Impact of metadata in multimodal classification of bone tumours.

The accurate classification of bone tumours is crucial for guiding clinical decisions regarding treatment and follow-up. However, differentiating between various tumour types is challenging due to the rarity of certain entities, high intra-class variability, and limited training data in clinical practice. This study proposes a multimodal deep learning model that integrates clinical metadata and X-ray imaging to improve the classification of primary bone tumours. The dataset comprises 1,785 radiographs from 804 patients collected between 2000 and 2020, including metadata such as age, affected bone site, tumour position, and gender. Ten tumour types were selected, with histopathology or tumour board decisions serving as the reference standard. METHODS: Our model is based on the NesT image classification model and a multilayer perceptron with a joint fusion architecture. Descriptive statistics included incidence and percentage ratios for discrete parameters, and mean, standard deviation, median, and interquartile range for continuous parameters. RESULTS: The mean age of the patients was 33.62 ± 18.60 years, with 54.73% being male. Our multimodal deep learning model achieved 69.7% accuracy in classifying primary bone tumours, outperforming the Vision Transformer model by five percentage points. SHAP values indicated that age had the most substantial influence among the considered metadata. CONCLUSION: The joint fusion approach developed in this study, integrating clinical metadata and imaging data, outperformed state-of-the-art models in classifying primary bone tumours. The use of SHAP values provided insights into the impact of different metadata on the model's performance, highlighting the significant role of age. This approach has potential implications for improving diagnostic accuracy and understanding the influence of clinical factors in tumour classification.

Evaluation of the sagittal vertical axis with postural and 3D motion analyses in lumbar spinal stenosis.

BACKGROUND: Sagittal balance, commonly impaired in lumbar spinal stenosis (LSS) patients, is typically assessed using the sagittal vertical axis (SVA) with EOS imaging. However, to limit X-ray exposure and medical costs, it could be interesting to evaluate the capacity of quantified motion analysis to estimate the 3D modeling of SVA in patients with symptomatic LSS. METHODS: An estimation of the SVA in patients with LSS was performed with 3D motion analysis. SVA and "C7_PSI" (orthogonal horizontal distance between the vertical lines through the markers of the C7 vertebra and the middle of the posterosuperior iliac spine) were measured on 37 LSS patients using EOS radiography and postural and 3D motion analysis, respectively. Multiple stepwise linear regressions were performed with EOS SVA according to age, body mass index, C7_PSI and/or postural variables. RESULTS: A highly significant relationship was found between SVA and C7_PSI, mediolateral amplitude of CoP displacements and age (adjusted R²=0.69, p < 0.0001). While the postural analysis did not reveal significant relationships, the model using 3D parameters revealed significant relationships between radiographic SVA and C7_PSI and age (adjusted R²=0.65, p < 0.0001). 3D motion parameters with or without postural parameters may explain more than 65% of the variance seen in EOS imaging performed on LSS patients. CONCLUSIONS: These promising results in LSS patients suggest that the estimation of SVA with 3D motion analysis offers an alternative to EOS. In addition, SVA could be assessed at rest and during dynamic tasks. TRIAL REGISTRATION: This study has been published in Clinical Trial registration (reference NCT03194607).

Sex differences in the radiographic and symptomatic prevalence of knee and hip osteoarthritis.

Recognising sex differences in disease prevalence can lead to clues as to its pathogenesis, for example the role of hormonal factors and related influences such as body composition, as well as forming the basis for new treatments. However, if different methods are used to define the disorder it can be difficult to explore differences in prevalence, making it necessary to draw on multiple sources of evidence. This narrative review addresses sex differences in the prevalence of knee and hip osteoarthritis, which are the most common forms of large joint osteoarthritis. Females appear to have a higher prevalence of knee osteoarthritis across a wide range of disease definitions, while findings for the hip vary depending on how the disease is defined. Clinically or symptomatically defined hip osteoarthritis is more common in females, whereas radiographically defined hip osteoarthritis is more common in males. Therefore, understanding sex differences in large joint arthritis requires consideration that osteoarthritis, as defined structurally, more commonly affects females at the knee, whereas the opposite is true at the hip. Furthermore, despite structural changes in hip osteoarthritis being more common in males, symptomatic hip osteoarthritis is more common in females. The basis for these disparities is currently unclear, but may reflect a combination of hormonal, biomechanical and behavioural factors.

Revision knee arthroplasty using a modular system manufactured in Brazil. Clinical and radiographic results with a mean nine-year follow-up.

OBJECTIVE: To demonstrate the clinical and radiographic results of revision knee arthroplasty using a modular system manufactured in Brazil. METHODS: Between November 2010 and January 2017, 31 revision knee arthroplasties were performed in 30 patients, using the MB-V system. Patients were assessed clinically and radiographically after a minimum follow-up of two years using the following scores: Knee Society Score (KSS), Knee Society Score ‒ Function (KSS ‒ Function), and Knee Society Roentgenographic Evaluation System (KSRES). RESULTS: Among the patients, 19 were women. The mean age at surgery was 68 years. The mean follow-up was 9.1 years. There was no aseptic loosening in this period. The mean values of KSS, KSS-function, and KSRES were, respectively, 82, 77, and less than 4. One patient evolved with postoperative flexion instability. Two knees became infected, requiring the removal of the prosthesis and implantation of spacers loaded with antibiotics. One patient developed arthrofibrosis. One patient had a dislocation with full medial collateral avulsion and had to be re-operated with an insert exchange to a CCK liner and medial collateral reconstruction as proposed by Krakow. CONCLUSION: This implant had 93.5 % survivorship (no cases of aseptic failure) and good results in 27 out of 31 knees (87 % of the surgeries).

Influence of Knee Osteoarthritis Severity, Knee Pain, and Depression on Physical Function: A Cross-Sectional Study.

Purpose: Decreased physical function with increasing life expectancy is a public health concern worldwide. Knee osteoarthritis (KOA) is considered one of the primary illnesses causing decreased physical function. Depression affects decreased physical function and is closely related to knee pain in KOA. However, the effect of these interacting factors on physical function is not clear. Patients and Methods: We conducted a cross-sectional analysis of the baseline data of 1106 subjects of the 2009 Locomotive Syndrome and Health Outcome in Aizu Cohort Study (LOHAS). We determined the association between their Timed Up and Go test (TUG) scores and radiographic KOA, knee pain, and depression in a multivariate analysis. Results: Severe knee pain was significantly associated with decreased physical function (the odds ratio [OR] was 2.13, 95% confidence interval [CI]: 1.32-4.89), as was depression (OR 2.64, 95% CI 1.61-4.33). Only Kellgren-Lawrence (KL) grade 4 was significantly associated with decreased physical function in the radiographic KOA severity (OR 6.58, 95% CI 1.75-24.68). Conclusion: Severe knee pain and depression were significantly associated with decreased physical function, but not radiographic KOA severity except for KL grade 4. The limitations of using radiographic KOA severity alone as the indicator of assessment for physical function were suggested. When assessing decreased physical function, the clinical focus tends to be on radiographic KOA severity, but it is important to consider the patient's knee pain and psychological factors.

Radiographic features and subtypes of congenital thumb duplication type C3 according to Wu et al. and their potential implications for surgical management: new classification and preliminary results.

BACKGROUND: Wu et al. introduced a modified radiographic system that allows classification of all forms of CTD with excellent interobserver and intraobserver reliability. No study to date has evaluated the radiographic characteristics of Wu et al. type C3 CTD with osseous attachment at the level of the metacarpal. OBJECTIVE: This study aimed to evaluate the radiographic features of type C3 CTD according to the system of Wu et al., to describe the different anatomical subtypes of the duplication, and to propose a categorization approach to distinguish diverse surgical strategies based on the radiographic anatomy of this specific subtype of duplication. METHODS: We performed a retrospective analysis of 215 patients (221 thumbs) diagnosed with Wu et al. type C3 CTD at our Institution between 2015 and 2021. We evaluated all CTDs by examining the alignment of the interphalangeal (IP) and metacarpophalangeal (MP) joints and by assessing the presence of abnormal hypertrophic epiphysis of the primary thumb on posteroanterior (PA) radiographs. The proposed classification system has four types: Type I with good alignment of both MP and IP joints, Type II with ulnar deviation of the MP joint, Type III with radial deviation in the MP joint and Type IV with abnormal hypertrophic epiphysis of the distal phalanx of the main thumb with ulnar deviation of the IP joint with or without ulnar deviation of the MP joint. RESULTS: There were 140 male and 75 female patients with CTD (221 thumbs). There were 65 left, 144 right and 6 bilateral forms. The right-to-left, male-to-female and unilateral-to-bilateral ratios were 2.2:1, 1.9:1 and 35.8:1 respectively. The mean age at surgery was 22.3 ± 11.8 months (range, 8-80). The proposed classification system allowed the classification of all CTDs (n = 221). Specifically, 53 fingers were classified as Type I (24%), 136 as Type II (61.5%), 21 as Type III (9.5%), and 11 as Type IV (5%). CONCLUSION: The proposed system is based on radiographic pathoanatomy and complements that of Wu et al. by identifying four distinct subtypes of deformity. It has the potential to improve inter-professional communication and guide surgery in patients with Wu et al. type C3 CTD. However, our results are preliminary and further research is needed to validate them. LEVEL OF EVIDENCE: III.

Distal sagittal forelimb conformation in young Walloon horses: Radiographic assessment and its relationship with osteochondral fragments.

Osteochondral fragments within equine joints are commonly encountered and may predispose to lameness and limitation to sport purposes. Factors leading to this condition include genetic, nutritional and environmental conditions. However, few studies have evaluated the impact of conformation traits and their correlation with osteochondrosis. This study, based on the radiographic screenings of young horses born in Wallonia (266 individuals, 532 forelimbs), evaluated the correlation between foot, fetlock conformations of the front limb, height at the withers and presence of osteochondral fragments. Moreover, for all traits significantly associated with the presence of osteochondral fragments, a Receiver Operator Characteristic (ROC) curve, area under the curve and optimal cut-off value were calculated to predict the occurrence of fragments. Mean dorsal hoof wall angle was 52.36°, dorsal and palmar angle of the third phalanx were respectively 49.83° and 2.99°, and dorsal metacarpophalangeal angle 147.99°. Moreover, the prevalence of upright feet, defined as having an inclined profile of >2° steeper in relation to its contralateral counterpart, was 24%. Increased palmar angle of the distal phalanx was significantly correlated (P < 0.05) with presence of fragments located at the dorso-proximal margin of the proximal phalanx. The associated area under the curve was 0.623 (95% CI: 0528-0.717, P < 0.05) and the optimal cut-off value to predict fragment occurrence was 2.95° (sensitivity 77.3%; specificity 52.9%). Furthermore, the third metacarpal bone diameter of the left forelimb and height at the withers were significantly (P < 0.05) correlated with the presence of osteochondral fragments in general and within tarsocrural and metatarsophalangeal joints specifically. The area under the curve was 0.585 (95% CI: 0.513-0.656, P < 0.05) with an optimal cut-off value of 152.5 cm (sensitivity 85.1%; specificity 31.2%) for height at the withers to predict presence of osteochondral fragment; to predict the occurrence of osteochondral fragment in any joint on the basis of the third metacarpal bone diameter, the area under the curve was 0.595 (95% CI: 0.524-0.667, P <0.05) and the optimal cut-off value 34.9 mm (sensitivity 52.5%; specificity 64.9%). This study provides information about phenotypic traits associated with osteochondral fragments in horses. Although the diagnostic accuracy of these traits to detect osteochondral fragment was limited, the identification of more phenotypic characteristics could, in the future, make it possible to generate models for accurately identifying individuals at high risk of osteochondral fragments on the basis of their phenotype.

Evaluation of Intervertebral Disc Space Width Between 10th and 11th Thoracic Vertebrae in Cats Using Radiography and Magnetic Resonance Imaging.

BACKGROUND: It is well known that the T10-T11 intervertebral disc space is usually narrower than the adjacent intervertebral disc spaces in most dogs. However, whether the T10-T11 disc space is narrower than the adjacent disc spaces is not well-established in cats. OBJECTIVES: To measure the intervertebral disc space width of the T10-T11 disc space and to compare it with that of the adjacent disc spaces using radiography and magnetic resonance imaging (MRI). METHODS: In Groups 1 (101 clinically normal cats) and 2 (32 cats without spinal diseases from T8 to T13), dorsal width (DW), central width (CW) and ventral width (VW) of the T8-T13 disc spaces were measured by radiography or MRI. Significant differences of the DW, CW and VW among disc spaces measured by radiography and MRI were evaluated. Intra- and interobserver reliabilities were measured by intraclass correlation coefficient. RESULTS: In Groups 1 and 2 using radiography, the DW and CW of the T10-T11 disc space were significantly narrower than those of adjacent disc spaces (p < 0.05). The DW of the T8-T12 disc spaces was significantly narrower than that of the T12-T13 disc space in Group 1 using radiography. In Groups 1 and 2 using radiography, the CW and VW of the T8-T11 disc spaces were significantly narrower than those of the T11-T13 disc spaces. In Group 2 using MRI, the DW of the T10-T11 disc space was significantly narrower than that of the T11-T13 disc spaces. The VW of the T8-T11 disc spaces was significantly narrower than that of the T11-T13 disc spaces. CONCLUSION: The T10-T11 intervertebral disc space on radiography is statistically narrower than the adjacent intervertebral disc spaces in normal cats. The cranial thoracic intervertebral disc spaces (T8-T11) are usually narrower than the caudal thoracic intervertebral disc spaces (T11-T13) on radiography and MRI.

Clinical benchmark dataset for AI accuracy analysis: quantifying radiographic annotation of pelvic tilt.

Radiographic landmark annotation determines patients' anatomical parameters and influences diagnoses. However, challenges arise from ambiguous region-based definitions, human error, and image quality variations, potentially compromising patient care. Additionally, AI landmark localization often presents its predictions in a probability-based heatmap format, which lacks a corresponding clinical standard for accuracy validation. This Data Descriptor presents a clinical benchmark dataset for pelvic tilt landmarks, gathered through a probabilistic approach to measure annotation accuracy within clinical environments. A retrospective analysis of 115 pelvic sagittal radiographs was conducted for annotating pelvic tilt parameters by five annotators, revealing landmark cloud sizes of 6.04 mm-17.90 mm at a 95% dataset threshold, corresponding to 9.51°-16.55° maximum angular disagreement in clinical settings. The outcome provides a quantified point cloud dataset for each landmark corresponding to different probabilities, which enables assessment of directional annotation distribution and parameter-wise impact, providing clinical benchmarks. The data is readily reusable for AI studies analyzing the same landmarks, and the method can be easily replicated for establishing clinical accuracy benchmarks of other landmarks.

Global tilt angle in spinal alignment of asymptomatic older women with low bone mass.

This study investigated the relationship between sagittal spinal alignment and musculoskeletal health in older women, particularly those exhibiting diminished bone density without apparent symptoms. The study assessed the impact of global tilt (GT) and the presence of coronal malalignment on spinal health. The research involved 165 asymptomatic older women with an average age of 68.91 ± 7.25 years and average body mass index of 24.26 ± 3.66 kg/m². Comprehensive standing anteroposterior and lateral spine radiographs were used for assessment. Through multivariate linear regression analysis, the study identified significant correlations between increased GT angles and various factors, including coronal malalignment, the C7 slope, and pelvic incidence. This led to the formulation of a predictive GT model: GT = -9.79 + 0.06 × menopausal period + 0.19 × body mass index (BMI)- 0.81 × average T score - 0.11 × grip strength + 3.03 × (presence of coronal malalignment) + 0.08 × sagittal vertical axis (SVA) + 0.12 × C7 slope - 0.35 × upper lumbar lordosis (ULL) - 0.43 × lower lumbar lordosis (LLL) + 0.70 × pelvic incidence (PI), with an adjusted R² of 0.816. The study findings highlight the prevalence of coronal malalignment in this demographic and its significant associations with critical spinal parameters. The proposed GT predictive model may enable development of personalized treatment plans for older women with low bone mass.

Artificial intelligence in commercial fracture detection products: a systematic review and meta-analysis of diagnostic test accuracy.

Conventional radiography (CR) is primarily utilized for fracture diagnosis. Artificial intelligence (AI) for CR is a rapidly growing field aimed at enhancing efficiency and increasing diagnostic accuracy. However, the diagnostic performance of commercially available AI fracture detection solutions (CAAI-FDS) for CR in various anatomical regions, their synergy with human assessment, as well as the influence of industry funding on reported accuracy are unknown. Peer-reviewed diagnostic test accuracy (DTA) studies were identified through a systematic review on Pubmed and Embase. Diagnostic performance measures were extracted especially for different subgroups such as product, type of rater (stand-alone AI, human unaided, human aided), funding, and anatomical region. Pooled measures were obtained with a bivariate random effects model. The impact of rater was evaluated with comparative meta-analysis. Seventeen DTA studies of seven CAAI-FDS analyzing 38,978 x-rays with 8,150 fractures were included. Stand-alone AI studies (n = 15) evaluated five CAAI-FDS; four with good sensitivities (> 90%) and moderate specificities (80-90%) and one with very poor sensitivity (< 60%) and excellent specificity (> 95%). Pooled sensitivities were good to excellent, and specificities were moderate to good in all anatomical regions (n = 7) apart from ribs (n = 4; poor sensitivity / moderate specificity) and spine (n = 4; excellent sensitivity / poor specificity). Funded studies (n = 4) had higher sensitivity (+ 5%) and lower specificity (-4%) than non-funded studies (n = 11). Sensitivity did not differ significantly between stand-alone AI and human AI aided ratings (p = 0.316) but specificity was significantly higher the latter group (p < 0.001). Sensitivity was significant lower in human unaided compared to human AI aided respectively stand-alone AI ratings (both p ≤ 0.001); specificity was higher in human unaided ratings compared to stand-alone AI (p < 0.001) and showed no significant differences AI aided ratings (p = 0.316). The study demonstrates good diagnostic accuracy across most CAAI-FDS and anatomical regions, with the highest performance achieved when used in conjunction with human assessment. Diagnostic accuracy appears lower for spine and rib fractures. The impact of industry funding on reported performance is small.

Radiographic Measurements of the Hip Joint in Capuchin Monkeys (Sapajus spp.).

BACKGROUND: Radiographs are useful for the initial evaluation of the hip joints. The information can be utilized for the betterment of animal health or other goals such as anatomic studies and gait analysis, among others. Therefore, this study aimed to evaluate radiographic measurements of the hip joint in capuchin monkeys, kept under human care at a reference center for wildlife. METHODS: Twelve capuchin monkeys (Sapajus spp.) (three adult males, seven adult females, and two sub-adult females) were evaluated. Ventrodorsal radiographic views were taken under chemical restraint. All measurements on the digital images were performed in triplicate by one examiner. RESULTS AND CONCLUSIONS: None of the measurements evaluated were statistically different between males and females. No statistical differences were found between hind limbs. The mean (±SD) Norberg angle was 104.92° (±2.82°) and the Wiberg angle was 15.26° (±1.86°). The percentage of the femoral head covered by the acetabulum was 68.57% (±3.65%) and the acetabular index depth to width ratio was 54.66% (±3.85%). In conclusion, the radiographic measurements showed certain morphological features of the hip joint in Sapajus spp. that contribute to improving species knowledge.

A systematic review on how to diagnose deltoid ligament injuries-are we missing a uniform standard?

BACKGROUND: Up to now, there is no convincing evidence, that surgical treatment of deltoid ligament injuries, especially in the setting of ankle fractures, does result in improved outcome. One reason could be a missing diagnostic standard. The aim of the current systematic review was to analyze the applied diagnostic strategies for acute deltoid ligament injuries in outcome studies. METHODS: MEDLINE, Scopus, Central, and EMBASE were searched through February 2022 for any original studies addressing diagnostics of acute deltoid injuries. The study was conducted per the PRISMA guidelines. The inclusion criteria were formed according to the PICOS criteria. The data assessed were study type, level of evidence, included fractures, time point and method of diagnosing deltoid ligament layers, differentiation between layers and syndesmotic injuries. RESULTS: 31 studies were included in the final analysis. Most studies (n = 28) based their decision to treat the deltoid ligament injury on radiologic findings only, with stressed radiographs (n = 18) being the most common. The radiographs were applied at one or more time points (preoperative, before ORIF, after ORIF, after ORIF and syndesmotic repair). The most frequently assessed parameter was the Medial Clear Space (MCS, n = 27) with cut-off-values considered pathological ranging between MCS > 1 mm and MCS > 6 mm. CONCLUSION: Comparing the 31 studies shows that a standardized method to diagnose deltoid ligament injuries is missing. Further research is needed to establish evidence-based guidelines on how to diagnose acute deltoid ligament injuries. TRIAL REGISTRATION: Prospero ID: CRD42022307112. CLINICAL TRIAL NUMBER: not applicable.

Different Radiological Indices of Patellar Height Predict Patients' Diverse Outcomes Following Total Knee Arthroplasty.

Background: Total knee arthroplasty (TKA) is a common surgical procedure for patients with knee osteoarthritis. The patellar component plays a crucial role in knee biomechanics and can influence postoperative outcomes. This study aimed to investigate the relationship between radiological indices of patellar height and patient outcomes following TKA. Methods: A retrospective analysis was conducted on patients who underwent TKA for osteoarthritis. Radiographic measurements of patellar height, including the Insall-Salvati (IS) ratio, modified Blackburne-Peel (mBP) ratio, Caton-Deschamps ratio, and plateau-patellar angle (PPA), were obtained. Clinical outcomes were assessed using the Knee Society Score (KSS) and the Forgotten Joint Score-12 (FJS-12). Patient satisfaction and postoperative complications were also evaluated. Statistical analyses, including correlation analysis and multiple regression models, were performed to determine the association between radiological indices and patient outcomes. Results: The study included 330 cases that met the inclusion criteria. The analysis revealed significant correlations between different radiological indices of patellar height and patient outcomes. Lower postoperative PPA was correlated with worse KSS and range of motion scores. A decreased mBP ratio was associated with poorer FJS-12 responses and higher risks of dissatisfaction and patellar clunk or crepitus. Increased IS ratio was linked to a lower likelihood of incidental giving way of the knee. Advanced age was associated with reduced dissatisfaction and incidental giving way probabilities. Conclusions: The findings of this study demonstrate that radiological indices of patellar height can predict patient outcomes following TKA. Assessing patellar height using various radiographic measurements provides valuable information for surgical planning and prognostic evaluation. Understanding the impact of patellar height on clinical outcomes can aid in optimizing TKA procedures and improving patient satisfaction. These findings emphasize the importance of considering patellar height as a predictive factor in TKA and highlight its potential role in guiding postoperative management and rehabilitation strategies.

Pelvic Binder Radiography Is Safe and Feasible for Quantifying Fracture Instability in LC1 Pelvis Fractures: A Clinical Trial.

OBJECTIVES: To demonstrate fracture displacement, patient tolerance, and in-hospital safety quantifying lateral compression type 1 (LC1) pelvis fracture stability with awake stress radiography using a pelvic binder (PBR). DESIGN: Prospective clinical trial of PBR diagnostic intervention. SETTING: Two centers, Level I trauma center and academic hospital. PATIENT SELECTION CRITERIA: Alert adults with LC1 (OTA/AO B1.1/2.1) pelvic fractures from blunt mechanism presenting within 3 weeks of injury between February and December 2023 without hypotension or injury precluding pelvic binder application. OUTCOME MEASURES AND COMPARISONS: The primary outcome was fracture displacement on PBR. Secondary outcomes included feasibility (≥85% patient tolerance) and safety (no major adverse events). Exploratory outcomes included pain during examination, opioid medication consumption within 96 hours, length of hospital stay, discharge destination, and correlation between fracture displacement on PBR and examination under anesthetic. RESULTS: One hundred sixty-nine patients with pelvis fractures were screened, 58 were eligible, and 31 were enrolled. The mean age of enrolled patients was 58.7 ± 23 years, 54.8% were female, 22.6% were White, and 67.7% Hispanic. Fracture displacement was 4.3 ± 4.7 mm on PBR at 5 kg and 8.4 ± 9.0 mm at 10 kg. All patients tolerated PBR. Sixteen patients (52%) reported pain during PBR but completed the test. No patient experienced an adverse event. Patients received 40.4 ± 56.4 mg oral morphine equivalents within 96 hours of PBR. Ten patients (32.2%) received internal fixation. The median hospital stay was 6 days (interquartile range 5.5 days) with no inpatient mortality. Fourteen patients (45.2%) discharged home. CONCLUSIONS: PBR is feasible and safe as a point-of-care test for LC1 pelvis fracture instability. LEVEL OF EVIDENCE: Level II Diagnostic.

External validation of an artificial intelligence multi-label deep learning model capable of ankle fracture classification.

BACKGROUND: Advances in medical imaging have made it possible to classify ankle fractures using Artificial Intelligence (AI). Recent studies have demonstrated good internal validity for machine learning algorithms using the AO/OTA 2018 classification. This study aimed to externally validate one such model for ankle fracture classification and ways to improve external validity. METHODS: In this retrospective observation study, we trained a deep-learning neural network (7,500 ankle studies) to classify traumatic malleolar fractures according to the AO/OTA classification. Our internal validation dataset (IVD) contained 409 studies collected from Danderyd Hospital in Stockholm, Sweden, between 2002 and 2016. The external validation dataset (EVD) contained 399 studies collected from Flinders Medical Centre, Adelaide, Australia, between 2016 and 2020. Our primary outcome measures were the area under the receiver operating characteristic (AUC) and the area under the precision-recall curve (AUPR) for fracture classification of AO/OTA malleolar (44) fractures. Secondary outcomes were performance on other fractures visible on ankle radiographs and inter-observer reliability of reviewers. RESULTS: Compared to the weighted mean AUC (wAUC) 0.86 (95%CI 0.82-0.89) for fracture detection in the EVD, the network attained wAUC 0.95 (95%CI 0.94-0.97) for the IVD. The area under the precision-recall curve (AUPR) was 0.93 vs. 0.96. The wAUC for individual outcomes (type 44A-C, group 44A1-C3, and subgroup 44A1.1-C3.3) was 0.82 for the EVD and 0.93 for the IVD. The weighted mean AUPR (wAUPR) was 0.59 vs 0.63. Throughout, the performance was superior to that of a random classifier for the EVD. CONCLUSION: Although the two datasets had considerable differences, the model transferred well to the EVD and the alternative clinical scenario it represents. The direct clinical implications of this study are that algorithms developed elsewhere need local validation and that discrepancies can be rectified using targeted training. In a wider sense, we believe this opens up possibilities for building advanced treatment recommendations based on exact fracture types that are more objective than current clinical decisions, often influenced by who is present during rounds.

Clinical and radiological outcomes of flexible flatfoot correction with double calcaneal osteotomy.

BACKGROUND: Although double calcaneal osteotomy (medial displacement calcaneal osteotomy with lateral column lengthening) is widely regarded as an effective treatment option for flexible flatfoot, limited studies have extensively analyzed the degree of deformity correction in three dimensions following double calcaneal osteotomy. This study was performed to evaluate the radiographic and clinical effectiveness of double calcaneal osteotomy to correct flexible flatfoot deformities. METHODS: Thirty-one patients who had 44 symptomatic flexible flatfeet and underwent double calcaneal osteotomy were examined retrospectively with a mean follow-up of 50 months. Visual analog scale, foot and ankle activity measure, and other clinical data were obtained from medical records. Various radiographic variables for assessing flatfoot and osteoarthritic change in tarsal joints were analyzed from weightbearing radiographs. RESULTS: Clinical scores and radiographic variables were significantly improved postoperatively. The mean values of medial sliding and lateral lengthening were 7.6 and 8.7 mm, respectively. No osteoarthritic changes were observed. CONCLUSIONS: Double calcaneal osteotomy could be used to correct flatfoot deformities effectively and sustainably and provide symptomatic relief and patient satisfaction. LEVEL OF EVIDENCE: Level 4, retrospective case series.

[The Seymour fracture - a fracture that easily can be missed]. / Seymourfraktur ­ öppen fraktur som lätt missas.

The Seymour fracture is a pediatric fracture involving the growth plate of the distal phalanx. It is associated with an overlying nail bed injury, and the child often presents with difficulties to extend fully in the distal joint. If not treated properly it can result in infections such as osteomyelitis or lead to deformities and growth disturbances. Since the extent of the damage of the nail bed can range from more obvious crush injuries to small lacerations of the matrix it can easily be missed in the more subtle cases. Early administration of prophylactic antibiotics, reduction of the fracture and repair of the nail bed are necessary for a good prognosis.

Demographic characteristics and clinical-radiological correlation in patients with indications for Total Knee Arthroplasty: A cross-sectional study.

OBJECTIVE: The objective of this study is to correlate clinical symptoms and functionality, using the KOOS (Knee Injury and Osteoarthritis Outcome Score) and KSS (Knee Society Score System) scores, with the radiographic changes, using the Kellgren-Lawrence classification, in patients with knee osteoarthritis and indications for Total Knee Arthroplasty (TKA). METHODS: 120 patients (189 knees) with gonarthrosis and indications for TKA were included in the study. Demographic questionnaires were applied, and clinical and functional assessment was carried out using the KOOS and KSS scores. Knee radiographs were taken and graded according to Kellgren-Lawrence. The clinical scores were compared with the radiographic classification to establish a correlation between these two measurements. Statistical analysis was performed using the τ-Kendall correlation test. RESULTS: Weak and inversely proportional correlations were found between the clinical scores and the radiographic classification. Among clinical scores evaluated, KSS showed the highest correlation with Kellgren-Lawrence classification (τ = -0.356; p < 0.001), followed by KOOS-quality of life (τ = -0.176; p = 0.004), KOOS-total score (τ = -0.166; p = 0.004), KOOS-function in daily living (τ = -0.160; p = 0.005) and KOOS-symptoms (τ = -0.159; p = 0.006). KOOS-pain (τ = -0.149; p = 0.01) and KOOS-sport and recreation function (τ = -0.142; p = 0.025) scores had the weakest correlations. CONCLUSION: There is a weak correlation between the clinical-functional scores of TKA candidates and their radiographic classification by Kellgren-Lawrence. Among clinical scores evaluated, KSS had the strongest negative correlation with the radiographic classification.