Search

BACKGROUND: The polyarticular nature of Osteoarthritis (OA) tends to manifest in multi-joints. Associations between cartilage health in connected joints can help identify early degeneration and offer the potential for biomechanical intervention. Such associations between hip and knee cartilages remain understudied. PURPOSE: To investigate T1p associations between hip-femoral and acetabular-cartilage subregions with Intra-limb and Inter-limb patellar cartilage; whole and deep-medial (DM), deep-lateral (DL), superficial-medial (SM), superficial-lateral (SL) subregions. STUDY TYPE: Prospective. SUBJECTS: Twenty-eight subjects (age 55.1 ± 12.8 years, 15 females) with none-to-moderate hip-OA while no radiographic knee-OA. FIELD STRENGTH/SEQUENCE: 3-T, bilateral hip, and knee: 3D-proton-density-fat-saturated (PDFS) Cube and Magnetization-Prepared-Angle-Modulated-Partitioned-k-Space-Spoiled-Gradient-Echo-Snapshots (MAPSS). ASSESSMENT: Ages of subjects were categorized into Group-1 (≤40), Group-2 (41-50), Group-3 (51-60), Group-4 (61-70), Group-5 (71-80), and Group-6 (≥81). Hip T1p maps, co-registered to Cube, underwent an atlas-based algorithm to quantify femoral and acetabular subregional (R2-R7) cartilage T1p. For knee Cube, a combination of V-Net architectures was used to segment the patellar cartilage and subregions (DM, DL, SM, SL). T1p values were computed from co-registered MAPSS. STATISTICAL TESTS: For Intra-and-Inter-limb, 5 optimum predictors out of 13 (Hip subregional T1p, age group, gender) were selected by univariate linear-regression, to predict outcome (patellar T1p). The top five predictors were stepwise added to six linear mixed-effect (LME) models. In all LME models, we assume the data come from the same subject sharing the same random effect. The best-performing models (LME-modelbest) selected via ANOVA, were tested with DM, SM, SL, and DL subregional-mean T1p. LME assumptions were verified (normality of residuals, random-effects, and posterior-predictive-checks). RESULTS: LME-modelbest (Intra-limb) had significant negative and positive fixed-effects of femoral-R5 and acetabular-R2 T1p, respectively (conditional-R2 = 0.581). LME-modelbest (Inter-limb) had significant positive fixed-effects of femoral-R3 T1p (conditional-R2 = 0.26). DATA CONCLUSION: Significant positive and negative T1p associations were identified between load-bearing hip cartilage-subregions vs. ipsilateral and contralateral patellar cartilages respectively. The effects were localized on medial subregions of Inter-limb, in particular. EVIDENCE LEVEL: 1 TECHNICAL EFFICACY: Stage 1.

PurposeHigh quality scan prescription that optimally covers the area of interest with scan planes aligned to relevant anatomical structures is crucial for error-free radiologic interpretation. The goal of this project was to develop a machine learning pipeline for oblique scan prescription that could be trained on localizer images and metadata from previously acquired MR exams.MethodsA novel Multislice Rotational Region-based Convolutional Neural Network (MS-R2CNN) architecture was developed. Based on this architecture, models for automated prescription sagittal lumbar spine acquisitions from axial, sagittal, and coronal localizer slices were trained. The automated prescription pipeline was integrated with the scanner console software and evaluated in experiments with healthy volunteers (N = 3) and patients with lower-back pain (N = 20).ResultsExperiments in healthy volunteers demonstrated high accuracy of automated prescription in all subjects. There was good agreement between alignment and coverage of manual and automated prescriptions, as well as consistent views of the lumbar spine at different positions of the subjects within the scanner bore. In patients with lower-back pain, the generated prescription was applied in 18 cases (90% of the total number). None of the cases required major adjustment, while in 11 cases (55%) there were minor manual adjustments to the generated prescription.ConclusionsThis study demonstrates the ability of oriented object detection-based models to be trained to prescribe oblique lumbar spine MRI acquisitions without the need of manual annotation or feature engineering and the feasibility of using machine learning-based pipelines on the scanner for automated prescription of MRI acquisitions.

BackgroundParaspinal muscle fat infiltration is associated with spinal degeneration and low back pain, however, quantifying muscle fat using clinical magnetic resonance imaging (MRI) techniques continues to be a challenge. Advanced MRI techniques, including chemical-shift encoding (CSE) based water-fat MRI, enable accurate measurement of muscle fat, but such techniques are not widely available in routine clinical practice.MethodsTo facilitate assessment of paraspinal muscle fat using clinical imaging, we compared four thresholding approaches for estimating muscle fat fraction (FF) using T1- and T2-weighted images, with measurements from water-fat MRI as the ground truth: Gaussian thresholding, Otsu's method, K-mean clustering, and quadratic discriminant analysis. Pearson's correlation coefficients (r), mean absolute errors, and mean bias errors were calculated for FF estimates from T1- and T2-weighted MRI with water-fat MRI for the lumbar multifidus (MF), erector spinae (ES), quadratus lumborum (QL), and psoas (PS), and for all muscles combined.ResultsWe found that for all muscles combined, FF measurements from T1- and T2-weighted images were strongly positively correlated with measurements from the water-fat images for all thresholding techniques (r = 0.70-0.86, p

A 2D U-Net was trained to generate synthetic T1p maps from T2 maps for knee MRI to explore the feasibility of domain adaptation for enriching existing datasets and enabling rapid, reliable image reconstruction. The network was developed using 509 healthy contralateral and injured ipsilateral knee images from patients with ACL injuries and reconstruction surgeries acquired across three institutions. Network generalizability was evaluated on 343 knees acquired in a clinical setting and 46 knees from simultaneous bilateral acquisition in a research setting. The deep neural network synthesized high-fidelity reconstructions of T1p maps, preserving textures and local T1p elevation patterns in cartilage with a normalized mean square error of 2.4% and Pearsons correlation coefficient of 0.93. Analysis of reconstructed T1p maps within cartilage compartments revealed minimal bias (-0.10 ms), tight limits of agreement, and quantification error (5.7%) below the threshold for clinically significant change (6.42%) associated with osteoarthritis. In an out-of-distribution external test set, synthetic maps preserved T1p textures, but exhibited increased bias and wider limits of agreement. This study demonstrates the capability of image synthesis to reduce acquisition time, derive meaningful information from existing datasets, and suggest a pathway for standardizing T1p as a quantitative biomarker for osteoarthritis.

BACKGROUND: Rates of cartilage degeneration in asymptomatic elite basketball players are significantly higher compared with the general population due to excessive loads on the knee. Compositional quantitative magnetic resonance imaging (qMRI) techniques can identify local biochemical changes of macromolecules observed in cartilage degeneration. PURPOSE/HYPOTHESIS: The purpose of this study was to utilize multiparametric qMRI to (1) quantify how T1ρ and T2 relaxation times differ based on the presence of anatomic abnormalities and (2) correlate T1ρ and T2 with self-reported functional deficits. It was hypothesized that prolonged relaxation times will be associated with knees with MRI-graded abnormalities and knees belonging to basketball players with greater self-reported functional deficits. STUDY DESIGN: Cross-sectional study; Level of evidence, 3. METHODS: A total of 75 knees from National Collegiate Athletic Association Division I basketball players (40 female, 35 male) were included in this multicenter study. All players completed the Knee injury and Osteoarthritis Outcome Score (KOOS) and had bilateral knee MRI scans taken. T1ρ and T2 were calculated on a voxel-by-voxel basis. The cartilage surfaces were segmented into 6 compartments: lateral femoral condyle, lateral tibia, medial femoral condyle, medial tibia (MT), patella (PAT), and trochlea (TRO). Lesions from the MRI scans were graded for imaging abnormalities, and statistical parametric mapping was performed to study cross-sectional differences based on MRI scan grading of anatomic knee abnormalities. Pearson partial correlations between relaxation times and KOOS subscore values were computed, obtaining r value statistical parametric mappings and P value clusters. RESULTS: Knees without patellar tendinosis displayed significantly higher T1ρ in the PAT compared with those with patellar tendinosis (average percentage difference, 10.4%; P = .02). Significant prolongation of T1ρ was observed in the MT, TRO, and PAT of knees without compared with those with quadriceps tendinosis (average percentage difference, 12.7%, 13.3%, and 13.4%, respectively; P ≤ .05). A weak correlation was found between the KOOS-Symptoms subscale values and T1ρ/T2. CONCLUSION: Certain tissues that bear the brunt of impact developed tendinosis but spared cartilage degeneration. Whereas participants reported minimal functional deficits, their high-impact activities resulted in structural damage that may lead to osteoarthritis after their collegiate careers.

ObjectiveTo assess (i) the impact of changes in body weight on changes in joint-adjacent subcutaneous fat (SCF) and cartilage thickness over 4 years and (ii) the relation between changes in joint-adjacent SCF and knee cartilage thickness.DesignIndividuals from the Osteoarthritis Initiative (total=399) with > 10% weight gain (n=100) and > 10% weight loss (n=100) over 4 years were compared to a matched control cohort with less than 3% change in weight (n=199). 3.0T Magnetic Resonance Imaging (MRI) of the right knee was performed at baseline and after 4 years to quantify joint-adjacent SCF and cartilage thickness. Linear regression models were used to evaluate the associations between the (i) weight change group and 4-year changes in both knee SCF and cartilage thickness, and (ii) 4-year changes in knee SCF and in cartilage thickness. Analyses were adjusted for age, sex, baseline body mass index (BMI), tibial diameter (and weight change group in analysis (ii)).ResultsIndividuals who lost weight over 4-years had significantly less joint-adjacent SCF (beta range, medial/lateral joint sides: 2.2-4.2 mm, p

BACKGROUND: The purpose of this study was to develop a deep learning approach to automatically segment the scapular bone on magnetic resonance imaging (MRI) images and to compare the accuracy of these three-dimensional (3D) models with that of 3D computed tomography (CT). METHODS: Fifty-five patients with high-resolution 3D fat-saturated T2 MRI were retrospectively identified. The underlying pathology included rotator cuff tendinopathy and tears, shoulder instability, and impingement. Two experienced musculoskeletal researchers manually segmented the scapular bone. Five cross-validation training and validation splits were generated to independently train two-dimensional (2D) and 3D models using a convolutional neural network approach. Model performance was evaluated using the Dice similarity coefficient (DSC). All models with DSC > 0.70 were ensembled and used for the test set, which consisted of four patients with matching high-resolution MRI and CT scans. Clinically relevant glenoid measurements, including glenoid height, width, and retroversion, were calculated for two of the patients. Paired t-tests and Wilcoxon signed-rank tests were used to compare the DSC of the models. RESULTS: The 2D and 3D models achieved a best DSC of 0.86 and 0.82, respectively, with no significant difference observed. Augmentation of imaging data significantly improved 3D but not 2D model performance. In comparing clinical measurements of 3D MRI and CT, there was a mean difference ranging from 1.29 mm to 3.46 mm and 0.05° to 7.47°. CONCLUSION: We have presented a fully automatic, deep learning-based strategy for extracting scapular shape from a high-resolution MRI scan. Further developments of this technology have the potential to allow for surgeons to obtain all clinically relevant information from MRI scans and reduce the need for multiple imaging studies for patients with shoulder pathology.

STUDY DESIGN: In vivo retrospective study of fully automatic quantitative imaging feature extraction from clinically acquired lumbar spine magnetic resonance imaging (MRI). OBJECTIVE: To demonstrate the feasibility of substituting automatic for human-demarcated segmentation of major anatomic structures in clinical lumbar spine MRI to generate quantitative image-based features and biomechanical models. SETTING: Previous studies have demonstrated the viability of automatic segmentation applied to medical images; however, the feasibility of these networks to segment clinically acquired images has not yet been demonstrated, as they largely rely on specialized sequences or strict quality of imaging data to achieve good performance. METHODS: Convolutional neural networks were trained to demarcate vertebral bodies, intervertebral disc, and paraspinous muscles from sagittal and axial T1-weighted MRIs. Intervertebral disc height, muscle cross-sectional area, and subject-specific musculoskeletal models of tissue loading in the lumbar spine were then computed from these segmentations and compared against those computed from human-demarcated masks. RESULTS: Segmentation masks, as well as the morphological metrics and biomechanical models computed from those masks, were highly similar between human- and computer-generated methods. Segmentations were similar, with Dice similarity coefficients of 0.77 or greater across networks, and morphological metrics and biomechanical models were similar, with Pearson R correlation coefficients of 0.69 or greater when significant. CONCLUSIONS: This study demonstrates the feasibility of substituting computer-generated for human-generated segmentations of major anatomic structures in lumbar spine MRI to compute quantitative image-based morphological metrics and subject-specific musculoskeletal models of tissue loading quickly, efficiently, and at scale without interrupting routine clinical care.

In 2019, the National Health Interview survey found that nearly 59% of adults reported pain some, most, or every day in the past 3 months, with 39% reporting back pain, making back pain the most prevalent source of pain, and a significant issue among adults. Often, identifying a direct, treatable cause for back pain is challenging, especially as it is often attributed to complex, multifaceted issues involving biological, psychological, and social components. Due to the difficulty in treating the true cause of chronic low back pain (cLBP), an over-reliance on opioid pain medications among cLBP patients has developed, which is associated with increased prevalence of opioid use disorder and increased risk of death. To combat the rise of opioid-related deaths, the National Institutes of Health (NIH) initiated the Helping to End Addiction Long-TermSM (HEAL) initiative, whose goal is to address the causes and treatment of opioid use disorder while also seeking to better understand, diagnose, and treat chronic pain. The NIH Back Pain Consortium (BACPAC) Research Program, a network of 14 funded entities, was launched as a part of the HEAL initiative to help address limitations surrounding the diagnosis and treatment of cLBP. This paper provides an overview of the BACPAC research program's goals and overall structure, and describes the harmonization efforts across the consortium, define its research agenda, and develop a collaborative project which utilizes the strengths of the network. The purpose of this paper is to serve as a blueprint for other consortia tasked with the advancement of pain related science.

OBJECTIVES: To evaluate whether combining fast acquisitions with deep-learning reconstruction can provide diagnostically useful images and quantitative assessment comparable to standard-of-care acquisitions for lumbar spine magnetic resonance imaging (MRI). METHODS: Eighteen patients were imaged with both standard protocol and fast protocol using reduced signal averages, each protocol including sagittal fat-suppressed T2-weighted, sagittal T1-weighted, and axial T2-weighted 2D fast spin-echo sequences. Fast-acquisition data was additionally reconstructed using vendor-supplied deep-learning reconstruction with three different noise reduction factors. For qualitative analysis, standard images as well as fast images with and without deep-learning reconstruction were graded by three radiologists on five different categories. For quantitative analysis, convolutional neural networks were applied to sagittal T1-weighted images to segment intervertebral discs and vertebral bodies, and disc heights and vertebral body volumes were derived. RESULTS: Based on noninferiority testing on qualitative scores, fast images without deep-learning reconstruction were inferior to standard images for most categories. However, deep-learning reconstruction improved the average scores, and noninferiority was observed over 24 out of 45 comparisons (all with sagittal T2-weighted images while 4/5 comparisons with sagittal T1-weighted and axial T2-weighted images). Interobserver variability increased with 50 and 75% noise reduction factors. Deep-learning reconstructed fast images with 50% and 75% noise reduction factors had comparable disc heights and vertebral body volumes to standard images (r2≥ 0.86 for disc heights and r2≥ 0.98 for vertebral body volumes). CONCLUSIONS: This study demonstrated that deep-learning-reconstructed fast-acquisition images have the potential to provide noninferior image quality and comparable quantitative assessment to standard clinical images.

OBJECTIVES: To evaluate a deep learning model for automated and interpretable classification of central canal stenosis, neural foraminal stenosis, and facet arthropathy from lumbar spine MRI. METHODS: T2-weighted axial MRI studies of the lumbar spine acquired between 2008 and 2019 were retrospectively selected (n = 200) and graded for central canal stenosis, neural foraminal stenosis, and facet arthropathy. Studies were partitioned into patient-level train (n = 150), validation (n = 20), and test (n = 30) splits. V-Net models were first trained to segment the dural sac and the intervertebral disk, and localize facet and foramen using geometric rules. Subsequently, Big Transfer (BiT) models were trained for downstream classification tasks. An interpretable model for central canal stenosis was also trained using a decision tree classifier. Evaluation metrics included linearly weighted Cohens kappa score for multi-grade classification and area under the receiver operator characteristic curve (AUROC) for binarized classification. RESULTS: Segmentation of the dural sac and intervertebral disk achieved Dice scores of 0.93 and 0.94. Localization of foramen and facet achieved intersection over union of 0.72 and 0.83. Multi-class grading of central canal stenosis achieved a kappa score of 0.54. The interpretable decision tree classifier had a kappa score of 0.80. Pairwise agreement between readers (R1, R2), (R1, R3), and (R2, R3) was 0.86, 0.80, and 0.74. Binary classification of neural foraminal stenosis and facet arthropathy achieved AUROCs of 0.92 and 0.93. CONCLUSION: Deep learning systems can be performant as well as interpretable for automated evaluation of lumbar spine MRI including classification of central canal stenosis, neural foraminal stenosis, and facet arthropathy. KEY POINTS: • Interpretable deep-learning systems can be developed for the evaluation of clinical lumbar spine MRI. Multi-grade classification of central canal stenosis with a kappa of 0.80 was comparable to inter-reader agreement scores (0.74, 0.80, 0.86). Binary classification of neural foraminal stenosis and facet arthropathy achieved favorable and accurate AUROCs of 0.92 and 0.93, respectively. • While existing deep-learning systems are opaque, leading to clinical deployment challenges, the proposed system is accurate as well as interpretable, providing valuable information to a radiologist in clinical practice.

BACKGROUND: Although T1ρ and T2 have emerged as early indicators for hip osteoarthritis (OA), there is little information regarding longitudinal changes across the cartilage in the early stages of this disease. PURPOSE: To characterize the variability in 2-year hip cartilage T1ρ and T2 changes and investigate associations between these patterns of change and common indicators of hip OA. STUDY TYPE: Prospective. POPULATION: A total of 25 women (age: 51.9 ± 16.3 years old; BMI: 22.6 ± 2.0 kg/m2 ) and 17 men (age: 55.8 ± 14.9 years old; body mass index (BMI): 24.4 ± 3.8 kg/m2 ) who were healthy or with early-to-moderate hip OA. FIELD STRENGTH/SEQUENCE: A 3 T MRI (GE), 3D combined T1ρ /T2 magnetization-prepared angle-modulated partitioned k-space spoiled gradient echo snapshots. ASSESSMENT: Principal component (PC) analysis of Z-score difference maps of 2-year changes in hip cartilage T1ρ and T2 relaxation times, participant hip disability and osteoarthritis outcome scores (HOOS) and functional tests at 2-year follow-up. STATISTICAL TESTS: Shapiro-Wilk test, unpaired t-tests, Kruskal Wallis tests, Pearson or Spearman (ρ) correlations. Significance was set at P

Radiologists today play a central role in making diagnostic decisions and labeling images for training and benchmarking artificial intelligence (AI) algorithms. A key concern is low inter-reader reliability (IRR) seen between experts when interpreting challenging cases. While team-based decisions are known to outperform individual decisions, inter-personal biases often creep up in group interactions which limit nondominant participants from expressing true opinions. To overcome the dual problems of low consensus and interpersonal bias, we explored a solution modeled on bee swarms. Two separate cohorts, three board-certified radiologists, (cohort 1), and five radiology residents (cohort 2) collaborated on a digital swarm platform in real time and in a blinded fashion, grading meniscal lesions on knee MR exams. These consensus votes were benchmarked against clinical (arthroscopy) and radiological (senior-most radiologist) standards of reference using Cohen's kappa. The IRR of the consensus votes was then compared to the IRR of the majority and most confident votes of the two cohorts. IRR was also calculated for predictions from a meniscal lesion detecting AI algorithm. The attending cohort saw an improvement of 23% in IRR of swarm votes (k = 0.34) over majority vote (k = 0.11). Similar improvement of 23% in IRR (k = 0.25) in 3-resident swarm votes over majority vote (k = 0.02) was observed. The 5-resident swarm had an even higher improvement of 30% in IRR (k = 0.37) over majority vote (k = 0.07). The swarm consensus votes outperformed individual and majority vote decision in both the radiologists and resident cohorts. The attending and resident swarms also outperformed predictions from a state-of-the-art AI algorithm.

OBJECTIVE: To evaluate whether the deep learning (DL) segmentation methods from the six teams that participated in the IWOAI 2019 Knee Cartilage Segmentation Challenge are appropriate for quantifying cartilage loss in longitudinal clinical trials. DESIGN: We included 556 subjects from the Osteoarthritis Initiative study with manually read cartilage volume scores for the baseline and 1-year visits. The teams used their methods originally trained for the IWOAI 2019 challenge to segment the 1130 knee MRIs. These scans were anonymized and the teams were blinded to any subject or visit identifiers. Two teams also submitted updated methods. The resulting 9,040 segmentations are available online.The segmentations included tibial, femoral, and patellar compartments. In post-processing, we extracted medial and lateral tibial compartments and geometrically defined central medial and lateral femoral sub-compartments. The primary study outcome was the sensitivity to measure cartilage loss as defined by the standardized response mean (SRM). RESULTS: For the tibial compartments, several of the DL segmentation methods had SRMs similar to the gold standard manual method. The highest DL SRM was for the lateral tibial compartment at 0.38 (the gold standard had 0.34). For the femoral compartments, the gold standard had higher SRMs than the automatic methods at 0.31/0.30 for medial/lateral compartments. CONCLUSION: The lower SRMs for the DL methods in the femoral compartments at 0.2 were possibly due to the simple sub-compartment extraction done during post-processing. The study demonstrated that state-of-the-art DL segmentation methods may be used in standardized longitudinal single-scanner clinical trials for well-defined cartilage compartments.

To determine the cross-sectional and longitudinal associations of patella alignment with cartilage relaxation and patients self-reported symptoms. Thirty participants with isolated patellofemoral joint (PFJ) degeneration (six males, 53.7 ± 9.3 years) and 24 controls (12 males, 47.6 ± 10.7 years) were included. Magnetic resonance assessment was performed to provide grading of structural abnormalities, cartilage relaxation times, and patella alignment. Self-reported symptoms were assessed using the self-administrated knee injury and osteoarthritis outcome score (KOOS). All participants were examined at baseline and 3 years. Statistical parametric mapping and Pearson partial correlation were used to evaluate the associations between patella alignment with cartilage relaxation times and self-reported symptoms, respectively. The analyses were performed between baseline (cross-sectional) as well as the baseline against 3 years (longitudinal). Results indicated that patella height and patella flexion were associated with T1ρ and T2 relaxation times at baseline (percentages of voxels showing significant correlation [PSV] = 10.1%-24.8%; mean correlations [R] = 0.34-0.36; mean p = 0.015-0.026). Furthermore, greater patella lateral alignment, lateral tilt, and lateral spin were associated with longer T2 times at 3 years (PSV = 11.0%-14.4%, R = 0.39-0.44, p = 0.017-0.028). Last, a higher patella was associated with a lower KOOS at baseline and at 3 years (R = -0.33 to -0.35). The study suggests that patella malalignment is a risk factor for worsening cartilage health, informing clinicians of a better rehabilitation program that targets PFJ degeneration.

Magnetic Resonance Imaging (MRI) offers strong soft tissue contrast but suffers from long acquisition times and requires tedious annotation from radiologists. Traditionally, these challenges have been addressed separately with reconstruction and image analysis algorithms. To see if performance could be improved by treating both as end-to-end, we hosted the K2S challenge, in which challenge participants segmented knee bones and cartilage from 8× undersampled k-space. We curated the 300-patient K2S dataset of multicoil raw k-space and radiologist quality-checked segmentations. 87 teams registered for the challenge and there were 12 submissions, varying in methodologies from serial reconstruction and segmentation to end-to-end networks to another that eschewed a reconstruction algorithm altogether. Four teams produced strong submissions, with the winner having a weighted Dice Similarity Coefficient of 0.910 ± 0.021 across knee bones and cartilage. Interestingly, there was no correlation between reconstruction and segmentation metrics. Further analysis showed the top four submissions were suitable for downstream biomarker analysis, largely preserving cartilage thicknesses and key bone shape features with respect to ground truth. K2S thus showed the value in considering reconstruction and image analysis as end-to-end tasks, as this leaves room for optimization while more realistically reflecting the long-term use case of tools being developed by the MR community.

Background: Gadolinium (Gd)-enhanced Magnetic Resonance Imaging (MRI) is crucial in several applications, including oncology, cardiac imaging, and musculoskeletal inflammatory imaging. One use case is rheumatoid arthritis (RA), a widespread autoimmune condition for which Gd MRI is crucial in imaging synovial joint inflammation, but Gd administration has well-documented safety concerns. As such, algorithms that could synthetically generate post-contrast peripheral joint MR images from non-contrast MR sequences would have immense clinical utility. Moreover, while such algorithms have been investigated for other anatomies, they are largely unexplored for musculoskeletal applications such as RA, and efforts to understand trained models and improve trust in their predictions have been limited in medical imaging. Methods: A dataset of 27 RA patients was used to train algorithms that synthetically generated post-Gd IDEAL wrist coronal T1-weighted scans from pre-contrast scans. UNets and PatchGANs were trained, leveraging an anomaly-weighted L1 loss and global generative adversarial network (GAN) loss for the PatchGAN. Occlusion and uncertainty maps were also generated to understand model performance. Results: UNet synthetic post-contrast images exhibited stronger normalized root mean square error (nRMSE) than PatchGAN in full volumes and the wrist, but PatchGAN outperformed UNet in synovial joints (UNet nRMSEs: volume = 6.29 ± 0.88, wrist = 4.36 ± 0.60, synovial = 26.18 ± 7.45; PatchGAN nRMSEs: volume = 6.72 ± 0.81, wrist = 6.07 ± 1.22, synovial = 23.14 ± 7.37; n = 7). Occlusion maps showed that synovial joints made substantial contributions to PatchGAN and UNet predictions, while uncertainty maps showed that PatchGAN predictions were more confident within those joints. Conclusions: Both pipelines showed promising performance in synthesizing post-contrast images, but PatchGAN performance was stronger and more confident within synovial joints, where an algorithm like this would have maximal clinical utility. Image synthesis approaches are therefore promising for RA and synthetic inflammatory imaging.

BackgroundTo assess the compound effects of BMI and sustained depressive symptoms on changes in knee structure, cartilage composition, and knee pain over 4 years using statistical interaction analyses.MethodsOne thousand eight hundred forty-four individuals from the Osteoarthritis Initiative Database were analyzed at baseline and 4-year follow-up. Individuals were categorized according to their BMI and presence of depressive symptoms (based on the Center for Epidemiological Studies Depression Scale (threshold≥16)) at baseline and 4-year follow-up. 3 T MRI was used to quantify knee cartilage T2 over 4 years, while radiographs were used to assess joint space narrowing (JSN). Mixed effects models examined the effect of BMI-depressive symptoms interactions on outcomes of cartilage T2, JSN, and knee pain over 4-years.ResultsThe BMI-depressive symptoms interaction was significantly associated with knee pain (p

BackgroundModic changes (MCs) are the most prevalent classification system for describing magnetic resonance imaging (MRI) signal intensity changes in the vertebrae. However, there is a growing need for novel quantitative and standardized methods of characterizing these anomalies, particularly for lesions of transitional or mixed nature, due to the lack of conclusive evidence of their associations with low back pain. This retrospective imaging study aims to develop an interpretable deep learning-based detection tool for voxel-wise mapping of MCs.MethodsSeventy-five lumbar spine MRI exams that presented with acute-to-chronic low back pain, radiculopathy, and other symptoms of the lumbar spine were enrolled. The pipeline consists of two deep convolutional neural networks to generate an interpretable voxel-wise Modic map. First, an autoencoder was trained to segment vertebral bodies from T1-weighted sagittal lumbar spine images. Next, two radiologists segmented and labeled MCs from a combined T1- and T2-weighted assessment to serve as ground truth for training a second autoencoder that performs segmentation of MCs. The voxels in the detected regions were then categorized to the appropriate Modic type using a rule-based signal intensity algorithm. Post hoc, three radiologists independently graded a second dataset with the aid of the model predictions in an artificial (AI)-assisted experiment.ResultsThe model successfully identified the presence of changes in 85.7% of samples in the unseen test set with a sensitivity of 0.71 (±0.072), specificity of 0.95 (±0.022), and Cohen's kappa score of 0.63. In the AI-assisted experiment, the agreement between the junior radiologist and the senior neuroradiologist significantly improved from Cohen's kappa score of 0.52 to 0.58 (p

PurposeTo determine the incidence of preoperative shoulder imaging, explore the prevalence of obtaining multiple advanced imaging studies, and identify patient characteristics associated with specific imaging studies before anterior versus posterior shoulder stabilization surgery.MethodsThe PearlDiver database was queried for patients who underwent anterior or posterior shoulder stabilization surgery from 2010 to 2019. The incidence of imaging studies within a year of surgery was collected. Patient characteristics were compared between groups using one-way analysis of variance or χ2 test.ResultsIn total, 10,252 patients underwent anterior shoulder stabilization surgery, and 1,108 patients underwent posterior shoulder stabilization surgery. Imaging use before anterior and posterior shoulder stabilization surgery included plain radiographs (69%, 70%, respectively), magnetic resonance imaging (MRI; 43%, 33%), and computed tomography (CT; 22%, 22%). In total, 1,098 patients (11%) received MRI and CT before anterior stabilization surgery and 85 patients (8%) received MRI and CT before posterior stabilization surgery. Over time, the incidence of obtaining MRI and CT increased before anterior (z = 2.54, P = .011) and posterior (z = 2.36, P = .018) stabilization surgery.ConclusionsThis study highlights the increasing use of multiple imaging studies before shoulder stabilization surgery over recent years, including plain radiographs, MRI, and CT imaging. In total, 45% of anterior shoulder stabilization patients and 41% of posterior shoulder stabilization patients obtained more than 1 imaging study within a year of surgery, with a recent increase in patients obtaining both MR and CT scans preoperatively.Statement of clinical relevanceThe increasing use of multiple preoperative imaging studies observed in this study highlights an opportunity for new imaging technology to streamline and improve the preoperative workup.

BackgroundDiagnostic classification of diffuse gliomas now requires an assessment of molecular features, often including IDH-mutation and 1p19q-codeletion status. Because genetic testing requires an invasive process, an alternative noninvasive approach is attractive, particularly if resection is not recommended. The goal of this study was to evaluate the effects of training strategy and incorporation of biologically relevant images on predicting genetic subtypes with deep learning.MethodsOur dataset consisted of 384 patients with newly diagnosed gliomas who underwent preoperative MRI with standard anatomical and diffusion-weighted imaging, and 147 patients from an external cohort with anatomical imaging. Using tissue samples acquired during surgery, each glioma was classified into IDH-wildtype (IDHwt), IDH-mutant/1p19q-noncodeleted (IDHmut-intact), and IDH-mutant/1p19q-codeleted (IDHmut-codel) subgroups. After optimizing training parameters, top performing convolutional neural network (CNN) classifiers were trained, validated, and tested using combinations of anatomical and diffusion MRI with either a 3-class or tiered structure. Generalization to an external cohort was assessed using anatomical imaging models.ResultsThe best model used a 3-class CNN containing diffusion-weighted imaging as an input, achieving 85.7% (95% CI: [77.1, 100]) overall test accuracy and correctly classifying 95.2%, 88.9%, 60.0% of the IDHwt, IDHmut-intact, and IDHmut-codel tumors. In general, 3-class models outperformed tiered approaches by 13.5%-17.5%, and models that included diffusion-weighted imaging were 5%-8.8% more accurate than those that used only anatomical imaging.ConclusionTraining a classifier to predict both IDH-mutation and 1p19q-codeletion status outperformed a tiered structure that first predicted IDH-mutation, then 1p19q-codeletion. Including apparent diffusion coefficient (ADC), a surrogate marker of cellularity, more accurately captured differences between subgroups.

ObjectiveThe aim of this systematic literature review was to provide a comprehensive and exhaustive overview of the use of machine learning (ML) in the clinical care of osteoarthritis (OA).MethodsA systematic literature review was performed in July 2021 using MEDLINE PubMed with key words and MeSH terms. For each selected article, the number of patients, ML algorithms used, type of data analysed, validation methods and data availability were collected.ResultsFrom 1148 screened articles, 46 were selected and analysed; most were published after 2017. Twelve articles were related to diagnosis, 7 to prediction, 4 to phenotyping, 12 to severity and 11 to progression. The number of patients included ranged from 18 to 5749. Overall, 35% of the articles described the use of deep learning And 74% imaging analyses. A total of 85% of the articles involved knee OA and 15% hip OA. No study investigated hand OA. Most of the studies involved the same cohort, with data from the OA initiative described in 46% of the articles and the MOST and Cohort Hip and Cohort Knee cohorts in 11% and 7%. Data and source codes were described as publicly available respectively in 54% and 22% of the articles. External validation was provided in only 7% of the articles.ConclusionThis review proposes an up-to-date overview of ML approaches used in clinical OA research and will help to enhance its application in this field.

PurposeTo validate the potential of quantifying R2 -R1ρ using one pair of signals with T1ρ preparation and T2 preparation incorporated to magnetization-prepared angle-modulated partitioned k-space spoiled gradient-echo snapshots (MAPSS) acquisition and to find an optimal preparation time (Tprep ) for in vivo knee MRI.MethodsBloch equation simulations were first performed to assess the accuracy of quantifying R2 -R1ρ using T1ρ - and T2 -prepared signals with an equivalent Tprep . For validation of this technique in comparison to the conventional approach that calculates R2 -R1ρ after estimating both T2 and T1ρ , phantom experiments and in vivo validation with five healthy subjects and five osteoarthritis patients were performed at a clinical 3T scanner.ResultsBloch equation simulations demonstrated that the accuracy of this efficient R2 -R1ρ quantification method and the optimal Tprep can be affected by image signal-to-noise ratio (SNR) and tissue relaxation times, but quantification can be closest to the reference with an around 25 ms Tprep for knee cartilage. Phantom experiments demonstrated that the proposed method can depict R2 -R1ρ changes with agarose gel concentration. With in vivo data, significant correlation was observed between cartilage R2 -R1ρ measured from the conventional and the proposed methods, and a Tprep of 25.6 ms provided the most agreement by Bland-Altman analysis. R2 -R1ρ was significantly lower in patients than in healthy subjects for most cartilage compartments.ConclusionAs a potential biomarker to indicate cartilage degeneration, R2 -R1ρ can be efficiently measured using one pair of T1ρ -prepared and T2 -prepared signals with an optimal Tprep considering cartilage relaxation times and image SNR.

BackgroundThe comprehensiveness and maintenance of the American College of Radiology (ACR) Appropriateness Criteria (AC) makes it a unique resource for evidence-based clinical imaging decision support, but it is underutilized by clinicians. To facilitate the use of imaging recommendations, we develop a natural language processing (NLP) search algorithm that automatically matches clinical indications that physicians write into imaging orders to appropriate AC imaging recommendations.MethodsWe apply a hybrid model of semantic similarity from a sent2vec model trained on 223 million scientific sentences, combined with term frequency inverse document frequency features. AC documents are ranked based on their embeddings' cosine distance to query. For model testing, we compiled a dataset of simulated simple and complex indications for each AC document (n = 410) and another with clinical indications from randomly sampled radiology reports (n = 100). We compare our algorithm to a custom google search engine.ResultsOn the simulated indications, our algorithm ranked ground truth documents as top 3 for 98% of simple queries and 85% of complex queries. Similarly, on the randomly sampled radiology report dataset, the algorithm ranked 86% of indications with a single match as top 3. Vague and distracting phrases present in the free-text indications were main sources of errors. Our algorithm provides more relevant results than a custom Google search engine, especially for complex queries.ConclusionsWe have developed and evaluated an NLP algorithm that matches clinical indications to appropriate AC guidelines. This approach can be integrated into imaging ordering systems for automated access to guidelines.

MRI T2 mapping sequences quantitatively assess tissue health and depict early degenerative changes in musculoskeletal (MSK) tissues like cartilage and intervertebral discs (IVDs) but require long acquisition times. In MSK imaging, small features in cartilage and IVDs are crucial for diagnoses and must be preserved when reconstructing accelerated data. To these ends, we propose region of interest-specific postprocessing of accelerated acquisitions: a recurrent UNet deep learning architecture that provides T2 maps in knee cartilage, hip cartilage, and lumbar spine IVDs from accelerated T2-prepared snapshot gradient-echo acquisitions, optimizing for cartilage and IVD performance with a multi-component loss function that most heavily penalizes errors in those regions. Quantification errors in knee and hip cartilage were under 10% and 9% from acceleration factors R = 2 through 10, respectively, with bias for both under 3 ms for most of R = 2 through 12. In IVDs, mean quantification errors were under 12% from R = 2 through 6. A Gray Level Co-Occurrence Matrix-based scheme showed knee and hip pipelines outperformed state-of-the-art models, retaining smooth textures for most R and sharper ones through moderate R. Our methodology yields robust T2 maps while offering new approaches for optimizing and evaluating reconstruction algorithms to facilitate better preservation of small, clinically relevant features.

Clinicians often examine movement patterns to design hip osteoarthritis (OA) interventions, yet traditional biomechanical analyses only report a single timepoint. Multivariate principal component analysis (MFPCA) analyzes the entire waveform (i.e., movement pattern), which clinicians observe to direct treatment. This study investigated hip OA indicators, by (1) employing MFPCA to characterize variance across the hip, knee, and ankle angles in healthy and early-to-moderate hip OA participants; and (2) investigating relationships between these waveform features and hip cartilage health. Bilateral hip magnetic resonance images from 72 participants with Kellgren-Lawrence grades ranging from 0 to 3 were used to calculate mean T 1ρ and T 2 relaxation times in the femoral and acetabular cartilage. MFPCA was performed on lower-limb gait biomechanics and used to identify primary modes of variation, which were related to T 1ρ and T 2 relaxation times. Here, a MFPC = mode of variation = waveform feature. In the femoral cartilage, transverse plane MFPCs 3 and 5 and body mass index (BMI) was related to T 1ρ , while MFPC 2 and BMI were related to T 2 relaxation times. In the acetabular cartilage, sagittal plane MFPC 1 and BMI were related to T 1ρ , while BMI was related to T 2 relaxation times. Greater internal rotation was related to increased T 1ρ and T 2 relaxation times in the femoral cartilage, while the greater extension was related to increased T 1ρ relaxation times in the acetabular cartilage. This study established a data-driven framework to assess relationships between multi-joint biomechanics and quantitative assessments of cartilage health and identified waveform features that could be evaluated in future hip OA intervention studies.

BackgroundAnatomic parameters, such as the critical shoulder angle and acromion index, have emerged as methods to quantify scapular anatomy and may contribute to rotator cuff pathology. The purpose of this paper is to investigate the published literature on influences of scapular morphology on the development of re-tears and patient-reported outcomes following rotator cuff repair.MethodsA systematic review of the Embase and PubMed databases was performed to identify published studies on the potential influence of scapular bony morphology and re-tear rates and patient-reported outcomes after rotator cuff repair. Studies were reviewed by two authors.ResultsA total of 615 unique titles and 49 potentially relevant abstracts were reviewed, with eight published manuscripts identified for inclusion. Two of three papers reported no relationship between these acromion index and rotator cuff re-tear rate, while one paper found an increased re-tear rate. All three studies on critical shoulder angle found a significant association between critical shoulder angle and cuff re-tear rate. There was no clear relationship between any bony morphologic measurement and patient-reported outcomes after rotator cuff repair.ConclusionsRotator cuff re-tear rate appears to be significantly associated with the critical shoulder angle and glenoid inclination, while not clearly associated with acromial morphologic measurements.

Magnetic resonance imaging (MRI) is commonly used to evaluate the morphology of the knee in athletes with high-knee impact; however, complex repeated loading of the joint can lead to biochemical and structural degeneration that occurs before visible morphological changes. In this study, we utilized multiparametric quantitative MRI to compare morphology and composition of articular cartilage and subchondral bone shape between young athletes with high-knee impact (basketball players; n = 40) and non-knee impact (swimmers; n = 25). We implemented voxel-based relaxometry to register all cases to a single reference space and performed a localized compositional analysis of T 1ρ - and T 2 -relaxation times on a voxel-by-voxel basis. Additionally, statistical shape modeling was employed to extract differences in subchondral bone shape between the two groups. Evaluation of cartilage composition demonstrated a significant prolongation of relaxation times in the medial femoral and tibial compartments and in the posterolateral femur of basketball players in comparison to relaxation times in the same cartilage compartments of swimmers. The compositional analysis also showed depth-dependent differences with prolongation of the superficial layer in basketball players. For subchondral bone shape, three total modes were found to be significantly different between groups and related to the relative sizes of the tibial plateaus, intercondylar eminences, and the curvature and concavity of the patellar lateral facet. In summary, this study identified several characteristics associated with a high-knee impact which may expand our understanding of local degenerative patterns in this population.

Osteoarthritis (OA) develops through heterogenous pathophysiologic pathways. As a result, no regulatory agency approved disease modifying OA drugs are available to date. Stratifying knees into MRI-based morphological phenotypes may provide insight into predicting future OA incidence, leading to improved inclusion criteria and efficacy of therapeutics. We trained convolutional neural networks to classify bone, meniscus/cartilage, inflammatory, and hypertrophy phenotypes in knee MRIs from participants in the Osteoarthritis Initiative (n = 4791). We investigated cross-sectional association between baseline morphological phenotypes and baseline structural OA (Kellgren Lawrence grade > 1) and symptomatic OA. Among participants without baseline OA, we evaluated association of baseline phenotypes with 48-month incidence of structural OA and symptomatic OA. The area under the curve of bone, meniscus/cartilage, inflammatory, and hypertrophy phenotype neural network classifiers was 0.89 ± 0.01, 0.93 ± 0.03, 0.96 ± 0.02, and 0.93 ± 0.02, respectively (mean ± standard deviation). Among those with no baseline OA, bone phenotype (OR: 2.99 (95%CI: 1.59-5.62)) and hypertrophy phenotype (OR: 5.80 (95%CI: 1.82-18.5)) each respectively increased odds of developing incident structural OA and symptomatic OA at 48 months. All phenotypes except meniscus/cartilage increased odds of undergoing total knee replacement within 96 months. Artificial intelligence can rapidly stratify knees into structural phenotypes associated with incident OA and total knee replacement, which may aid in stratifying patients for clinical trials of targeted therapeutics.

[This corrects the article DOI: 10.1148/ryai.2021200165.].

PurposeTo organize a multi-institute knee MRI segmentation challenge for characterizing the semantic and clinical efficacy of automatic segmentation methods relevant for monitoring osteoarthritis progression.Materials and methodsA dataset partition consisting of three-dimensional knee MRI from 88 retrospective patients at two time points (baseline and 1-year follow-up) with ground truth articular (femoral, tibial, and patellar) cartilage and meniscus segmentations was standardized. Challenge submissions and a majority-vote ensemble were evaluated against ground truth segmentations using Dice score, average symmetric surface distance, volumetric overlap error, and coefficient of variation on a holdout test set. Similarities in automated segmentations were measured using pairwise Dice coefficient correlations. Articular cartilage thickness was computed longitudinally and with scans. Correlation between thickness error and segmentation metrics was measured using the Pearson correlation coefficient. Two empirical upper bounds for ensemble performance were computed using combinations of model outputs that consolidated true positives and true negatives.ResultsSix teams (T 1-T 6) submitted entries for the challenge. No differences were observed across any segmentation metrics for any tissues (P = .99) among the four top-performing networks (T 2, T 3, T 4, T 6). Dice coefficient correlations between network pairs were high (> 0.85). Per-scan thickness errors were negligible among networks T 1-T 4 (P = .99), and longitudinal changes showed minimal bias (< 0.03 mm). Low correlations (ρ < 0.41) were observed between segmentation metrics and thickness error. The majority-vote ensemble was comparable to top-performing networks (P = .99). Empirical upper-bound performances were similar for both combinations (P = .99).ConclusionDiverse networks learned to segment the knee similarly, where high segmentation accuracy did not correlate with cartilage thickness accuracy and voting ensembles did not exceed individual network performance.See also the commentary by Elhalawani and Mak in this issue.Keywords: Cartilage, Knee, MR-Imaging, Segmentation © RSNA, 2020Supplemental material is available for this article.

ObjectiveThe aim of this study was to investigate the associations between weight cycling and knee joint degeneration in individuals with overweight or obesity with different patterns of weight change over 4 years.MethodsA total of 2,271 individuals from the Osteoarthritis Initiative database were assessed (case-control study). Linear regression models using annual BMI measurements over 4 years were used to classify participants as weight cyclers or noncyclers. 3-T magnetic resonance imaging was used to quantify knee cartilage transverse relaxation time (T2) and cartilage thickness annually over 4 years in all subjects. Whole-Organ Magnetic Resonance Imaging Scores (WORMS) were obtained for cartilage, meniscus, and bone-marrow abnormalities in 958 subjects at baseline and at the 4-year follow-up. The longitudinal differences in cartilage T2 and thickness between weight cyclers and noncyclers were assessed using general estimating equations, whereas the differences in WORMS outcomes were compared using general linear models.ResultsNo significant differences in the rate of change of cartilage thickness or T2 were found between weight cyclers and noncyclers. However, increases in maximum cartilage WORMS (P = 0.0025) and bone-marrow abnormalities (P = 0.04) were significantly greater in weight cyclers than in noncyclers.ConclusionsAlthough participants' intent for weight cycling in this study was unknown, weight cyclers had significantly greater increases in cartilage and bone-marrow abnormalities over 4 years than noncyclers, independent of weight gain and loss.

PurposeTo test the hypothesis that artificial intelligence (AI) techniques can aid in identifying and assessing lesion severity in the cartilage, bone marrow, meniscus, and anterior cruciate ligament (ACL) in the knee, improving overall MRI interreader agreement.Materials and methodsThis retrospective study was conducted on 1435 knee MRI studies (n = 294 patients; mean age, 43 years ± 15 [standard deviation]; 153 women) collected within three previous studies (from 2011 to 2014). All MRI studies were acquired using high-spatial-resolution three-dimensional fast-spin-echo CUBE sequence. Three-dimensional convolutional neural networks were developed to detect the regions of interest within MRI studies and grade abnormalities of the cartilage, bone marrow, menisci, and ACL. Evaluation included sensitivity, specificity, and Cohen linear-weighted ĸ. The impact of AI-aided grading in intergrader agreement was assessed on an external dataset.ResultsBinary lesion sensitivity reported for all tissues was between 70% and 88%. Specificity ranged from 85% to 89%. The area under the receiver operating characteristic curve for all tissues ranged from 0.83 to 0.93. Deep learning-assisted intergrader Cohen ĸ agreement significantly improved in 10 of 16 comparisons among two attending physicians and two trainees for all tissues.ConclusionThe three-dimensional convolutional neural network had high sensitivity, specificity, and accuracy for knee-lesion-severity scoring and also increased intergrader agreement when used on an external dataset.Supplemental material is available for this article. Keywords: Bone Marrow, Cartilage, Computer Aided Diagnosis (CAD), Computer Applications-3D, Computer Applications-Detection/Diagnosis, Knee, Ligaments, MR-Imaging, Neural Networks, Observer Performance, Segmentation, Statistics © RSNA, 2021See also the commentary by Li and Chang in this issue.: An earlier incorrect version of this article appeared online. This article was corrected on April 16, 2021.

To explore bone shape features that are associated with patellofemoral joint (PFJ) osteoarthritic features. Thirty subjects with PFJ degeneration (six males, 53.2 ± 9.8 years) and 23 controls (12 males, 48.1 ± 10.6 years) were included. Magnetic resonance (MR) assessment was performed to provide bone segmentation, morpholgocial grading, and cartilage relaxation times. In addition, subject self-reported symptoms were reported. Logistic regressions were used to identify the shape features that were associated with the presence and worsening of PFJ morphological lesions over 3 years, and worsening of self-reported symptoms. Statistical parametric mapping was used to evaluate the associations between shape features and cartilage relaxation times at 3 years. Results indicated that subjects with PFJ degeneration exhibited a trochlea with longer lateral condyle and shallower trochlear groove (adjusted odds ratio [OR] = 0.30; 95% confidence interval [CI]: 0.10, 0.86; P = .025). Subjects with worsening of PFJ degeneration exhibited a patella with equally distributed facets (adjusted OR = 3.14; 95% CI: 1.05, 9.37; P = .040) and lateral bump (adjusted OR = 0.14; 95% CI: 0.02, 0.83; P = .030). No shape features were associated with worsening of self-reported symptoms. Elevated T1ρ and T2 times at 3 years were associated with a patella with a lateral hook, equally distributed facets, round and thick as well as a trochlea larger in size (R = 0.38~0.46, P = .015~.025). The study demonstrated the ability of 3D statistical shape modeling to quantify patella and trochlear bone shape features that are associated with the presence and progression of PFJ osteoarthritic features.

Knee pain is the most common and debilitating symptom of knee osteoarthritis (OA). While there is a perceived association between OA imaging biomarkers and pain, there are weak or conflicting findings for this relationship. This study uses Deep Learning (DL) models to elucidate associations between bone shape, cartilage thickness and T2 relaxation times extracted from Magnetic Resonance Images (MRI) and chronic knee pain. Class Activation Maps (Grad-CAM) applied on the trained chronic pain DL models are used to evaluate the locations of features associated with presence and absence of pain. For the cartilage thickness biomarker, the presence of features sensitive for pain presence were generally located in the medial side, while the features specific for pain absence were generally located in the anterior lateral side. This suggests that the association of cartilage thickness and pain varies, requiring a more personalized averaging strategy. We propose a novel DL-guided definition for cartilage thickness spatial averaging based on Grad-CAM weights. We showed a significant improvement modeling chronic knee pain with the inclusion of the novel biomarker definition: likelihood ratio test p-values of 7.01 × 10-33 and 1.93 × 10-14 for DL-guided cartilage thickness averaging for the femur and tibia, respectively, compared to the cartilage thickness compartment averaging.

BackgroundBone-cartilage interactions have been implicated in causing osteoarthritis (OA).PurposeTo use [18 F]-NaF PET-MRI to 1) develop automatic image processing code in MatLab to create a model of bone-cartilage interactions and 2) find associations of bone-cartilage interactions with known manifestations of OA.Study typeProspective study aimed to evaluate a data analysis method.PopulationTwenty-nine patients with knee pain or joint stiffness.Field strength/sequence3T MRI (GE), 3D CUBE FSE, 3D combined T1 ρ/T2 MAPSS, [18F]-sodium fluoride, SIGNA TOF (OSEM).AssessmentCorrelation between MRI (cartilage) and PET (bone) quantitative parameters, bone-cartilage interactions model described by modes of variation as derived by principal component analysis (PCA), WORMS scoring on cartilage lesions, bone marrow abnormalities, subchondral cysts.Statistical testsLinear regression, Pearson correlation.ResultsMode 1 was a positive predictor of the bone abnormality score (P = 0.0003, P = 0.001, P = 0.0007) and the cartilage lesion score (P = 0.03, P = 0.01, P = 0.02) in the femur, tibia, and patella, respectively. For the cartilage lesion scores, mode 5 was the most important positive predictor in the femur (P = 3.9E-06), and mode 2 were predictors, significant negative predictor in the tibia (P = 0.007). In the patella, mode 1 was a significant positive predictor of the bone abnormality score (P = 0.0007).Data conclusionBy successfully building an automatic code to create a bone-cartilage interface, we were able to observe dynamic relationships between biochemical changes in the cartilage accompanied with bone remodeling, extended to the whole knee joint instead of simple colocalized observations, shedding light on the interactions that occur between bone and cartilage in OA. Evidence Level: 3 Technical Efficacy: Stage 3 J. Magn. Reson. Imaging 2020;52:1462-1474.

Osteoarthritis (OA) of the knee is a major source of disability that has no known treatment or cure. Morphological and compositional MRI is commonly used for assessing the bone and soft tissues in the knee to enhance the understanding of OA pathophysiology. However, it is challenging to extend these imaging methods and their subsequent analysis techniques to study large population cohorts due to slow and inefficient imaging acquisition and postprocessing tools. This can create a bottleneck in assessing early OA changes and evaluating the responses of novel therapeutics. The purpose of this review article is to highlight recent developments in tools for enhancing the efficiency of knee MRI methods useful to study OA. Advances in efficient MRI data acquisition and reconstruction tools for morphological and compositional imaging, efficient automated image analysis tools, and hardware improvements to further drive efficient imaging are discussed in this review. For each topic, we discuss the current challenges as well as potential future opportunities to alleviate these challenges. LEVEL OF EVIDENCE: 5 TECHNICAL EFFICACY STAGE: 3.