Search Results (111)

Ultrasound enables radiation-free longitudinal monitoring of scoliosis, but rib shadowing and speckle noise often obscure vertebral structures. Current deep-learning methods present results in terms of localisation accuracy, without directly measuring anatomical completeness. We introduce a vertebra-level completeness model that includes a YOLO-based detection framework and an explicit representation of completeness, the Vertebra Presence Matrix (VPM). The VPM provides visibility into detections across 17 ordinal vertebral levels (T1–T12, L1–L5), allowing us to measure completeness across anatomy rather than just detections. Thoracolumbar ultrasound scans were annotated and divided into train/test sets using a patient-wise split to avoid data leakage. Four model variants were evaluated, including full-spine and vertebra-centric crop representations with single-class and 17-class detection heads. The full-spine detector was less stable in regions of high anatomical variability, such as the upper thoracic and lower lumbar spine. Crops of individual vertebrae were more stable under partial fields of view. The 17-class crop model achieved an mAP50 of 0.929 and a scan-level completeness score of 0.74 using the VPM. These results demonstrate that vertebral completeness can be explicitly quantified and integrated with localisation-based metrics for completeness-aware automated scoliosis evaluation. Full article

Automated identification of lumbar vertebrae plays a key role in modern spine analysis, offering valuable assistance for diagnostic assessment and preoperative decision-making. Despite recent progress in deep learning-based detection methods, accurately localizing vertebral structures remains challenging due to anatomical variability and heterogeneous image quality. To address the difficulty of capturing subtle vertebral structures, we introduce a Multi-Scale Channel Attention Block (MSCABlock) integrated into the YOLO11 backbone. Unlike conventional attention-based or multi-scale convolutional designs, MSCABlock jointly exploits channel-wise feature interaction and multi-scale receptive fields to enhance both local detail sensitivity and contextual representation, while preserving computational efficiency. The proposed approach is designed to improve detection performance without significantly increasing model complexity. Our model is trained and validated using only the AP-view images from the Burapha University Lumbar-Spine Dataset (BUU-LSPINE), which provides well-annotated lumbar spine X-ray images from 400 unique patients. The proposed approach operates in a fully end-to-end manner, allowing vertebrae to be identified directly from input images without relying on handcrafted feature engineering or complex preprocessing pipelines. Experimental evaluations show that the proposed model achieves strong detection performance, with mAP@0.5 and mAP@0.5–0.95 reaching 0.982 and 0.79, respectively, alongside a precision of 0.93 and a recall of 0.975. Compared with the YOLO11 baseline, ablation and efficiency analyses demonstrate that MSCABlock consistently improves detection performance. It introduces only marginal increases in model parameters and computational cost, thereby preserving a lightweight architecture and maintaining efficient inference. These results show that the optimized YOLO11-based system generalizes well across lumbar levels. It maintains reliable detection under challenging conditions, providing robust automated localization to support large-scale clinical spine analysis. Full article

Background: Bone Mineral Density (BMD) plays a crucial role in diagnosing osteoporosis, and early detection is essential to preventing complications such as osteoporotic fractures. However, access to dual-energy X-ray absorptiometry (DXA) screening remains limited in many healthcare settings. Objective: This study presents a fully automated artificial intelligence pipeline for BMD prediction from lumbar spine radiographs to enable opportunistic osteoporosis screening. Methods: The proposed system integrates automatic vertebral segmentation and a machine learning-based regression model for BMD prediction. A YOLO-based instance segmentation model was trained to automatically segment four lumbar vertebrae, achieving a high Intersection over Union (IoU) of 0.9. Radiomic features were extracted from the segmented vertebrae to capture advanced image characteristics and combined with clinical features from 2875 female patients. An eXtreme Gradient Boosting (XGBoost) regressor was trained to provide opportunistic BMD estimation. Results: The model achieved a mean absolute percentage error (MAPE) of 6% for BMD prediction. A classification model built from segmented vertebrae distinguished between osteoporosis, osteopenia, and normal bone with approximately 90% accuracy. Strong agreement between predicted and ground-truth BMD values was confirmed using Pearson correlation coefficient and Bland–Altman analysis. Conclusions: The proposed fully automated system demonstrates strong agreement with DXA measurements and potential for opportunistic osteoporosis screening in settings with limited DXA access. Further validation and refinement are needed to achieve clinical-grade precision for diagnostic applications. Full article

In the context of minimally invasive spine surgery, accurately estimating the 3D coordinates of the vertebrae from intraoperative 2D X-ray images is crucial for aligning preoperative data with the patient’s real-time posture. However, existing methods are hindered by the ill-posed nature of 2D-to-3D localization and the distinctive anatomical features of the spinal column, leading to ambiguities and reduced accuracy. In this paper, we introduce X2P-net, a novel prompt-guided and semantic context-enhanced 2D/3D vertebra detection framework. To achieve this, we design a novel Transformer architecture, referred to as BrickFormer, which can automatically extract the refined vertebral foreground context at low computational cost using a dual-attention mechanism. Comprehensive experiments were conducted to validate the proposed approach on two datasets: a large-scale synthetic dataset (BiSpineX) and a sheep spine dataset (SheepSpineX). Results obtained from these experiments demonstrate superior landmark localization performance of the proposed method compared to other state-of-the-art methods. Specifically, on the BiSpineX dataset, X2P-Net achieves percentages of 96.9% and 98.8% at 10 mm and 20 mm thresholds, respectively, a mean position error of 2.99 mm, and an $AUC$ of 0.9923. Similar superior performance was also observed when the proposed method was applied to the SheepSpineX dataset, with percentages of 98.4% and 100.0% at 10 mm and 20 mm thresholds, respectively, a mean position error of 1.08 mm, and an $AUC$ of 0.9972. Full article

Background: Spondylolysis is a stress fracture of the pars interarticularis, most common in adolescents and athletes involved in sports requiring repetitive spinal loading, extension, and rotation. The condition is often underdiagnosed due to delays in presentation and diagnosis, particularly among non-orthopedic providers. Aims: This review aims to summarize the current understanding of spondylolysis, focusing on its etiology, diagnosis, management strategies, and identify gaps in research for future exploration. Methods: A structured literature search was conducted in PubMed to identify studies relevant to pediatric and adolescent spondylolysis, spondylosis, and spondylolisthesis, particularly in the context of athletic injuries. The initial search yielded 143 citations. Applying filters for English language publications within the past five years reduced this to 125 citations. Limiting to populations that were aged 18 years and under returned 50 studies. After screening the titles and abstracts, 12 non-specific or irrelevant articles (including letters to the editor) were excluded, leaving a final dataset of 38 articles for detailed review. In addition, foundational and landmark studies outside this window were included to provide historical and conceptual context, bringing the total evidence base to 50 papers. Findings: Spondylolysis most commonly affects the L5 vertebra, with a higher incidence in male athletes. Conservative treatments like physical therapy and bracing are effective, especially when initiated early. However, the efficacy of bracing remains debated, with limited evidence on long-term clinical benefits. Surgical intervention is considered for severe or non-responsive cases. Diagnostic methods, including CT and MRI, are preferred, with emerging techniques like ultrasound showing potential for non-ionizing, cost-effective, early detection. Implications: Early diagnosis and treatment are crucial for preventing progression to spondylolisthesis. While conservative treatments often yield favorable outcomes, more research is needed to compare the effectiveness of bracing and pharmacological interventions. Future studies should focus on long-term outcomes, cost-effective, non-ionizing diagnostic methods, and the role of emerging therapies like regenerative medicine. A multi-disciplinary approach is vital for optimal patient care, particularly in young athletes. Full article

Axial spondyloarthritis (axSpA) is a chronic inflammatory disorder causing structural spinal damage and pathological thoracic kyphosis. Accurate quantification of spinal curvature is crucial for monitoring disease progression and guiding treatment. Conventional Cobb angle measurement on radiographs or DEXA images is widely used but is time-consuming and prone to inter-observer variability. This study evaluates an automated deep learning-based approach using a You Only Look Once (YOLO) model for vertebral detection on lateral morphometric DEXA scans and estimation of thoracic kyphosis angles. A dataset of 512 annotated DEXA images, including 182 from axSpA patients, was used to train and test the model. Kyphosis angles were computed by fitting a circle through detected vertebral centroids (Th4–Th12) and calculating the corresponding curvature angle. Model-predicted angles demonstrated strong agreement with physician-measured Cobb angles (r = 0.92, p < 0.001), low mean squared error (4.2°) and high sensitivity and specificity for detecting clinically significant kyphosis. Automated lateral DEXA morphometry provides a rapid, reproducible and clinically interpretable method for assessing thoracic kyphosis and bone density in axSpA, representing a practical tool for integrated structural and metabolic evaluation. Full article

Radiographically visible carotid artery calcifications are typically seen at the level of the C3–C4 cervical vertebrae and can be detected on panoramic dental radiographs. Their early identification is clinically relevant, as they represent a potential marker for increased risk of stroke. In this context, the present study proposes a deep learning method for automatic identification of carotid atheromas using MobileNetV2. From a publicly available dataset, 378 region-of-interest (ROI) images (640 × 320) were prepared and split into train/val/test = 264/57/57 with class counts train 157/107, val 34/23, test 34/23 (negatives/positives). Images underwent standardized preprocessing and on-the-fly augmentation; training used a two-stage scheme (backbone frozen “head” training followed by partial fine-tuning of the top layers), class-weighting, dropout = 0.3, batch normalization (BN) head, early stopping, and partial unfreezing (~70% of the backbone). The decision threshold was selected on validation by Youden’s J. On the independent test set, the model achieved an accuracy (ACC) of 94.7%, sensitivity (SEN) of 95,7%, specificity (SPE) of 0.941, area under the receiver operating characteristic curve (AUC) 0.963, and area under the precision–recall curve (AUPRC) of 0.968. Using a sensitivity-targeted threshold (SEN ≈ 0.80), the model yielded ACC = 91.2%, SEN = 82.6%, and SPE = 97.1%. These results support panoramic radiographs as an opportunistic screening modality for systemic vascular risk and highlight the potential of artificial intelligence (AI)-assisted methods to enable earlier identification within preventive healthcare. Full article

Background/Objectives: Some classification problems involve ordered categories (e.g., low–medium–high), which are better modeled as ordinal regression. This study aimed to propose and evaluate a dual loss framework—Ordinal Residual Dual Loss—for lumbar vertebra classification on CT images to assist L3 identification and sarcopenia detection. Methods: In this retrospective study, lumbar spine CT images were used to train a deep learning model based on a MobileNet-v3-Large network. The proposed framework combines standard cross-entropy loss for classification with an Ordinal Residual Loss defined on the difference between output probabilities and target ordinal probabilities. Results: Experimental results show that the Ordinal Residual Dual Loss approach outperforms using cross-entropy alone and also surpasses methods from previous studies in lumbar vertebra classification performance. Conclusions: Leveraging a dual loss design that incorporates ordinal information improves vertebral level classification on CT images and has potential to support more accurate automated L3 localization and sarcopenia assessment in clinical practice. Full article

Objectives: The aim of this study was to characterize the developmental trajectories of the fifth lumbar vertebra in human fetuses by assessing the growth of its vertebral body and ossification center using linear, planar, and volumetric measurements. Methods: A total of 54 human fetuses (26 male and 28 female) aged 17–30 weeks of gestation were examined. Computed tomography, digital image analysis, 3D reconstruction, and statistical modeling were used to quantify morphometric parameters of the L5 vertebral body and its ossification center. Results: All measured parameters demonstrated consistent age-related growth following a linear pattern. No statistically significant differences between sexes were observed in any measured diameter of the L5 vertebra or its ossification center within the examined gestational age range. Conclusions: The normative morphometric data and growth curves obtained for the L5 vertebra and its ossification center provide age-specific reference values that may aid in prenatal diagnostics. These findings can support clinicians in estimating gestational age, assessing vertebral development on ultrasound, and detecting congenital spinal anomalies and skeletal dysplasias at an early stage. Further multicenter studies including a broader gestational age range are warranted to strengthen the generalizability and clinical applicability of these results. Full article

Background: Free fatal falls (FFF) are a frequent occurrence in forensic medicine. Many variables, such as the victim’s sex, BMI, intoxication, height of the fall, and mental illness, can influence injury patterns. Previous studies identified fracture patterns and frequencies mostly with general anatomical detail, focusing on broad areas. As specific fractures might be roots for new statistical connections, this leaves a gap in our understanding. Using postmortem computed tomography, we aim to establish fracture frequencies and identify possible new statistical connections. Methods: In total, we retrospectively analyzed seventy-nine cases of confirmed deaths due to falls using the database of the Department and Institute of Forensic Medicine in Lublin. Our inclusion criteria were death due to free fall onto hard, non-deformable surfaces. We excluded cases of ground-level falls. All victims must have undergone postmortem computed tomography. Furthermore, each analyzed case documented individual intrinsic variables (sex, age, body mass, height, pre-existing mental conditions, and drug or alcohol use) and extrinsic variables (fall height, landing surface, time between the fall and death, and known cause of the fall). Results: Injuries in free fatal falls tend to focus on the axial skeleton. Suicides experience more severe, bilateral fractures, often involving the pelvis and limbs, while accidents tend to have unilateral injuries with rare limb involvement. We established new correlations with the height of the fall for the maxilla, mandible, anterior and posterior regions of the occipital bone, and the temporal bone. Moreover, our research confirmed previously noted correlations between the height of the fall and fractures of the limbs (and their individual bones), the lumbar vertebrae, and the chest. Conclusions: Our findings highlight that free fatal falls are characterized by distinct skeletal injury patterns that differ between accidents and suicides, with bilateral pelvic and limb fractures being particularly indicative of intentional falls. The integration of PMCT with autopsy improves the detection of these patterns. It provides valuable diagnostic and medico-legal insights, supporting a more precise determination of the cause and manner of death. Full article

Background/Objectives: Lumbosacral transitional vertebra (LSTV) is a common anatomical variant, but its identification on plain radiographs is often inconsistent. This inconsistency can lead to clinical complications such as chronic low back pain, misinterpretation of spinal parameters, and an increased risk of wrong-level surgery. This study aimed to develop and validate a deep learning-based artificial intelligence (AI) model for the automated detection of LSTV on plain lumbar radiographs. Methods: This retrospective observational study included a total of 3116 standing lumbar lateral radiographs. The presence or absence of lumbosacral transitional vertebra (LSTV) was definitively established using whole-spine imaging, CT, or MRI. Multiple deep learning architectures, including DINOv2, CLIP (ViT-B/32), and ResNet-50, were initially evaluated for binary classification of LSTV. Among these, the ResNet-50 model with partial fine-tuning achieved the best test performance and was subsequently selected for fivefold cross-validation using the training set. Model performance was assessed using accuracy, sensitivity, specificity, and the area under the receiver operating characteristic curve (AUROC), and interpretability was evaluated using gradient-weighted class activation mapping (Grad-CAM). Results: On the independent test set of 313 radiographs, the final model demonstrated robust diagnostic performance. It achieved an accuracy of 76.4%, a sensitivity of 85.1%, a specificity of 61.9%, and an AUC of 0.84. The model correctly identified 166 out of 195 LSTV cases and 73 out of 118 normal cases. Conclusions: This AI-based system offers a highly accurate and reliable method for the automated detection of LSTV on plain radiographs. It shows strong potential as a clinical decision-support tool to reduce diagnostic errors, improve pre-operative planning, and enhance patient safety. Full article

Background/Objective: Major research demonstrates that longitudinal loading affects the vertebral growth and disc wedging in the scoliotic animal models; however, there is a scarcity of research on the effect of torque on the vertebral growth. Comparison of the effect of static and dynamic torque on growth is also lacking. The aims of this study were to assess the morphological, histological, and immunohistochemical changes in caudal vertebrae of rats under controlled, static, and dynamic torque. Methods: Adjacent vertebral bodies of female Sprague-Dawley rats were loaded with a torque for 4 weeks. Six rats received a static torque of 1.25 Nm while 6 additional rats received a dynamic torque (2.4 Nm, 1.0 Hz for 15 min/time, 3 times/week). An additional 6 rats formed the control group and received no torque at all. All the rats were later sacrificed, and the tails for histological analysis, immunocytochemistry, and X-rays were obtained. Results: Among the three groups, there were significant differences in right side disc height and average disc height on the proximal vertebrae space in the coronal plane of the X-ray. There were significant differences in the physeal height between static torque and control, or between dynamic torque and control (p < 0.05). The proliferating cell nuclear antigens were detected with variable percentages in samples among the three physeal zones for all groups. Conclusions: Both static and dynamic torque induced asymmetric reduction in the physis and intervertebral disc, which may help to explain the development and vertebral tethering of scoliosis. Full article

Nocardiosis is an infection caused by Gram-positive, saprophytic bacteria most often affecting immunocompromised hosts. The lungs, central nervous system, and skin are the sites most typically involved, although any organ may be affected. Skeletal involvement, particularly osteomyelitis, remains uncommon. This study is a review of all published cases of Nocardia osteomyelitis in humans, emphasizing epidemiology, microbiology, clinical features, management, and patient outcomes. A narrative review was performed using data from the PubMed/MedLine and Scopus databases. Fifty studies describing 55 patients were included. The median age was 54 years, and 65.5% were male. The main risk factors were immunosuppression (21.8%) and trauma (18.2%). The vertebrae constituted the most commonly affected site (25.5%), followed by the lower limb bones (20%); 23.6% had multifocal disease. Nocardia asteroides accounted for the majority of cases (34.8%). Trimethoprim-sulfamethoxazole was the most frequently administered agent (81.5%), followed by cephalosporins (29.6%) and carbapenems (27.8%). Overall mortality was 9.3%, with 5.6% of reported deaths directly attributed to the infection. Although uncommon, osteomyelitis due to Nocardia spp. should be considered when Gram-positive, filamentous microorganisms are detected in bone specimens, particularly in immunocompromised or post-trauma patients, as early suspicion and targeted therapy may improve survival. Full article

Objectives: The present study aimed to analyze the growth dynamics of the ossification centers of the seventh cervical (C7) vertebra in the human fetus, focusing on linear, planar, and volumetric parameters of both the vertebral body and neural processes. Methods: The study was conducted on 55 human fetuses of both sexes (27 males and 28 females), aged 17–30 weeks’ gestation. High-resolution computed tomography, three-dimensional reconstruction, digital image analysis, and appropriate statistical modeling were used to obtain detailed morphometric measurements of the C7 ossification centers. Results: All morphometric parameters—length, cross-sectional area, and volume—of the vertebral body ossification center increased linearly with gestational age, except for the sagittal diameter, which followed a logarithmic growth pattern. Linear growth was likewise observed in all diameters of the neural process ossification centers, including length, width, cross-sectional area, and volume. No statistically significant sex-related or side-related differences were detected. Conclusions: The CT-based morphometric data and growth models for the ossification centers of C7 presented in this study offer preliminary reference values for the vertebra prominens during fetal development. Although limited by sample size, these results establish a baseline that may assist anatomists, radiologists, obstetricians, pediatricians, and spinal surgeons in assessing cervical-spine maturation and in detecting congenital anomalies prenatally. Further studies involving larger and more diverse fetal cohorts are warranted to validate and extend these observations. Full article

Background/Objectives: Fluorine-18 fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) is a valuable imaging modality for detecting malignancies and diagnosing fever of unknown origin (FUO). However, data regarding FDG accumulation in bone marrow among pediatric acute leukemia (AL) cases are limited. In this study, we aimed to compare FDG-PET/CT findings between children with AL and those with inflammatory diseases (IDs), including FUO, and develop a scoring system for differential diagnoses. Methods: We retrospectively analyzed FDG-PET/CT findings in six children with AL and 22 with IDs. The maximum standardized uptake value (SUV max), visual score (VS), and spread score (SS) were evaluated across various bone marrow sites, including vertebrae, pelvic bone, humerus, forearm, and femur. Statistical analysis consisted of Mann–Whitney U test for group comparisons and receiver operating characteristic curve (ROC)/area under the curve (AUC) analyses to assess diagnostic performance. Results: SUV max, VS, and SS were significantly higher in children with AL across all evaluated sites. The combined VS + SS scoring system yielded the highest diagnostic accuracy. A simplified version using only the VS of the middle humerus and femur plus the SS showed comparable effectiveness. Conclusions: FDG-PET/CT in children with AL showed high FDG accumulation in bone marrow areas in the whole body. The simple scoring system, which comprises FDG accumulation in the middle portion of the extremities and the whole body, appears to be helpful in distinguishing AL from IDs in children. FDG-PET/CT-based visual scoring may provide supportive information alongside conventional diagnostics in pediatric acute leukemia. Full article