showing differences in the performance of individual optimization strategies against Otsu and K-means method. Lastly, as a part of this study, we publish a software environment, integrating all the methods from this study.

Automated quantification of cartilage quality for hip treatment decision support

PurposePreservation surgery can halt the progress of joint degradation, preserving the life of the hip; however, outcome depends on the existing cartilage quality. Biochemical analysis of the hip cartilage utilizing MRI sequences such as delayed gadolinium-enhanced MRI of cartilage (dGEMRIC), in addition to morphological analysis, can be used to detect early signs of cartilage degradation. However, a complete, accurate 3D analysis of the cartilage regions and layers is currently not possible due to a lack of diagnostic tools.MethodsA system for the efficient automatic parametrization of the 3D hip cartilage was developed. 2D U-nets were trained on manually annotated dual-flip angle (DFA) dGEMRIC for femoral head localization and cartilage segmentation. A fully automated cartilage sectioning pipeline for analysis of central and peripheral regions, femoral-acetabular layers, and a variable number of section slices, was developed along with functionality for the automatic calculation of dGEMRIC index, thickness, surface area, and volume.ResultsThe trained networks locate the femoral head and segment the cartilage with a Dice similarity coefficient of 88 ± 3 and 83 ± 4% on DFA and magnetization-prepared 2 rapid gradient-echo (MP2RAGE) dGEMRIC, respectively. A completely automatic cartilage analysis was performed in 18s, and no significant difference for average dGEMRIC index, volume, surface area, and thickness calculated on manual and automatic segmentation was observed.ConclusionAn application for the 3D analysis of hip cartilage was developed for the automated detection of subtle morphological and biochemical signs of cartilage degradation in prognostic studies and clinical diagnosis. The segmentation network achieved a 4-time increase in processing speed without loss of segmentation accuracy on both normal and deformed anatomy, enabling accurate parametrization. Retraining of the networks with the promising MP2RAGE protocol would enable analysis without the need for B1 inhomogeneity correction in the future.

Automatic femoral articular cartilage segmentation using deep learning in three-dimensional ultrasound images of the knee

ObjectiveThis study aimed to develop a deep learning-based approach to automatically segment the femoral articular cartilage (FAC) in 3D ultrasound (US) images of the knee to increase time efficiency and decrease rater variability. DesignOur method involved deep learning predictions on 2DUS slices sampled in the transverse plane to view the cartilage of the femoral trochlea, followed by reconstruction into a 3D surface. A 2D U-Net was modified and trained using a dataset of 200 2DUS images resliced from 20 3DUS images. Segmentation accuracy was evaluated using a holdout dataset of 50 2DUS images resliced from 5 3DUS images. Absolute and signed error metrics were computed and FAC segmentation performance was compared between rater 1 and 2 manual segmentations. ResultsOur U-Net-based algorithm performed with mean 3D DSC, recall, precision, VPD, MSD, and HD of 73.1 ​± ​3.9%, 74.8 ​± ​6.1%, 72.0 ​± ​6.3%, 10.4 ​± ​6.0%, 0.3 ​± ​0.1 ​mm, and 1.6 ​± ​0.7 ​mm, respectively. Compared to the individual 2D predictions, our algorithm demonstrated a decrease in performance after 3D reconstruction, but these differences were not found to be statistically significant. The percent difference between the manually segmented volumes of the 2 raters was 3.4%, and rater 2 demonstrated the largest VPD with 14.2 ​± ​11.4 ​mm3 compared to 10.4 ​± ​6.0 ​mm3 for rater 1. ConclusionThis study investigated the use of a modified U-Net algorithm to automatically segment the FAC in 3DUS knee images of healthy volunteers, demonstrating that this segmentation method would increase the efficiency of anterior femoral cartilage volume estimation and expedite the post-acquisition processing for 3D US images of the knee.

Deep collaborative network with alpha matte for precise knee tissue segmentation from MRI

Background and objectivePrecise segmentation of knee tissues from magnetic resonance imaging (MRI) is critical in quantitative imaging and diagnosis. Convolutional neural networks (CNNs), being state of the art, often challenged by the lack of image-specific adaptation, such as low tissue contrasts and structural inhomogeneities, thereby leading to incomplete segmentation results. MethodsThis paper presents a deep learning–based automatic segmentation framework for precise knee tissue segmentation. A novel deep collaborative method is proposed, which consists of an encoder–decoder-based segmentation network in combination with a low rank tensor-reconstructed segmentation network. Low rank reconstruction in MRI tensor sub-blocks is introduced to exploit the morphological variations in knee tissues. To model the tissue boundary regions and effectively utilize the superimposed regions, trimap generation is proposed for defining high, medium and low confidence regions from the multipath CNNs. The secondary path with low rank reconstructed input mitigates the conditions in which the primary segmentation network can potentially fail and overlook the boundary regions. The outcome of the segmentation is solved as an alpha matting problem by blending the trimap with the source input. ResultsExperiments on Osteoarthritis Initiative (OAI) datasets with all the 6 musculoskeletal tissues provide an overall segmentation dice score of 0.8925, where Femoral and Tibial part of cartilage achieving an average dice exceeding 0.9. The volumetric metrics also indicate the superior performances in all tissue compartments. ConclusionsExperiments on Osteoarthritis Initiative (OAI) datasets and a self-prepared scan validate the effectiveness of the proposed method. Inclusion of extra prediction scale allowed the model to distinguish and segment the tissue boundary accurately. We specifically demonstrate the application of the proposed method in a cartilage segmentation–based thickness map for diagnosis purposes.

Segmentation of knee MRI data with convolutional neural networks for semi-automated three-dimensional surface-based analysis of cartilage morphology and composition

To assess automatic segmentations for surface-based analysis of cartilage morphology and composition on knee magnetic resonance (MR) images. 2D and 3D U-Nets were trained on double echo steady state (DESS) images from the publicly available Osteoarthritis Initiative (OAI) dataset with femoral and tibial bone and cartilage segmentations provided by the Zuse Institute Berlin (ZIB). The U-Nets were used to perform automatic segmentation of femoral and tibial bone-cartilage structures (bone and cartilage segmentations combined into one structure) from the DESS images. T 2 -weighted images from the OAI dataset were registered to the DESS images and used for T 2 map calculation. Using the 3D cartilage surface mapping (3D-CaSM) method, surface-based analysis of cartilage morphology (thickness) and composition ( T 2 ) was performed using both manual and network-generated segmentations from OAI ZIB testing images. Bland-Altman analyses were performed to evaluate the accuracy of the extracted cartilage thickness and T 2 measurements from both U-Nets compared to manual segmentations. Bland-Altman analysis showed a mean bias [95% limits of agreement] for femoral and tibial cartilage thickness measurements ranging between -0.12 to 0.33 [-0.28, 0.96] mm with 2D U-Net and 0.07 to 0.14 [-0.14, 0.39] mm with 3D U-Net. For T 2 , the mean bias [95% limits of agreement] ranged between -0.16 to 1.32 [-4.71, 4.83] ms with 2D U-Net and -0.05 to 0.46 [-2.47, 3.39] ms with 3D U-Net. While both 2D and 3D U-Nets exemplified the time-efficiency benefit of using deep learning methods for generating the required segmentations, segmentations from 3D U-Nets demonstrated higher accuracy in the extracted thickness and T 2 features using 3D-CaSM compared to the segmentations from 2D U-Nets.

Quantitative T2 and T1ρ mapping are sensitive to ischemic injury to the epiphyseal cartilage in an in vivo piglet model of Legg-Calvé-Perthes disease

To determine if the quantitative MRI techniques T2 and T1ρ mapping are sensitive to ischemic injury to epiphyseal cartilage in vivo in a piglet model of Legg-Calvé-Perthes disease using a clinical 3T MRI scanner. We hypothesized that T2 and T1ρ relaxation times would be increased in the epiphyseal cartilage of operated vs contralateral-control femoral heads 1week following onset of ischemia. Unilateral femoral head ischemia was surgically induced in eight piglets. Piglets were imaged 1week post-operatively in vivo at 3T MRI using a magnetization-prepared 3D fast spin echo sequence for T2 and T1ρ mapping and a 3D gradient echo sequence for cartilage segmentation. Ischemia was confirmed in all piglets using gadolinium contrast-enhanced MRI. Median T2 and T1ρ relaxation times were measured in the epiphyseal cartilage of the ischemic and control femoral heads and compared using paired t-tests. Histological assessment was performed on a subset of five piglets. T2 and T1ρ relaxation times were significantly increased in the epiphyseal cartilage of the operated vs control femoral heads (ΔT2=11.9±3.7ms, 95% CI=[8.8, 15.0] ms, P<0.0001; ΔT1ρ=12.8±4.1ms, 95% CI=[9.4, 16.2] ms, P<0.0001). Histological assessment identified chondronecrosis in the hypertrophic and deep proliferative zones within ischemic epiphyseal cartilage. T2 and T1ρ mapping are sensitive to ischemic injury to the epiphyseal cartilage in vivo at clinical 3T MRI. These techniques may be clinically useful to assess injury and repair to the epiphyseal cartilage to better stage the extent of ischemic damage in Legg-Calvé-Perthes disease.

Knee Bone and Cartilage Segmentation Based on a 3D Deep Neural Network Using Adversarial Loss for Prior Shape Constraint.

Fast and accurate segmentation of knee bone and cartilage on MRI images is becoming increasingly important in the orthopaedic area, as the segmentation is an essential prerequisite step to a patient-specific diagnosis, optimising implant design and preoperative and intraoperative planning. However, manual segmentation is time-intensive and subjected to inter- and intra-observer variations. Hence, in this study, a three-dimensional (3D) deep neural network using adversarial loss was proposed to automatically segment the knee bone in a resampled image volume in order to enlarge the contextual information and incorporate prior shape constraints. A restoration network was proposed to further improve the bone segmentation accuracy by restoring the bone segmentation back to the original resolution. A conventional U-Net-like network was used to segment the cartilage. The ultimate results were the combination of the bone and cartilage outcomes through post-processing. The quality of the proposed method was thoroughly assessed using various measures for the dataset from the Grand Challenge Segmentation of Knee Images 2010 (SKI10), together with a comparison with a baseline network U-Net. A fine-tuned U-Net-like network can achieve state-of-the-art results without any post-processing operations. This method achieved a total score higher than 76 in terms of the SKI10 validation dataset. This method showed to be robust to extract bone and cartilage masks from the MRI dataset, even for the pathological case.

Analysis of Knee Joint Injury Caused by Physical Training of Freshmen Students Based on 3T MRI and Automatic Cartilage Segmentation Technology: A Prospective Study.

BackgroundThe differential effects of various exercises on knee joint injury have not been well documented. Improper physical training can cause irreversible damage to the knee joint. MRI is generally used to precisely analyze morphological and biochemical changes in the knee cartilage. We compared the effects of long-walking and regular daily physical training on acute and chronic knee joint injuries as well as cartilage structure in freshmen students.MethodsA total of 23 young male college freshmen were recruited to participate in an 8-day 240 km long distance walk and a one-year daily training. 3D-DESSwe, 2D T2 mapping, DIXON, and T1WI of the right knee joint were performed using the MAGNETOM Spectra 3T MR scanner. The injury of meniscus, bone marrow edema, ligaments and joint effusion is graded. Cartilage volume, thickness and T2 values of 21 sub-regions of the knee cartilage were estimated using automatic cartilage segmentation prototype software. Friedman’s test and Wilcoxon paired rank-sum test were used to compare quantitative indices of knee cartilage in three groups.ResultsThe injury to the medial meniscus and anterior cruciate ligament of the knee joint, joint effusion, and bone marrow edema was significantly higher in the long-walking group compared to the baseline and daily groups. Furthermore, injury to the lateral meniscus was significantly worse in the long-walking group compared to the baseline group but was significantly better in the daily group compared to the baseline group. No significant changes to the posterior cruciate ligament were observed among the three groups. Knee cartilage volume was significantly increased, mainly in the stress surface of the femur, patella, and the lateral area of the tibial plateau. Regular daily training did not significantly change the thickness of the knee cartilage. Conversely, knee cartilage thickness decreased in the long-walking group, especially in the medial and lateral areas of the femur and tibial plateau. Moreover, no significant changes were observed in the knee cartilage volume of the long-walking group. Both long-walking and daily groups showed reduced T2 values of the knee joint compared to the baseline.ConclusionAmong freshmen students and the training of this experimental intensity, our results show that regular daily training does not cause high-level injury to the knee joint, but improve the knee joint function adaptability by increasing cartilage volume. Moreover, knee injury caused by short-term long walking can be reversible.

Automated knee cartilage segmentation for heterogeneous clinical MRI using generative adversarial networks with transfer learning.

This study aimed to build a deep learning model to automatically segment heterogeneous clinical MRI scans by optimizing a pre-trained model built from a homogeneous research dataset with transfer learning. Conditional generative adversarial networks pretrained on the Osteoarthritis Initiative MR images was transferred to 30 sets of heterogenous MR images collected from clinical routines. Two trained radiologists manually segmented the 30 sets of clinical MR images for model training, validation and test. The model performance was compared to models trained from scratch with different datasets, as well as two radiologists. A 5-fold cross validation was performed. The transfer learning model obtained an overall averaged Dice coefficient of 0.819, an averaged 95 percentile Hausdorff distance of 1.463 mm, and an averaged average symmetric surface distance of 0.350 mm on the 5 random holdout test sets. A 5-fold cross validation had a mean Dice coefficient of 0.801, mean 95 percentile Hausdorff distance of 1.746 mm, and mean average symmetric surface distance of 0.364 mm. It outperformed other models and performed similarly as the radiologists. A transfer learning model was able to automatically segment knee cartilage, with performance comparable to human, using heterogeneous clinical MR images with a small training data size. In addition, the model proved robust when tested through cross validation and on images from a different vendor. We found it feasible to perform fully automated cartilage segmentation of clinical knee MR images, which would facilitate the clinical application of quantitative MRI techniques and other prediction models for improved patient treatment planning.

Improved-Mask R-CNN: Towards an accurate generic MSK MRI instance segmentation platform (data from the Osteoarthritis Initiative).

Objective assessment of osteoarthritis (OA) Magnetic Resonance Imaging (MRI) scans can address the limitations of the current OA assessment approaches. Detecting and extracting bone, cartilage, and joint fluid is a necessary component for the objective assessment of OA, which helps to quantify tissue characteristics such as volume and thickness. Many algorithms, based on Artificial Intelligence (AI), have been proposed over recent years for segmenting bone and soft tissues. Most of these segmentation methods suffer from the class imbalance problem, can't differentiate between the same anatomic structure, or do not support segmenting different rang of tissue sizes. Mask R-CNN is an instance segmentation framework, meaning it segments and distinct each object of interest like different anatomical structures (e.g. bone and cartilage) using a single model. In this study, the Mask R-CNN architecture was deployed to address the need for a segmentation method that is applicable to use for different tissue scales, pathologies, and MRI sequences associated with OA, without having a problem with imbalanced classes. In addition, we modified the Mask R-CNN to improve segmentation accuracy around instance edges. A total of 500 adult knee MRI scans from the publicly available Osteoarthritis Initiative (OAI), and 97 hip MRI scans from adults with symptomatic hip OA, evaluated by two readers, were used for training and validating the network. Three specific modifications to Mask R-CNN yielded the improved-Mask R-CNN (iMaskRCNN): an additional ROIAligned block, an extra decoder block in the segmentation header, and connecting them using a skip connection. The results were evaluated using Hausdorff distance, dice score for bone and cartilage segmentation, and differences in detected volume, dice score, and coefficients of variation (CoV) for effusion segmentation. The iMaskRCNN led to improved bone and cartilage segmentation compared to Mask RCNN as indicated with the increase in dice score from 95% to 98% for the femur, 95-97% for the tibia, 71-80% for the femoral cartilage, and 81-82% for the tibial cartilage. For the effusion detection, the dice score improved with iMaskRCNN 72% versus Mask R-CNN 71%. The CoV values for effusion detection between Reader1 and Mask R-CNN (0.33), Reader1 and iMaskRCNN (0.34), Reader2 and Mask R-CNN (0.22), Reader2 and iMaskRCNN (0.29) are close to CoV between two readers (0.21), indicating a high agreement between the human readers and both Mask R-CNN and iMaskRCNN. Mask R-CNN and iMaskRCNN can reliably and simultaneously extract different scale articular tissues involved in OA, forming the foundation for automated assessment of OA. The iMaskRCNN results show that the modification improved the network performance around the edges.

Detection of Differences in Longitudinal Cartilage Thickness Loss Using a Deep-Learning Automated Segmentation Algorithm: Data From the Foundation for the National Institutes of Health Biomarkers Study of the Osteoarthritis Initiative.

ObjectiveTo study the longitudinal performance of fully automated cartilage segmentation in knees with radiographic osteoarthritis (OA), we evaluated the sensitivity to change in progressor knees from the Foundation for the National Institutes of Health OA Biomarkers Consortium between the automated and previously reported manual expert segmentation, and we determined whether differences in progression rates between predefined cohorts can be detected by the fully automated approach.MethodsThe OA Initiative Biomarker Consortium was a nested case–control study. Progressor knees had both medial tibiofemoral radiographic joint space width loss (≥0.7 mm) and a persistent increase in Western Ontario and McMaster Universities Osteoarthritis Index pain scores (≥9 on a 0–100 scale) after 2 years from baseline (n = 194), whereas non‐progressor knees did not have either of both (n = 200). Deep‐learning automated algorithms trained on radiographic OA knees or knees of a healthy reference cohort (HRC) were used to automatically segment medial femorotibial compartment (MFTC) and lateral femorotibial cartilage on baseline and 2‐year follow‐up magnetic resonance imaging. Findings were compared with previously published manual expert segmentation.ResultsThe mean ± SD MFTC cartilage loss in the progressor cohort was –181 ± 245 μm by manual segmentation (standardized response mean [SRM] –0.74), –144 ± 200 μm by the radiographic OA–based model (SRM –0.72), and –69 ± 231 μm by HRC‐based model segmentation (SRM –0.30). Cohen's d for rates of progression between progressor versus the non‐progressor cohort was –0.84 (P < 0.001) for manual, –0.68 (P < 0.001) for the automated radiographic OA model, and –0.14 (P = 0.18) for automated HRC model segmentation.ConclusionA fully automated deep‐learning segmentation approach not only displays similar sensitivity to change of longitudinal cartilage thickness loss in knee OA as did manual expert segmentation but also effectively differentiates longitudinal rates of loss of cartilage thickness between cohorts with different progression profiles.

AGREEMENT AND SENSITIVITY TO BETWEEN-GROUP DIFFERENCES OF U-NET-BASED, AUTOMATED CARTILAGE T2 ANALYSES FROM MULTI-ECHO SPIN ECHO MRI - DATA FROM THE OA-BIO PROJECT

Purpose: Previous studies reported a high agreement for cartilage segmentations from high resolution 3D gradient echo MRIs using deep learning techniques when compared to manual cartilage segmentations with Dice similarity coefficients (DSC) of up to ∼0.90 (scale: 0.0 to 1.0). Only few studies applied deep learning techniques to multi-echo spin echo (MESE) MRIs for analysis of cartilage T2 relaxation time, an imaging biomarker sensitive to alterations in cartilage composition. These reported a somewhat lower agreement compared to manual segmentations with DSCs of up to 0.75.

RELEVANCE OF WITHIN-GRADE CARTILAGE ASSESSMENT USING THE MOAKS SCORING INSTRUMENT AND CORRESPONDING QUANTITATIVE CARTILAGE THICKNESS LOSS: DATA FROM THE OAI FNIH BIOMARKERS CONSORTIUM STUDY

Purpose: Both semi-quantitative (SQ) MOAKS cartilage scores and quantitative (Q) cartilage thickness measurements are used as outcome measures in observational studies and interventional knee osteoarthritis (OA) trials. Both methods are considered complementary. SQ scoring by experts is considered less sensitive to change compared to Q approaches based on cartilage segmentation particularly regarding subtle wide-spread cartilage loss. On the other hand, change in focal defects commonly is not covered by Q approaches as these may be too subtle in the context of segmentation of an entire subregion.

Magnetic resonance imaging quantitative analysis of knee joint injury and cartilage before and after long-distance march and training for college students

Objective: To assess the effects of long-distance march and training on acute knee injury and knee cartilage sub-regions of college students using quantitatively magnetic resonance imaging analysis. Methods: Twenty-seven young male students from freshman classes in the Army Military Medical University were enrolled in September 2019, aged from 17 to 20 (19.48±0.14) years, participated in the whole 8-day, 240 km long-distance march and training. Three-dimensional quantitative MRI was performed on the right knee using high-field MRI before (baseline) and 1 day after (follow-up) march. The assessment indexes included: meniscus and cartilage injury(5-point scale), bone marrow and ligament injury, and joint effusion(3-point scale). Using semi-automatic cartilage segmentation and 3D data post-processing techniques, a total of 21 sub-regions of cartilage volume and thickness were measured in the medial and lateral femur, medial and lateral tibia. Paired-samples t-test was used to compare the changes in quantitative cartilage indices of the knee joint before and after march. Results: In terms of acute knee injuries, medial and lateral meniscus injuries (grade 1-3) occurred in 8 and 9 college students, respectively compared with those before march. Anterior cruciate ligaments injury occurred in 4 college students and developed from grade 0 to grade 1. Bone marrow edema occurred in 10 students and developed from grade 0 to grade 2, and in 5 students from grade 0 to grade 1. Joint effusion occurred in 5 college students and developed from grade 1 to grade 2 (all P<0.05). In terms of quantitative analysis of cartilage subregion of knee joint, the volume of central region of femoral pulley increased [(1.84±0.32) mm3 vs (1.67±0.29) mm3] and the volume of central region of medial femoral condyle decreased [(1.18±0.21) mm3 vs (1.26±0.17) mm3] compared with that before march (all P<0.05); The cartilage thickness of 11 cartilage subregion [(1.37±0.27) mm vs (1.53±0.18) mm], [(1.42±0.25) mm vs (1.54±0.17) mm], [(1.53±0.20) mm vs (1.62±0.20) mm], [(1.72±0.28) mm vs (1.83±0.28) mm], [(1.84±0.45) mm vs (2.04±0.42) mm], [(2.20±0.58) mm vs (2.46±0.50) mm], [(1.74±0.19) mm vs (1.85±0.21) mm] [(1.45±0.21) mm vs (1.58±0.16) mm], [(1.81±0.22) mm vs (1.91±0.15) mm], [(1.44±0.13) mm vs (1.53±0.15) mm] was thinner than that before march (all P<0.05). The T2 values of 7 cartilage subregion [(40.57±26.23) ms vs (67.10±47.46) ms], [(80.10±20.56) ms vs (98.42±23.58) ms], [(87.92±24.95) ms vs (108.84±29.24) ms], [(50.49±19.18) ms vs (76.97±37.16) ms], [(38.89±15.82) ms vs (69.70±40.16) ms] [(55.84±24.53) ms vs (106.35±50.01) ms] and [(72.38±36.64) ms vs (105.31±39.34) ms] were lower than those before march, while the T2 values of the two subregions of patellar cartilage [(102.13±44.47) ms vs (72.20±28.37) ms], [(97.42±44.86) ms vs (76.67±51.64) ms] were higher than those before march (P<0.05). There was no significant difference in other cartilage subareas (P>0.05). Conclusions: Long distance hiking will lead to acute injury of knee joint of young college students. The thickness of cartilage subregion of knee joint showed a thinning trend as a whole, while the volume and T2 values showed different trends.