Geometric deep learning is a relatively new field that combines the principles of deep learning with techniques from geometry and topology to analyze data with complex structures, such as graphs and manifolds. In orthopedic research, geometric deep learning has been applied to a variety of tasks, including the analysis of imaging data to detect and classify abnormalities, the prediction of patient outcomes following surgical interventions, and the identification of risk factors for degenerative joint disease. This review aims to summarize the current state of the field and highlight the key findings and applications of geometric deep learning in orthopedic research. The review also discusses the potential benefits and limitations of these approaches and identifies areas for future research. Overall, the use of geometric deep learning in orthopedic research has the potential to greatly advance our understanding of the musculoskeletal system and improve patient care

Abstract. OBJECTIVES. Application of deep learning approaches to marker trajectories and ground reaction forces (mocap data), is often hampered by small datasets. Enlarging dataset size is possible using some simple numerical approaches, although these may not be suited to preserving the physiological relevance of mocap data. We propose augmenting mocap data using a deep learning architecture called “generative adversarial networks” (GANs). We demonstrate appropriate use of GANs can capture variations of walking patterns due to subject- and task-specific conditions (mass, leg length, age, gender and walking speed), which significantly affect walking kinematics and kinetics, resulting in augmented datasets amenable to deep learning analysis approaches. METHODS. A publicly available (. https://www.nature.com/articles/s41597-019-0124-4. ) gait dataset (733 trials, 21 women and 25 men, 37.2 ± 13.0 years, 1.74 ± 0.09 m, 72.0 ± 11.4 kg, walking speeds ranging from 0.18 m/s to 2.04 m/s) was used as the experimental dataset. The GAN comprised three neural networks: an encoder, a decoder, and a discriminator. The encoder compressed experimental data into a fixed-length vector, while the decoder transformed the encoder's output vector and a condition vector (containing information about the subject and trial) into mocap data. The discriminator distinguished between the encoded experimental data from randomly sampled vectors of the same size. By training these networks jointly using the experimental dataset, the generator (decoder) could generate synthetic data respecting specified conditions from randomly sampled vectors. Synthetic mocap data and lower limb joint angles were generated and compared to the experimental data, by identifying the statistically significant differences across the gait cycle for a randomly selected subset of the experimental data from 5 female subjects (73 trials, aged 26–40, weighing 57–74 kg, with leg lengths between 868–931 mm, and walking speeds ranging from 0.81–1.68 m/s). By conducting these comparisons for this subset, we aimed to assess the synthetic data generated using multiple conditions. RESULTS. We visually inspected the synthetic trials to ensure that they appeared realistic. The statistical comparison revealed that, on average, only 2.5% of the gait cycle showed significantly differences in the joint angles of the two data groups. Additionally, the synthetic ground reaction forces deviated from the experimental data distribution for an average of 2.9% of the gait cycle. CONCLUSIONS. We introduced a novel approach for generating synthetic mocap data of human walking based on the conditions that influence walking patterns. The synthetic data closely followed the trends observed in the experimental data, also in the literature, suggesting that our approach can augment mocap datasets considering multiple conditions, an approach unfeasible in previous work. Creation of large, augmented datasets allows the application of other deep learning approaches, with the potential to generate realistic mocap data from limited and non-lab-based data. Our method could also enhance data sharing since synthetic data does not raise ethical concerns. You can generate and download virtual gait data using our GAN approach from . https://thisgaitdoesnotexist.streamlit.app/. . Declaration of Interest. (b) declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported:I declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research project

Introduction. Three-dimensional (3D) morphological understanding of the hip joint, specifically the joint space and surrounding anatomy, including the proximal femur and the pelvis bone, is crucial for a range of orthopedic diagnoses and surgical planning. While deep learning algorithms can provide higher accuracy for segmenting bony structures, delineating hip joint space formed by cartilage layers is often left for subjective manual evaluation. This study compared the performance of two state-of-the-art 3D deep learning architectures (3D UNET and 3D UNETR) for automated segmentation of proximal femur bone, pelvis bone, and hip joint space with single and multi-class label segmentation strategies. Method. A dataset of 56 3D CT images covering the hip joint was used for the study. Two bones and hip joint space were manually segmented for training and evaluation. Deep learning models were trained and evaluated for a single-class approach for each label (proximal femur, pelvis, and the joint space) separately, and for a multi-class approach to segment all three labels simultaneously. A consistent training configuration of hyperparameters was used across all models by implementing the AdamW optimizer and Dice Loss as the primary loss function. Dice score, Root Mean Squared Error, and Mean Absolute Error were utilized as evaluation metrics. Results. Both the models performed at excellent levels for single-label segmentations in bones (dice > 0.95), but single-label joint space performance remained considerably lower (dice < 0.87). Multi-class segmentations remained at lower performance (dice < 0.88) for both models. Combining bone and joint space labels may have introduced a class imbalance problem in multi-class models, leading to lower performance. Conclusion. It is not clear if 3D UNETR provides better performance as the selection of hyperparameters was the same across the models and was not optimized. Further evaluations will be needed with baseline UNET and nnUNET modeling architectures

Introduction. Inaccurate identification of implants on X-rays may lead to prolonged surgical duration as well as increased complexity and costs during implant removal. Deep learning models may help to address this problem, although they typically require large datasets to effectively train models in detecting and classifying objects, e.g. implants. This can limit applicability for instances when only smaller datasets are available. Transfer learning can be used to overcome this limitation by leveraging large, publicly available datasets to pre-train detection and classification models. The aim of this study was to assess the effectiveness of deep learning models in implant localisation and classification on a lower limb X-ray dataset. Method. Firstly, detection models were evaluated on their ability to localise four categories of implants, e.g. plates, screws, pins, and intramedullary nails. Detection models (Faster R-CNN, YOLOv5, EfficientDet) were pre-trained on the large, freely available COCO dataset (330000 images). Secondly, classification models (DenseNet121, Inception V3, ResNet18, ResNet101) were evaluated on their ability to classify five types of intramedullary nails. Localisation and classification accuracy were evaluated on a smaller image dataset (204 images). Result. The YOLOv5s model showed the best capacity to detect and distinguish between different types of implants (accuracy: plate=82.1%, screw=72.3%, intramedullary nail=86.9%, pin=79.9%). Screw implants were the most difficult implant to detect, likely due to overlapping screw implants visible in the image dataset. The DenseNet121 classification model showed the best performance in classifying different types of intramedullary nails (accuracy=73.7%). Therefore, a deep learning model pipeline with the YOLOv5s and DenseNet121 was proposed for the most optimal performance of automating implants localisation and classification for a relatively small dataset. Conclusion. These findings support the potential of deep learning techniques in enhancing implant detection accuracy. With further development, AI-based implant identification may benefit patients, surgeons and hospitals through improved surgical planning and efficient use of theatre time

Abstract. Objectives. Knee alignment affects both the development and surgical treatment of knee osteoarthritis. Automating femorotibial angle (FTA) and hip-knee-ankle angle (HKA) measurement from radiographs could improve reliability and save time. Further, if the gold-standard HKA from full-limb radiographs could be accurately predicted from knee-only radiographs then the need for more expensive equipment and radiation exposure could be reduced. The aim of this research is to assess if deep learning methods can predict FTA and HKA angle from posteroanterior (PA) knee radiographs. Methods. Convolutional neural networks with densely connected final layers were trained to analyse PA knee radiographs from the Osteoarthritis Initiative (OAI) database with corresponding angle measurements. The FTA dataset with 6149 radiographs and HKA dataset with 2351 radiographs were split into training, validation and test datasets in a 70:15:15 ratio. Separate models were learnt for the prediction of FTA and HKA, which were trained using mean squared error as a loss function. Heat maps were used to identify the anatomical features within each image that most contributed to the predicted angles. Results. FTA could be predicted with errors less than 3° for 99.8% of images, and less than 1° for 89.5%. HKA prediction was less accurate than FTA but still high: 95.7% within 3°, and 68.0 % within 1°. Heat maps for both models were generally concentrated on the knee anatomy and could prove a valuable tool for assessing prediction reliability in clinical application. Conclusions. Deep learning techniques could enable fast, reliable and accurate predictions of both FTA and HKA from plain knee radiographs. This could lead to cost savings for healthcare providers and reduced radiation exposure for patients

Quantitative ultrasound (QUS) is a promising tool to estimate bone structure characteristics and predict fragile fracture. The aim of this pilot cross-sectional study was to evaluate the performance of a multi-channel residual network (MResNet) based on ultrasonic radiofrequency (RF) signal to discriminate fragile fractures retrospectively in postmenopausal women. Methods. RF signal and speed of sound (SOS) were obtained using an axial transmission QUS at one‐third distal radius for 246 postmenopausal women. Based on the involved RF signal, we conducted a MResNet, which combines multi-channel training with original ResNet, to classify the high risk of fragility fractures patients from all subjects. The bone mineral density (BMD) at lumber, hip and femoral neck acquired with DXA was recorded on the same day. The fracture history of all subjects in adulthood were collected. To assess the ability of the different methods in the discrimination of fragile fracture, the odds ratios (OR) calculated using binomial logistic regression analysis and the area under the receiver operator characteristic curves (AUC) were analyzed. Results. Among the 246 postmenopausal women, 170 belonged to the non-fracture group, 50 to the vertebral group, and 26 to the non-vertebral fracture group. MResNet was discriminant for all fragile fractures (OR = 2.64; AUC = 0.74), for Vertebral fracture (OR = 3.02; AUC = 0.77), for non-vertebral fracture (OR = 2.01; AUC = 0.69). MResNet showed comparable performance to that of BMD of hip and lumbar with all types of fractures, and significantly better performance than SOS all types of fractures. Conclusions. the MResNet model based on the ultrasonic RF signal can significantly improve the ability of QUS device to recognize previous fragile fractures. Moreover, the performance of the proposed model modified by age, weight, and height is further optimized. These results open perspectives to evaluate the risk of fragile fracture applying a deep learning model to analyze ultrasonic RF signal

Introduction. The ability to walk over various surfaces such as cobblestones, slopes or stairs is a very patient centric and clinically meaningful mobility outcome. Current wearable sensors only measure step counts or walking speed regardless of such context relevant for assessing gait function. This study aims to improve deep learning (DL) models to classify surfaces of walking by altering and comparing model features and sensor configurations. Method. Using a public dataset, signals from 6 IMUs (Movella DOT) worn on various body locations (trunk, wrist, right/left thigh, right/left shank) of 30 subjects walking on 9 surfaces were analyzed (flat ground, ramps (up/down), stairs (up/down), cobblestones (irregular), grass (soft), banked (left/right)). Two variations of a CNN Bi-directional LSTM model, with different Batch Normalization layer placement (beginning vs end) as well as data reduction to individual sensors (versus combined) were explored and model performance compared in-between and with previous models using F1 scores. Result. The Bi-LSTM architecture improved performance over previous models, especially for subject-wise data splitting and when combining the 6 sensor locations (e.g. F1=0.94 versus 0.77). Placement of the Batch Normalization layer at the beginning, prior to the convolutional layer, enhanced model understanding of participant gait variations across surfaces. Single sensor performance was best on the right shank (F1=0.88). Conclusion. Walking surface detection using wearable IMUs and DL models shows promise for clinically relevant real-world applications, achieving high F1 levels (>0.9) even for subject-wise data splitting enhancing the model applicability in real-world scenarios. Normalization techniques, such as Batch Normalization, seem crucial for optimizing model performance across diverse participant data. Also single-sensor set-ups can give acceptable performance, in particular for specific surface types of potentially high clinical relevance (e.g. stairs, ramps), offering practical and cost-effective solutions with high usability. Future research will focus on collecting ground-truth labeled data to investigate system performance in real-world settings

Patients with advanced cancer can develop bone metastases in the femur which are often painful and increase the risk of pathological fracture. Accurate segmentation of bone metastases is, amongst others, important to improve patient-specific computer models which calculate fracture risk, and for radiotherapy planning to determine exact radiation fields. Deep learning algorithms have shown to be promising to improve segmentation accuracy for metastatic lesions, but require reliable segmentations as training input. The aim of this study was to investigate the inter- and intra-operator reliability of manual segmentation of femoral metastatic lesions and to define a set of lesions which can serve as a training dataset for deep learning algorithms. F. CT-scans of 60 advanced cancer patients with a femur affected with bone metastases (20 osteolytic, 20 osteoblastic and 20 mixed) were used in this study. Two operators were trained by an experienced radiologist and then segmented the metastatic lesions in all femurs twice with a four-week time interval. 3D and 2D Dice coefficients (DCs) were calculated to quantify the inter- and intra-operator reliability of the segmentations. We defined a DC>0.7 as good reliability, in line with a statistical image segmentation study. Mean first and second inter-operator 3D-DCs were 0.54 (±0.28) and 0.50 (±0.32), respectively. Mean intra-operator I and II 3D-DCs were 0.56 (±0.28) and 0.71 (±0.23), respectively. Larger lesions (>60 cm. 3. ) scored higher DCs in comparison with smaller lesions. This study reveals that manual segmentation of metastatic lesions is challenging and that the current manual segmentation approach resulted in dissatisfying outcomes, particularly for lesions with small volumes. However, segmentation of larger lesions resulted in a good inter- and intra-operator reliability. In addition, we were able to select 521 slices with good segmentation reliability that can be used to create a training dataset for deep learning algorithms. By using deep learning algorithms, we aim for more accurate automated lesion segmentations which might be used in computer modelling and radiotherapy planning

Introduction and Objective. Up to 30% of thoracolumbar (TL) fractures are missed in the emergency room. Failure to identify these fractures can result in neurological injuries up to 51% of the casesthis article aimed to clarify the incidence and risk factors of traumatic fractures in China. The China National Fracture Study (CNFS. Obtaining sagittal and anteroposterior radiographs of the TL spine are the first diagnostic step when suspecting a traumatic injury. In most cases, CT and/or MRI are needed to confirm the diagnosis. These are time and resource consuming. Thus, reliably detecting vertebral fractures in simple radiographic projections would have a significant impact. We aim to develop and validate a deep learning tool capable of detecting TL fractures on lateral radiographs of the spine. The clinical implementation of this tool is anticipated to reduce the rate of missed vertebral fractures in emergency rooms. Materials and Methods. We collected sagittal radiographs, CT and MRI scans of the TL spine of 362 patients exhibiting traumatic vertebral fractures. Cases were excluded when CT and/or MRI where not available. The reference standard was set by an expert group of three spine surgeons who conjointly annotated (fracture/no-fracture and AO Classification) the sagittal radiographs of 171 cases. CT and/or MRI were used confirm the presence and type of the fracture in all cases. 302 cropped vertebral images were labelled “fracture” and 328 “no fracture”. After augmentation, this dataset was then used to train, validate, and test deep learning classifiers based on the ResNet18 and VGG16 architectures. To ensure that the model's prediction was based on the correct identification of the fracture zone, an Activation Map analysis was conducted. Results. Vertebras T12 to L2 were the most frequently involved, accounting for 48% of the fractures. Accuracies of 88% and 84% were obtained with ResNet18 and VGG16 respectively. The sensitivity was 89% with both architectures but ResNet18 had a significantly higher specificity (88%) compared to VGG16 (79%). The fracture zone used was precisely identified in 81% of the heatmaps. Conclusions. Our AI model can accurately identify anomalies suggestive of TL vertebral fractures in sagittal radiographs precisely identifying the fracture zone within the vertebral body

Using deep learning and image processing technology, a standardized automatic quantitative analysis systerm of lumbar disc degeneration based on T2MRI is proposed to help doctors evaluate the prognosis of intervertebral disc (IVD) degeneration. A semantic segmentation network BianqueNet with self-attention mechanism skip connection module and deep feature extraction module is proposed to achieve high-precision segmentation of intervertebral disc related areas. A quantitative method is proposed to calculate the signal intensity difference (SI) in IVD, average disc height (DH), disc height index (DHI), and disc height-to-diameter ratio (DHR). According to the correlation analysis results of the degeneration characteristic parameters of IVDs, 1051 MRI images from four hospitals were collected to establish the quantitative ranges for these IVD parameters in larger population around China. The average dice coefficients of the proposed segmentation network for vertebral bodies and intervertebral discs are 97.04% and 94.76%, respectively. The designed parameters of intervertebral disc degeneration have a significant negative correlation with the Modified Pfirrmann Grade. This procedure is suitable for different MRI centers and different resolution of lumbar spine T2MRI (ICC=.874~.958). Among them, the standard of intervertebral disc signal intensity degeneration has excellent reliability according to the modified Pfirrmann Grade (macroF1=90.63%~92.02%). we developed a fully automated deep learning-based lumbar spine segmentation network, which demonstrated strong versatility and high reliability to assist residents on IVD degeneration grading by means of IVD degeneration quantitation

Total knee and hip arthroplasty (TKA and THA) are the most commonly performed surgical procedures, the costs of which constitute a significant healthcare burden. Improving access to care for THA/TKA requires better efficiency. It is hypothesized that this may be possible through a two-stage approach that utilizes prediction of surgical time to enable optimization of operating room (OR) schedules. Data from 499,432 elective unilateral arthroplasty procedures, including 302,490 TKAs, and 196,942 THAs, performed from 2014-2019 was extracted from the American College of Surgeons (ACS) National Surgical and Quality Improvement (NSQIP) database. A deep multilayer perceptron model was trained to predict duration of surgery (DOS) based on pre-operative clinical and biochemical patient factors. A two-stage approach, utilizing predicted DOS from a held out “test” dataset, was utilized to inform the daily OR schedule. The objective function of the optimization was the total OR utilization, with a penalty for overtime. The scheduling problem and constraints were simulated based on a high-volume elective arthroplasty centre in Canada. This approach was compared to current patient scheduling based on mean procedure DOS. Approaches were compared by performing 1000 simulated OR schedules. The predict then optimize approach achieved an 18% increase in OR utilization over the mean regressor. The two-stage approach reduced overtime by 25-minutes per OR day, however it created a 7-minute increase in underutilization. Better objective value was seen in 85.1% of the simulations. With deep learning prediction and mathematical optimization of patient scheduling it is possible to improve overall OR utilization compared to typical scheduling practices. Maximizing utilization of existing healthcare resources can, in limited resource environments, improve patient's access to arthritis care by increasing patient throughput, reducing surgical wait times and in the immediate future, help clear the backlog associated with the COVID-19 pandemic

Foot pain and related problems are quite common in the community. It is reported that 24% of individuals older than 45 experienced foot pain. Also, it is stated that at least two thirds of individuals experiences moderate physical disability due to foot problems. In the absence of evaluation of risk factors such as limited ankle dorsiflexion in the early period of the diseases (Plantar fasciitis, Achilles Tendinopathy e.g.) and the lack of mobile systems with portable remote access, foot pain becomes refractory/chronic foot pain, secondary pathologies and ends with workload of 1., 2. and 3rd level healthcare services. In the literature, manuel and dijital methods have been used to analyze the ankle range of motion (ROM). These studies are generally based on placing protractors on the image and / or angle detection from inclination measurement by using the gyroscope sensor of the mobile device. Some of these applications are effective and they are designed to be suitable for measuring in a clinical setting by a physician or physiotherapist. To the best of our knowledge, there is no system developed to measure real-time ankle ROM remotely with collaboration of the patients. In this research, we proposed to develop an ankle ROM analyze system with smart phone application that can be used comfortably by subjects. We present a case of a 22-year-old male with a symptomatic pes planus. The mobile application, which was used for data collection, was designed and implemented for Android devices. Initially, before the mobile application home page is opened, a consent page was submitted to the acceptance of individual within the scope of Law (KVKK) data privacy. Then, the participant was asked to state his sociodemographic characteristics [age, gender, height, weight] and dominant side. No history of foot-ankle injury, trauma, and surgery was recorded. Activity pain of the foot was 6 according to visual anolog scale (VAS) in the mobile application. His ankle dorsiflexion was 15 ° by manuel goniometer. Besides, server was responsible for storing the collected data and ROM measurement. ROM was calculated by processing the foot video which was sent through the mobile application. During the processing phase, a segmentation model was used which was trained with image process and deep learning methods. With the developed system, we obtained the manual goniometric measurement result with 2 degrees deviation. As the application is calibrated, it is expected to approach the actual measurement of ROM. We can conclude that mobile app-goniometer result in dorsiflexion measurement is a novel promising evaluation method for ankle ROM. it will be easy and practical to detect and monitor risk factor of the diseases, decrease medical costs, provide health services in rural areas, and contribution to life quality and to reduce the workload on physicians and physiotherapist