Detection Of Bone Marrow Edema Research Articles

Dual-energy CT for occult pelvic fractures: An audit and rocanalysis.

There is an increasing incidence of hip and pelvic fractures with an ageing population. Accurate and timely diagnosis is important in the emergency setting. While magnetic resonance imaging (MRI) is the gold standard, it is a limited resource. Dual energy CT (DECT) is comparable to MRI in detection of bone marrow oedema. Our hospital was the first centre in our country to introduce DECT for occult pelvic fractures. We aimed to describe its utility in occult pelvic fractures since commencement. Retrospective study of consecutive pelvic bone CT (conventional or DECT) performed to look for an occult fracture over a 10-month period. Sensitivity and specificity calculated based on clinical and imaging follow-up. ROC study performed where three observers visually interpreted pelvic radiographs, conventional CT and DECT and scored their confidence for an acute fracture from 1 to 5. The null hypothesis was that DECT would not improve observer performance compared with conventional CT. DECT studies were performed on 178 patients of whom 84 (47%) had acute fractures. Sensitivity on audit was 99% and specificity was 100%. ROC analysis showed that, for all observers, the area under curve increased from radiograph to conventional CT to DECT. The difference between conventional CT and DECT was statistically significant for all observers where metal implants were not present. DECT improves accuracy compared to conventional CT in the diagnosis of occult pelvic fractures and should be used for this indication when available.

Assessment of bone marrow edema on dual-energy CT scans in people with diabetes mellitus and suspected Charcot neuro-osteoarthropathy.

This study aimed to quantitatively assess the diagnostic value of bone marrow edema (BME) detection on virtual non-calcium (VNCa) images calculated from dual-energy CT (DECT) in people with diabetes mellitus and suspected Charcot neuro-osteoarthropathy (CN). People with diabetes mellitus and suspected CN who underwent DECT of the feet (80kVp/Sn150kVp) were included retrospectively. Two blinded observers independently measured CT values on VNCa images using circular regions of interest in five locations in the midfoot (cuneiforms, cuboid and navicular) and the calcaneus of the contralateral or (if one foot was available) the ipsilateral foot. Two clinical groups were formed, one with active CN and one without active CN (no-CN), based on the clinical diagnosis. Thirty-two people with diabetes mellitus and suspected CN were included. Eleven had clinically active CN. The mean CT value in the midfoot was significantly higher in the CN group (-55.6 ± 18.7 HU) compared to the no-CN group (-94.4 ± 23.5 HU; p < 0.001). In the CN group, the difference in CT value between the midfoot and calcaneus was statistically significant (p = 0.003); this was not the case in the no-CN group (p = 0.357). The overall observer agreement was good for the midfoot (ICC = 0.804) and moderate for the calcaneus (ICC = 0.712). Sensitivity was 100.0% and specificity was 71.4% using a cutoff value of -87.6 HU. The detection of BME on VNCa images has a potential value in people with diabetes mellitus and suspected active CN.

W-DRAG: A joint framework of WGAN with data random augmentation optimized for generative networks for bone marrow edema detection in dual energy CT

Dual-energy computed tomography (CT) is an excellent substitute for identifying bone marrow edema in magnetic resonance imaging. However, it is rarely used in practice owing to its low contrast. To overcome this problem, we constructed a framework based on deep learning techniques to screen for diseases using axial bone images and to identify the local positions of bone lesions. To address the limited availability of labeled samples, we developed a new generative adversarial network (GAN) that extends expressions beyond conventional augmentation (CA) methods based on geometric transformations. We theoretically and experimentally determined that combining the concepts of data augmentation optimized for GAN training (DAG) and Wasserstein GAN yields a considerably stable generation of synthetic images and effectively aligns their distribution with that of real images, thereby achieving a high degree of similarity. The classification model was trained using real and synthetic samples. Consequently, the GAN technique used in the diagnostic test had an improved F1 score of approximately 7.8% compared with CA. The final F1 score was 80.24%, and the recall and precision were 84.3% and 88.7%, respectively. The results obtained using the augmented samples outperformed those obtained using pure real samples without augmentation. In addition, we adopted explainable AI techniques that leverage a class activation map (CAM) and principal component analysis to facilitate visual analysis of the network’s results. The framework was designed to suggest an attention map and scattering plot to visually explain the disease predictions of the network.

The Clinical Utility of Lower Extremity Dual-Energy CT Angiography in the Detection of Bone Marrow Edema in Diabetic Patients with Peripheral Artery Disease.

(1) Background: Type 2 diabetes is a major cause of incidences and the progression of peripheral artery disease (PAD). Bone marrow edema (BME) is an important finding suggestive of underlying bone inflammation in non-traumatic diabetic patients with PAD. Our aim was to evaluate the presence, severity, and clinical implications of BME detected by virtual non-calcium application (VNCa) of dual-energy CT angiography (DE-CTA). (2) Methods: A consecutive series of 76 diabetic patients (55 men; mean age 71.6 ± 11.2 yrs) submitted to lower limb DE-CTA for PAD evaluation and revascularization planning, which were retrospectively analyzed. VNCa images were independently and blindly revised for the presence, location, and severity of BME by two radiologists with 10 years of experience. BME and non-BME groups were evaluated in terms of PAD clinical severity and 6-month secondary major amputation rate. (3) Results: BME was present in 17 (22%) cases, while 59 (78%) patients were non-BME. The BME group showed a significantly higher incidence of major amputation (p < 0.001) and a significantly higher number of patients with advanced clinical stages of PAD compared to the non-BME group (p = 0.024). (4) Conclusions: Lower limb DE-CTA with VNCa application is a useful tool in the detection of BME in diabetic patients with PAD, simultaneously enabling the evaluation of the severity and location of the arterial disease for revascularization planning. BME presence could be a marker of clinically severe PAD and a possible risk factor for revascularization failure.

The detecting value of virtual non-calcium technique of dual-energy CT for bone marrow edema around nontraumatic osteonecrosis of the femoral head

Objective: To evaluate the value of virtual non-calcium (VNCa) technique of dual-energy CT (DECT) for detecting bone marrow edema (BME) around nontraumatic osteonecrosis of the femoral head (ONFH) using MRI as reference standard. Methods: Nontraumatic ONFH patients were prospectively studied in the Fourth Medical Center of Chinese PLA General Hospital from October 2022 to May 2023, and their MRI and DECT images were analyzed. The diagnostic efficiency of the subjective assessment of BME around ONFH by two radiologists in VNCa color-coded images were calculated using the MRI results as the reference standard. The BME ranges were compared between VNCa images and MRI. Traditional CT values and VNCa CT values were compared between normal bone marrow and BME. The receiver operator characteristic (ROC) curve was established based on the statistically different CT values, and the area under the curve (AUC) was calculated to find the threshold to distinguish normal bone marrow from BME and evaluate the diagnostic efficacy. Results: Thirty patients with ONFH were included, including 24 males and 6 females, aged (39±12) years. There were 18 bilateral hips and 12 unilateral hips, with a total of 48 hips, 34 hips of which showed BME on MRI. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy of subjective detection of BME on VNCa color coded maps by two physicians were 97.1% (33/34) and 97.1% (33/34), 92.9% (13/14) and 71.4% (10/14), 97.1% (33/34) and 89.2% (33/37), 92.9% (13/14) and 90.9% (10/11), 95.8% (46/48) and 89.6% (43/48), respectively, with no statistical difference (all P>0.05).There was no statistical difference between VNCa color-coded images and MRI in the BME range (P=1.160). The traditional CT values measured by the two radiologists were in good agreement with VNCa CT values, with intraclass correlation coefficient (ICC) of 0.948 (95%CI: 0.908-0.971) and 0.982 (95%CI: 0.969-0.990), respectively. The traditional CT value of normal bone marrow was (400.7±82.8) HU, and that of BME was (443.7±65.7) HU, with no statistical difference (P=0.062). The VNCa CT value of normal bone marrow was (-103.1±27.8) HU, and that of BME was (-32.9±25.7) HU, with statistical difference (P<0.001). The AUC of distinguishing normal bone marrow from BME based on VNCa CT value was 0.958 (95%CI: 0.857-0.995). The best cut-off value was -74.5 HU, and when the VNCa CT value was higher than -74.5 HU, the sensitivity, specificity, PPV, NPV and accuracy of diagnosing BME were 97.1%, 92.9%, 97.1%, 92.9% and 95.8 %, respectively. Conclusion: The VNCa technique of DECT has high efficiency in detecting BME around ONFH, and can accurately demonstrate the range of BME.

Feasibility of bone marrow edema detection using dual-energy cone-beam computed tomography.

Dual-energy (DE) detection of bone marrow edema (BME) would be a valuable new diagnostic capability for the emerging orthopedic cone-beam computed tomography (CBCT) systems. However, this imaging task is inherently challenging because of the narrow energy separation between water (edematous fluid) and fat (health yellow marrow), requiring precise artifact correction and dedicated material decomposition approaches. We investigate the feasibility of BME assessment using kV-switching DE CBCT with a comprehensive CBCT artifact correction framework and a two-stage projection- and image-domain three-material decomposition algorithm. DE CBCT projections of quantitative BME phantoms (water containers 100-165mm in size with inserts presenting various degrees of edema) and an animal cadaver model of BME were acquired on a CBCT test bench emulating the standard wrist imaging configuration of a Multitom Rax twin robotic x-ray system. The slow kV-switching scan protocol involved a 60kV low energy (LE) beam and a 120kV high energy (HE) beam switched every 0.5° over a 200° angular span. The DE CBCT data preprocessing and artifact correction framework consisted of (i) projection interpolation onto matched LE and HE projections views, (ii) lag and glare deconvolutions, and (iii) efficient Monte Carlo (MC)-based scatter correction. Virtual non-calcium (VNCa) images for BME detection were then generated by projection-domain decomposition into an Aluminium (Al) and polyethylene basis set (to remove beam hardening) followed by three-material image-domain decomposition into water, Ca, and fat. Feasibility of BME detection was quantified in terms of VNCa image contrast and receiver operating characteristic (ROC) curves. Robustness to object size, position in the field of view (FOV) and beam collimation (varied 20-160mm) was investigated. The MC-based scatter correction delivered>69% reduction of cupping artifacts for moderate to wide collimations (>80mm beam width), which was essential to achieve accurate DE material decomposition. In a forearm-sized object, a 20% increase in water concentration (edema) of a trabecular bone-mimicking mixture presented as ∼15 HU VNCa contrast using 80-160mm beam collimations. The variability with respect to object position in the FOV was modest (<15% coefficient of variation). The areas under the ROC curve were>0.9. A femur-sized object presented a somewhat more challenging task, resulting in increased sensitivity to object positioning at 160mm collimation. In animal cadaver specimens, areas of VNCa enhancement consistent with BME were observed in DE CBCT images in regions of MRI-confirmed edema. Our results indicate that the proposed artifact correction and material decomposition pipeline can overcome the challenges of scatter and limited spectral separation to achieve relatively accurate and sensitive BME detection in DE CBCT. This study provides an important baseline for clinical translation of musculoskeletal DE CBCT to quantitative, point-of-care bone health assessment.

Dual-energy CT applications in musculoskeletal disorders.

Dual-energy CT (DECT) is an exciting application in CT technology conferring many advantages over conventional single-energy CT at no additional with comparable radiation dose to the patient. Various emerging and increasingly established clinical DECT applications in musculoskeletal (MSK) imaging such as bone marrow oedema detection, metal artefact reduction, monosodium urate analysis, and collagen analysis for ligamentous, meniscal, and disc injuries are made possible through its advanced DECT post-processing capabilities. These provide superior information on tissue composition, artefact reduction and image optimization. Newer DECT applications to evaluate fat fraction for sarcopenia, Rho/Z application for soft tissue calcification differentiation, 3D rendering, and AI integration are being assessed for future use. In this article, we will discuss the established and developing applications of DECT in the setting of MSK radiology as well as the basic principles of DECT which facilitate them.

Deep learning-based automated high-accuracy location and identification of fresh vertebral compression fractures from spinal radiographs: a multicenter cohort study.

Digital radiography (DR) is a common and widely available examination. However, spinal DR cannot detect bone marrow edema, therefore, determining vertebral compression fractures (VCFs), especially fresh VCFs, remains challenging for clinicians. We trained, validated, and externally tested the deep residual network (DRN) model that automated the detection and identification of fresh VCFs from spinal DR images. A total of 1,747 participants from five institutions were enrolled in this study and divided into the training cohort, validation cohort and external test cohorts (YHDH and BMUH cohorts). We evaluated the performance of DRN model based on the area under the receiver operating characteristic curve (AUC), feature attention maps, sensitivity, specificity, and accuracy. We compared it with five other deep learning models and validated and tested the model internally and externally and explored whether it remains highly accurate for an external test cohort. In addition, the influence of old VCFs on the performance of the DRN model was assessed. The AUC was 0.99, 0.89, and 0.88 in the validation, YHDH, and BMUH cohorts, respectively, for the DRN model for detecting and discriminating fresh VCFs. The accuracies were 81.45% and 72.90%, sensitivities were 84.75% and 91.43%, and specificities were 80.25% and 63.89% in the YHDH and BMUH cohorts, respectively. The DRN model generated correct activation on the fresh VCFs and accurate peak responses on the area of the target vertebral body parts and demonstrated better feature representation learning and classification performance. The AUC was 0.90 (95% confidence interval [CI] 0.84-0.95) and 0.84 (95% CI 0.72-0.93) in the non-old VCFs and old VCFs groups, respectively, in the YHDH cohort (p = 0.067). The AUC was 0.89 (95% CI 0.84-0.94) and 0.85 (95% CI 0.72-0.95) in the non-old VCFs and old VCFs groups, respectively, in the BMUH cohort (p = 0.051). In present study, we developed the DRN model for automated diagnosis and identification of fresh VCFs from spinal DR images. The DRN model can provide interpretable attention maps to support the excellent prediction results, which is the key that most clinicians care about when using the model to assist decision-making.

To study magnetic resonance imaging findings and inflammatory markers in inflammatory sacroiliitis

Objectives: The objectives of the study are to determine magnetic resonance imaging (MRI) findings in inflammatory sacroiliitis and assign scores and grades to it and to determine and correlate erythrocyte sedimentation rate, C-reactive protein (CRP), and human leukocyte antigen-B27 (HLA-B27) in various grades of sacroiliitis. Material and Methods: An observational cross-sectional study was conducted on 30 patients who clinically presented with features of sacroiliitis and underwent an MRI of sacroiliac joint (SIJ). Various inflammatory and structural findings on MRI were used to do Spondyloarthritis Research Consortium of Canada scoring and grading. Then inflammatory markers including erythrocyte sedimentation rate, CRP, and HLA-B27 were studied in various grades of sacroiliitis. Results: Inflammatory sacroiliitis affects commonly the age group of 21–40 years. Periarticular edema was the most common finding seen with the iliac aspect more commonly involved. The majority of the subjects were graded moderate (50%). Values of erythrocyte sedimentation rate and CRP levels were raised whereas HLA-B27 was positive in 9 patients (30%) of inflammatory sacroiliitis. Conclusion: Inflammatory sacroiliitis presents with a chief complaint of low back ache. MRI helps to grade it into mild, moderate, and severe. STIR is the most sensitive sequence for the detection of bone marrow edema with bilateral symmetrical involvement but the iliac bone of SIJ is more involved than the sacral side. Contrast-enhanced sequences and diffusion images add no significant statistical role in the diagnosis of bone marrow edema. Inflammatory laboratory parameters were increased in higher grades of sacroiliitis. HLA-B27, although not specific to inflammatory sacroiliitis, increases in higher grades of sacroiliitis.

Dual-energy CT Virtual Non-Calcium Image is Beneficial for the Detection of Non-displaced Knee Fractures.

Early and accurate diagnosis is vital for avoiding the development of nondisplaced fractures to displaced fractures. Dual-energy CT (Computed Tomography) can detect bone marrow edema (BME), which may help to detect non-displaced fractures. To evaluate the value of DECT (Dual-Energy Computed Tomography) VNCa (Virtual noncalcium) images for improving diagnostic performance and confidence in acute non-displaced knee fractures. 125 patients with clinical suspicion of knee fractures underwent both DECT and MR. Conventional linear-blended CT and VNCa images were obtained from DECT. First, five readers with varying levels of experience evaluated the presence of fractures on conventional linear-blended CT and graded their diagnostic confidence on a scale of 1 to 10. Then BME with VNCa images was evaluated and compared with MR. Finally, the VNCa images combined with conventional linear-blended CT images were used to reassess the presence of fractures and diagnostic confidence. Diagnostic performance and matched pair analyses were performed. 20 non-displaced knee fractures were detected. The consistency test of VNCa images and MR by five radiologists showed Kappa values are 0.76, 0.79, 0.81,0.85,and 0.90,respectively. The diagnostic performance of all readers was improved when using VNCa images combined with conventional linear-blended CT compared with that with conventional linear-blended CT alone. Diagnostic confidence was improved with combined conventional linear-blended CT and VNCa images (median score:8,8,9,9, and 10, respectively) compared with conventional linear-blended CT alone (median score:7,7,8,9, and 9). DECT VNCa images could improve the radiologists' diagnostic performance and confidence with varying levels of experience in the detection of non-displaced knee fractures.

Detecting Bone Marrow Edema of the Extremities on Spectral Computed Tomography Using a Three-Material Decomposition.

Detecting bone marrow edema (BME) as a sign of acute fractures is challenging on conventional computed tomography (CT). This study evaluated the diagnostic performance of a three-material decomposition (TMD) approach for detecting traumatic BME of the extremities on spectral computed tomography (SCT). This retrospective diagnostic study included 81 bone compartments with and 80 without BME. A TMD application to visualize BME was developed in collaboration with Philips Healthcare. The following bone compartments were included: distal radius, proximal femur, proximal tibia, distal tibia and fibula, and long bone diaphysis. Two blinded radiologists reviewed each case independently in random order for the presence or absence of BME. The interrater reliability was 0.84 (p < 0.001). The different bone compartments showed sensitivities of 86.7% to 93.8%, specificities of 84.2% to 94.1%, positive predictive values of 82.4% to 94.7%, negative predictive values of 87.5% to 93.3%, and area under the curve (AUC) values of 85.7% to 93.1%. The distal radius showed the highest sensitivity and the proximal femur showed the lowest sensitivity, while the proximal femur presented the highest specificity and the distal tibia presented the lowest specificity. Our TMD approach provides high diagnostic performance for detecting BME of the extremities. Therefore, this approach could be used routinely in the emergency setting.

Dual-Energy CT Arthrography: Advanced Muscolo-Skelatal Applications in Clinical Practice.

This paper provides a comprehensive overview of the potential applications of dual-energy CT (DECT) in improving image quality and the diagnostic capabilities of CT arthrography (CTA) in clinical practice. The paper covers the use of virtual non-contrast (VNC) images, in which the injected contrast medium is subtracted from the articular cavity in order to better analyze 2D and 3D images of the bone. Moreover, virtual monoenergetic imaging (VMI) applications and their potential use for the reduction of metal artifacts and improving image contrast are reviewed. The role of virtual non-calcium (VNCa) in detecting bone marrow edema surrounding the imaged joint will be discussed. Furthermore, the role of iodine maps in enhancing the contrast between soft tissues, optimizing the visualization of contrast material, and distinguishing contrast material from calcifications is described. Finally, a case series including different joints is provided to underline the additional advantages of high-spatial-resolution dual-energy CT reconstructed images.

Radiomics for the Detection of Active Sacroiliitis Using MR Imaging.

Detecting active inflammatory sacroiliitis at an early stage is vital for prescribing medications that can modulate disease progression and significantly delay or prevent debilitating forms of axial spondyloarthropathy. Conventional radiography and computed tomography offer limited sensitivity in detecting acute inflammatory findings as these methods primarily identify chronic structural lesions. Conversely, Magnetic Resonance Imaging (MRI) is the preferred technique for detecting bone marrow edema, although it is a complex process requiring extensive expertise. Additionally, ascertaining the origin of lesions can be challenging, even for experienced medical professionals. Machine learning (ML) has showcased its proficiency in various fields by uncovering patterns that are not easily perceived from multi-dimensional datasets derived from medical imaging. The aim of this study is to develop a radiomic signature to aid clinicians in diagnosing active sacroiliitis. A total of 354 sacroiliac joints were segmented from axial fluid-sensitive MRI images, and their radiomic features were extracted. After selecting the most informative features, a number of ML algorithms were utilized to identify the optimal method for detecting active sacroiliitis, leading to the selection of an Extreme Gradient Boosting (XGBoost) model that accomplished an Area Under the Receiver-Operating Characteristic curve (AUC-ROC) of 0.71, thus further showcasing the potential of radiomics in the field.

Performance of Fully Automated Algorithm Detecting Bone Marrow Edema in Sacroiliac Joints.

This study evaluates the performance of a fully automated algorithm to detect active inflammation in the form of bone marrow edema (BME) in iliac and sacral bones, depending on the quality of the coronal oblique plane in patients with axial spondyloarthritis (axSpA). The results were assessed based on the technical correctness of MRI examination of the sacroiliac joints (SIJs). A total of 173 patients with suspected axSpA were included in the study. In order to verify the correctness of the MRI, a deviation angle was measured on the slice acquired in the sagittal plane in the T2-weighted sequence. This angle was located between the line drawn between the posterior edges of S1 and S2 vertebrae and the line that marks the actual plane in which the slices were acquired in T1 and STIR sequences. All examinations were divided into quartiles according to the deviation angle measured in degrees as follows: 1st group [0; 2.2], 2nd group (2.2; 5.7], 3rd group (5.7; 10] and 4th group (10; 29.2]. Segmentations of the sacral and iliac bones were acquired manually and automatically using the fully automated algorithm on the T1 sequence. The Dice coefficient for automated bone segmentations with respect to reference manual segmentations was 0.9820 (95% CI [0.9804, 0.9835]). Examinations of BME lesions were assessed using the SPARCC scale (in 68 cases SPARCC > 0). Manual and automatic segmentations of the lesions were performed on STIR sequences and compared. The sensitivity of detection of BME ranged from 0.58 (group 1) to 0.83 (group 2) versus 0.76 (total), while the specificity was equal to 0.97 in each group. The study indicates that the performance of the algorithm is satisfactory regardless of the deviation angle.

Quantitative Dual-Energy Imaging of Bone Marrow Edema Using Multisource Cone-Beam CT with Model-Based Decomposition.

We investigated the feasibility of dual-energy (DE) detection of bone marrow edema (BME) using a dedicated extremity cone-beam CT (CBCT) with a unique three-source x-ray unit. The sources can be operated at different energies to enable single-scan DE acquisitions. However, they are arranged parallel to the axis of rotation, resulting in incomplete sampling and precluding the application of DE projection-domain decompositions (PDD) for beam-hardening reduction. Therefore, we propose a novel combination of a model-based "one-step" DE two-material decomposition followed by a constrained image-domain change-of-basis to obtain virtual non-calcium (VNCa) images for BME detection. DE projections were obtained using an "alternating-kV" protocol by operating the peripheral two sources of the CBCT system at low-energy (60 kV, 0.105 mAs/frame) and the central source at high-energy (100 kV, 0.028 mAs/frame), for a total of 600 frames over 216° of gantry rotation. Projections were processed with detector lag, glare and fast Monte Carlo (MC)-based iterative scatter corrections. Model-based material decomposition (MBMD) was then implemented to obtain aluminum (Al) and polyethylene (PE) volume fraction images with minimal beam-hardening. Statistical ray weights in MBMD were modified to account for regions with highly oblique sampling by the peripheral sources. To generate the VNCa maps, image-domain decomposition (IDD) constrained by the volume conservation principle (VCP) was performed to convert the Al and PE MBMD images into volume fractions of water, fat and cortical bone. Accuracy of BME detection was evaluated using physical phantom data acquired on the multi-source extremity CBCT scanner. The proposed framework estimated the volume of BME with ~10% error. The MC-based scatter corrections and the modified MBMD ray weights were essential to achieve such performance - the error without MC scatter corrections was >30%, whereas the uniformity of estimated VNCa images was 3x improved using the modified weights compared to the conventional weights. The proposed DE decomposition framework was able to overcome challenges of high scatter and incomplete sampling to achieve BME detection on a CBCT system with axially-distributed x-ray sources.

A deep learning model for the diagnosis of sacroiliitis according to Assessment of SpondyloArthritis International Society classification criteria with magnetic resonance imaging

PurposeThe purpose of this study was to develop and evaluate a deep learning model to detect bone marrow edema (BME) in sacroiliac joints and predict the MRI Assessment of SpondyloArthritis International Society (ASAS) definition of active sacroiliitis in patients with chronic inflammatory back pain. Materials and methodsMRI examinations of patients from the French prospective multicenter DESIR cohort (DEvenir des Spondyloarthropathies Indifférenciées Récentes) were used for training, validation and testing. Patients with inflammatory back pain lasting three months to three years were recruited. Test datasets were from MRI follow-ups at five years and ten years. The model was evaluated using an external test dataset from the ASAS cohort. A neuronal network classifier (mask-RCNN) was trained and evaluated for sacroiliac joints detection and BME classification. Diagnostic capabilities of the model to predict ASAS MRI active sacroiliitis (BME in at least two half-slices) were assessed using Matthews correlation coefficient (MCC), sensitivity, specificity, accuracy and AUC. The gold standard was experts' majority decision. ResultsA total of 256 patients with 362 MRI examinations from the DESIR cohort were included, with 27% meeting the ASAS definition for experts. A total of 178 MRI examinations were used for the training set, 25 for the validation set and 159 for the evaluation set. MCCs for DESIR baseline, 5-years, and 10-years follow-up were 0.90 (n = 53), 0.64 (n = 70), and 0.61 (n = 36), respectively. AUCs for predicting ASAS MRI were 0.98 (95% CI: 0.93–1), 0.90 (95% CI: 0.79–1), and 0.80 (95% CI: 0.62–1), respectively. The ASAS external validation cohort included 47 patients (mean age 36 ± 10 [SD] years; women, 51%) with 19% meeting the ASAS definition. MCC was 0.62, sensitivity 56% (95% CI: 42–70), specificity 100% (95% CI: 100–100) and AUC 0.76 (95% CI: 0.57–0.95). ConclusionThe deep learning model achieves performance close to those of experts for BME detection in sacroiliac joints and determination of active sacroiliitis according to the ASAS definition.

DCNAS-Net: deformation convolution and neural architecture search detection network for bone marrow oedema

BackgroundLumbago is a global disease that affects more than 500 million people worldwide. Bone marrow oedema is one of the main causes of the condition and clinical diagnosis is mainly made by radiologists manually reviewing MRI images to determine whether oedema is present. However, the number of patients with Lumbago has risen dramatically in recent years, which has brought a huge workload to radiologists. In order to improve the efficiency of diagnosis, this paper is devoted to developing and evaluating a neural network for detecting bone marrow edema in MRI images.Related workInspired by the development of deep learning and image processing techniques, we design a deep learning detection algorithm specifically for the detection of bone marrow oedema from lumbar MRI images. We introduce deformable convolution, feature pyramid networks and neural architecture search modules, and redesign the existing neural networks. We explain in detail the construction of the network and illustrate the setting of the network hyperparameters.Results and discussionThe detection accuracy of our algorithm is excellent. And its accuracy of detecting bone marrow oedema reached up to 90.6%, an improvement of 5.7% compared to the original. The recall of our neural network is 95.1%, and the F1-measure also reaches 92.8%. And our algorithm is fast in detecting it, taking only 0.144 s per image.ConclusionExtensive experiments have demonstrated that deformable convolution and aggregated feature pyramid structures are conducive for the detection of bone marrow oedema. Our algorithm has better detection accuracy and good detection speed compared to other algorithms.

Preliminary evaluation of dual-energy CT to quantitatively assess bone marrow edema in patients with diabetic foot ulcers and suspected osteomyelitis

ObjectivesThe purpose of this study is to evaluate the value of dual-energy CT (DECT) with virtual non-calcium (VNCa) in quantitatively assessing the presence of bone marrow edema (BME) in patients with diabetic foot ulcers and suspected osteomyelitis.MethodsPatients with a diabetic foot ulcer and suspected osteomyelitis that underwent DECT (80 kVp/Sn150 kVp) with VNCa were retrospectively included. Two observers independently measured CT values of the bone adjacent to the ulcer and a reference bone not related to the ulcer. The patients were divided into two clinical groups, osteomyelitis or no-osteomyelitis, based on the final diagnosis by the treating physicians.ResultsA total of 56 foot ulcers were identified of which 23 were included in the osteomyelitis group. The mean CT value at the ulcer location was significantly higher in the osteomyelitis group (− 17.23 ± 34.96 HU) compared to the no-osteomyelitis group (− 69.34 ± 49.40 HU; p < 0.001). Within the osteomyelitis group, the difference between affected bone and reference bone was statistically significant (p < 0.001), which was not the case in the group without osteomyelitis (p = 0.052). The observer agreement was good for affected bone measurements (ICC = 0.858) and moderate for reference bone measurements (ICC = 0.675). With a cut-off value of − 40.1 HU, sensitivity was 87.0%, specificity was 72.7%, PPV was 69.0%, and NPV was 88.9%.ConclusionDECT with VNCa has a potential value for quantitatively assessing the presence of BME in patients with diabetic foot ulcers and suspected osteomyelitis.Key Points• Dual-energy CT (DECT) with virtual non-calcium (VNCa) is promising for detecting bone marrow edema in the case of diabetic foot ulcers with suspected osteomyelitis.• DECT with VNCa has the potential to become a more practical alternative to MRI in assessing the presence of bone marrow edema in suspected osteomyelitis when radiographs are not sufficient to form a diagnosis.

Risk Factors for Symptomatic Bilateral Lumbar Bone Stress Injury in Adolescent Soccer Players: A Prospective Cohort Study

Background: Lumbar bone stress injury (BSI) is a high-risk long time—loss injury for adolescent soccer players. However, the risk factors for lumbar BSI are unclear. Purpose: To identify the risk factors for bilateral lumbar BSI for adolescent soccer players. Study Design: Case-control study; Level of evidence, 3. Methods: Adolescent soccer players underwent orthopaedic examination, whole-body dual energy x-ray scan, lumbar magnetic resonance imaging (MRI), and muscle tightness testing at baseline. Lumbar lordosis (LL), sacral slope, maturity stage of lumbar vertebral body, and bone marrow edema (BME) at the L5 were examined via MRI. In addition, bone mineral density and content; trunk lean body mass via dual energy x-ray scan; and bilateral muscle tightness including the iliopsoas, hamstrings, and quadriceps were measured. Lumbar BSI was diagnosed as positive bilateral BME and extension-based lumbar pain. All participants were examined twice, one at 6 months and one at 1 year, after the baseline examination. Multivariate logistic regression analysis was performed to identify the risk factors for bilateral lumbar BSI. Results: A total of 69 (26.3%) players were diagnosed with bilateral lumbar BSI. Asymptomatic BME (odds ratio [OR], 4.260; 95% CI, 2.153-8.431), apophyseal stage of the lumbar vertebral body (OR, 3.438; 95% CI, 1.698-6.959), sacral slope relative to LL ≥5° (OR, 4.067; 95% CI, 2.021-8.181), and hamstring tightness ≥50° (OR, 3.221; 95% CI, 1.385-7.489) were significantly associated with bilateral lumbar BSI. Conclusion: The incidence of bilateral lumbar BSI was common at 26.2%. Asymptomatic BME, sacral anterior tilt relative to LL, immature lumbar epiphyses, and hamstring tightness were found to be risk factors for bilateral lumbar BSI. The results of this study suggest that regular MRI examination could facilitate the early detection of BME, and improvement in hamstring flexibility and lumbosacral alignment may prevent bilateral lumbar BSI in young athletes.

Learning Models for Bone Marrow Edema Detection in Magnetic Resonance Imaging

Bone marrow edema (BME) is the term given to the abnormal fluid signal seen within the bone marrow on magnetic resonance imaging (MRI). It usually indicates the presence of underlying pathology and is associated with a myriad of conditions/causes. However, it can be misleading, as in some cases, it may be associated with normal changes in the bone, especially during the growth period of childhood, and objective methods for assessment are lacking. In this work, learning models for BME detection were developed. Transfer learning was used to overcome the size limitations of the dataset, and two different regions of interest (ROI) were defined and compared to evaluate their impact on the performance of the model: bone segmention and intensity mask. The best model was obtained for the high intensity masking technique, which achieved a balanced accuracy of 0.792 ± 0.034. This study represents a comparison of different models and data regularization techniques for BME detection and showed promising results, even in the most difficult range of ages: children and adolescents. The application of machine learning methods will help to decrease the dependence on the clinicians, providing an initial stratification of the patients based on the probability of edema presence and supporting their decisions on the diagnosis.