Dual Energy X-ray Absorptiometry: Radiographer'S Role in Assessing Fracture Risk Assessment Tool (FRAX) Questionnaire Variables.

BACKGROUND: The FRAX® algorithm is a tool used to calculate the 10-year probability of fracture in patients with osteoporosis and is based the assessment of several risk factors. We assessed the performance and accuracy of the completion of the FRAX® anamnestic questionnaire by the radiographer without impact on the clinical workflow. METHODOLOGY: We evaluated the accuracy of fracture risk calculation by the radiographer using the FRAX® algorithm before and after specific training. A total of 100 women were enrolled in the study. The radiographer preliminarily administered the FRAX® questionnaire to all subjects before the execution of the DXA examination. After the end of the examination, a radiologist administered the questionnaire to the patient. Women were divided into two groups: group A (pre-training) and group B (post-training). The radiographer in group A completed the FRAX® questionnaire for the patients before training. For group B, the same radiographer completed the FRAX® questionnaire after training. The results of the FRAX® questionnaire completed by radiographer were compared with that completed by the referring physician. RESULTS: Before training, radiographer's accuracy ranged from 92% (question 7, alcohol consumption) to 36% (question 6, secondary osteoporosis). After training, accuracy values improved substantially, ranging from 100% to 92%. Analysis of the absolute values of FRAX® showed that in the pre-training group data tended to be overestimated by the radiographer, with both major and fractures probabilities being significantly higher when assessed by the radiographer (12% and 5.8%, respectively). After the training, there was a marked decrease in the variation between the FRAX® data calculated by the radiographer and the radiologist. CONCLUSIONS: The accuracy of fracture risk calculation by the radiographer using the FRAX® algorithm is significantly improved after a specific training period. This study demonstrates the importance of dedicated training radiographers on the FRAX® algorithm.

DXA: New Concepts and Tools Beyond Bone Mineral Density.

Since its introduction in 1987, dual-energy X-ray absorptiometry (DXA) has revolutionized bone assessment, becoming the gold standard for measuring bone mineral density (BMD). Its low radiation exposure and high accuracy have made it indispensable in diagnosing osteoporosis, aligning with World Health Organization criteria. However, DXA evolution extends beyond BMD measurement, with emerging tools like the Trabecular Bone Score (TBS) and the DXA-based Bone Strain Index (BSI). TBS provides insights into trabecular bone architecture, enhancing the prediction of fracture risk. Despite limitations like body mass index correlation, TBS aids in evaluating patients with conditions such as diabetes and glucocorticoid exposure. BSI, introduced in 2019, evaluates bone strength using finite element analysis, complementing BMD and TBS by assessing bone fatigue.Advancements in DXA-based tools extend to Hip Structural Analysis and three-dimensional DXA software, offering valuable insights into hip fracture risk. Moreover, DXA serves beyond bone assessment, aiding in abdominal aortic calcification assessment, enhancing cardiovascular risk stratification. In summary, the expanding capabilities of DXA promise comprehensive skeletal and cardiovascular health evaluation, contributing significantly to clinical management and prevention strategies.

Radiofrequency Echographic Multi-Spectrometry in the Diagnosis of Metabolic Bone Disease.

Dual-energy X-ray absorptiometry (DXA) and bone mineral density (BMD) pose several limitations in some patient categories, such as pregnant women and young people. This review article explores whether the innovative radiofrequency echographic multi-spectrometry (REMS) technology is beneficial for assessing the bone condition of various patient groups. Common consequences in patients with acromegalia, prostate cancer undergoing hormone therapy, osteogenesis imperfecta, anorexia nervosa, and in a peritoneal dialysis setting include decreased BMD and an increased risk of fragility fracture.DXA is currently regarded as the gold standard for BMD assessment. However, using the DXA technique has several drawbacks in a young patient who requires repeated BMD tests because it uses ionizing radiation. Because of its precision and consistency, the REMS technique may be a valuable tool to assess changes in bone condition in patients of all ages, particularly in female patients who are fertile or who are pregnant or nursing.

Posteromedial Corner Injuries of the Knee: Imaging Findings.

The posteromedial corner (PMC) of the knee is an anatomical region formed by ligamentous structures (medial collateral ligament, posterior oblique ligament, oblique popliteal ligament), the semimembranosus tendon and its expansions, the posteromedial joint capsule, and the posterior horn of the medial meniscus. Injuries to the structures of the PMC frequently occur in acute knee trauma in association with other ligamentous or meniscal tears. The correct assessment of PMC injuries is crucial because the deficiency of these supporting structures can lead to anteromedial rotation instability or the failure of cruciate ligaments grafts. This article reviews the anatomy and biomechanics of the PMC to aid radiologists in identifying injuries potentially involving PMC components.

Imaging of Body Composition.

Body composition is now recognized to have a major impact on health and disease. Imaging enables its analysis in an objective and quantitative way through diverse techniques such as dual-energy X-ray absorptiometry, computed tomography, magnetic resonance imaging, and ultrasonography. This review article first surveys the methodological aspects underpinning the use of these modalities to assess body composition, highlighting their strengths and limitations as well as the set of parameters that they measure and their clinical relevance. It then provides an update on the main applications of body composition imaging in current practice, with a focus on sarcopenia, obesity, lipodystrophies, cancer, and critical care. We conclude by considering the emerging role of artificial intelligence in the analysis of body composition, enabling the extraction of numerous metrics with the potential to refine prognostication and management across a number of pathologies, paving the way toward personalized medicine.

Bone mineral density differences between femurs of scoliotic patients undergoing quantitative computed tomography analysis.

PURPOSE: Scoliosis is a cause of loading imbalance between the lower limbs, which can result in BMD differences between the two femurs. We investigated the discrepancy in BMD values assessed by quantitative computed tomography (QCT) between femurs in patients with and without scoliosis, also assessing if this difference can be related to spine convexity. METHODS: Abdominal CT examinations were retrospectively reviewed. An ''asynchronous'' calibration of CT images was performed to obtain BMD values from QCT. Scoliosis was evaluated on the antero-posterior CT localizer to calculate the Cobb angle. Differences between aBMD and vBMD of femurs were assessed in both scoliotic and non-scoliotic subjects. RESULTS: Final study cohort consisted of 263 subjects, 225 of them without scoliosis (85.6%) and 38 with scoliosis (14.4%). No significant differences were found in the general population without scoliosis, except for vBMD at the neck. Comparison of femurs in scoliotic patients showed statistically significant differences at neck aBMD -0.028 g/cm2, p = 0.004), total femur aBMD (--0.032 g/cm2, p = 0.008) and total femur vBMD (--8.9 mg/cm3, p = 0.011), with lower BMD values on the convexity side. In 10 cases (26%) a change in the final T-score diagnosis was observed. CONCLUSION: QCT analysis demonstrated a difference in both areal and volumetric BMD between the two femurs of scoliotic patients, in relation to the side of the scoliotic curve. If these data will be confirmed by larger studies, bilateral femoral DXA acquisition may be proposed for these patients.

Diagnostic performance of high-resolution ultrasound in the evaluation of intrinsic and extrinsic wrist ligaments after trauma.

PURPOSE: To investigate the role of ultrasound (US) in the evaluation of intrinsic and extrinsic ligaments of the wrist with magnetic resonance arthrography (MRA) as the reference standard. MATERIALS AND METHODS: This prospective study included patients referred for MRA after wrist trauma. US examination was performed just before MRA. On the dorsal and palmar sides of the wrist, the intrinsic interosseus and midcarpal, extrinsic, and collateral ligaments were evaluated. MRA was performed on a 1.5-T unit. In the first 20 patients included, ligament thickness was independently assessed using US and MRA and thickness reproducibility was calculated. Ligament integrity was evaluated in all patients. RESULTS: 38 patients (22 men, 16 women; mean age: 38 years) were included. Ligament thickness reproducibility ranged between 44% for the palmar ulnocapitate ligament and 71% for the palmar scaphotriquetral ligament. US had a sensitivity, specificity, positive and negative predictive values, and accuracy of 100% in the identification of tears of the palmar (n=8) and dorsal (n=3) bands of the scapholunate ligament and the ulnar collateral ligament (n=3). It had a sensitivity of 100%, specificity of 97%, positive predictive value of 50%, negative predictive value of 100%, and accuracy of 97% in the identification of tears of the palmar ulnolunate ligament (n=1). CONCLUSION: Compared to MRA, US showed good reproducibility in the assessment of wrist ligament thickness and similar accuracy with respect to identifying tears of the scapholunate, palmar ulnolunate, and ulnar collateral ligaments.

High incidence of RAMP lesions and a nonnegligible incidence of anterolateral ligament and posterior oblique ligament rupture in acute ACL injury.

PURPOSE: The purpose of this study was to assess the frequency of medial collateral ligament (MCL), posterior oblique ligament (POL) and anterolateral ligament (ALL) tears and different types of RAMP lesions of patients with verified acute anterior cruciate ligament (ACL) tears by magnetic resonance imaging (MRI). METHODS: MRI was performed on patients with a clinical diagnosis of acute ACL injury. Patients were eligible for inclusion if they had an initially clinically noted ACL tear confirmed on MRI within 30 days of trauma. RESULTS: A total of 146 patients were included in the study, 42 (28.8%) females and 104 (71.2%) males. The mean age at MRI was 27.2 ± 9.4 years, and the mean time from injury to MRI was 15.7 ± 7.8 days. Thirty-four (23.3%) patients had a complete MCL lesion, 32 (21.9%) had a complete POL lesion and 28 (19.2%) had a complete ALL lesion. One hundred and fourteen patients (78.1%) presented with RAMP lesions, while 20 (13.7%) patients reported other meniscal lesions. The mean medial and lateral tibial slopes were 4.0° ± 2.7° and 4.0° ± 3.1°, respectively. Only 10 (6.8%) patients reported no lesions associated with ACL rupture. The most common injuries were isolated RAMP type 3 (18-12.3%) and isolated RAMP type 1 (17-11.6%). Thirteen (8.9%) patients had a combination of MCL, POL and ALL rupture. CONCLUSIONS: Isolated lesions of the ACL are extremely rare. In most cases, a single RAMP lesion should be investigated. In the presence of MCL injury, POL injury should always be suspected as well, while nearly 20% of patients present a rupture of the ALL. About one in 10 patients had three lesions (MCL, ALL and POL), and most of them had a combined RAMP lesion. LEVEL OF EVIDENCE: Level IV.

Health technology assessment in musculoskeletal radiology: the case study of EOSedge™.

OBJECTIVES: Health technology assessment (HTA) is a systematic process used to evaluate the properties and effects of healthcare technologies within their intended use context. This paper describes the adoption of HTA process to assess the adoption of the EOSedge™ system in clinical practice. METHODS: The EOSedge™ system is a digital radiography system that delivers whole-body, high-quality 2D/3D biplanar images covering the complete set of musculoskeletal and orthopedic exams. Full HTA model was chosen using the EUnetHTA Core Model® version 3.0. The HTA Core Model organizes the information into nine domains. Information was researched and obtained by consulting the manufacturers' user manuals, scientific literature, and institutional sites for regulatory aspects. RESULTS: All nine domains of the EUnetHTA Core Model® helped conduct the HTA of the EOSedge, including (1) description and technical characteristics of the technology; (2) health problem and current clinical practice; (3) safety; (4) clinical effectiveness; (5) organizational aspects; (6) economic evaluation; (7) impact on the patient; (8) ethical aspects; and (9) legal aspects. CONCLUSIONS: EOS technologies may be a viable alternative to conventional radiographs. EOSedge has the same intended use and similar indications for use, technological characteristics, and operation principles as the EOS System and provides significant dose reduction factors for whole spine imaging compared to the EOS System without compromising image quality. Regarding the impact of EOS imaging on patient outcomes, most studies aim to establish technical ability without evaluating their ability to improve patient outcomes; thus, more studies on this aspect are warranted.

DXA beyond bone mineral density and the REMS technique: new insights for current radiologists practice.

Osteoporosis is the most prevalent skeletal disorder, a condition that is associated with significant social and healthcare burden. In the elderly, osteoporosis is commonly associated with sarcopenia, further increasing the risk of fracture. Several imaging techniques are available for a non-invasive evaluation of osteoporosis and sarcopenia. This review focuses on dual-energy X-ray absorptiometry (DXA), as this technique offers the possibility to evaluate bone mineral density and body composition parameters with good precision and accuracy. DXA is also able to evaluate the amount of aortic calcification for cardiovascular risk estimation. Additionally, new DXA-based parameters have been developed in recent years to further refine fracture risk estimation, such as the Trabecular Bone Score and the Bone Strain Index. Finally, we describe the recent advances of a newly developed ultrasound-based technology known as Radiofrequency Echographic Multi-Spectrometry, which represent the latest non-ionizing approach for osteoporosis evaluation at central sites.

The use of artificial intelligence in musculoskeletal ultrasound: a systematic review of the literature.

PURPOSE: To systematically review the use of artificial intelligence (AI) in musculoskeletal (MSK) ultrasound (US) with an emphasis on AI algorithm categories and validation strategies. MATERIAL AND METHODS: An electronic literature search was conducted for articles published up to January 2024. Inclusion criteria were the use of AI in MSK US, involvement of humans, English language, and ethics committee approval. RESULTS: Out of 269 identified papers, 16 studies published between 2020 and 2023 were included. The research was aimed at predicting diagnosis and/or segmentation in a total of 11 (69%) out of 16 studies. A total of 11 (69%) studies used deep learning (DL)-based algorithms, three (19%) studies employed conventional machine learning (ML)-based algorithms, and two (12%) studies employed both conventional ML- and DL-based algorithms. Six (38%) studies used cross-validation techniques with K-fold cross-validation being the most frequently employed (n = 4, 25%). Clinical validation with separate internal test datasets was reported in nine (56%) papers. No external clinical validation was reported. CONCLUSION: AI is a topic of increasing interest in MSK US research. In future studies, attention should be paid to the use of validation strategies, particularly regarding independent clinical validation performed on external datasets.

Shear-wave elastography for the evaluation of tendinopathies: a systematic review and meta-analysis.

PURPOSE: To compare pathologic and healthy tendons using shear-wave elastography (SWE). METHODS: A systematic review with meta-analysis was done searching Pubmed and EMBASE up to September 2022. Prospective, retrospective and cross-sectional studies that used SWE in the assessment of pathologic tendons versus control were included. Our primary outcome were SWE velocity (m/s) and stiffness (kPa). Methodological quality was assessed by the methodological index for non-randomized studies (MINORS). We used the mean difference (MD) with corresponding 95% confidence intervals (CIs) to quantify effects between groups. We performed sensitivity analysis in case of high heterogeneity, after excluding poor quality studies according to MINORS assessment. We used Grades of Recommendation, Assessment, Development and Evaluation to evaluate the certainty of evidence (CoE). RESULTS: Overall, 16 studies with 676 pathologic tendons (188 Achilles, 142 patellar, 96 supraspinatus, 250 mixed) and 723 control tendons (484 healthy; 239 contralateral tendon) were included. Five studies (31.3%) were judged as poor methodological quality. Shear-wave velocity and stiffness meta-analyses showed high heterogeneity. According to a sensitivity analysis, pathologic tendons had a lower shear wave velocity (MD of - 1.69 m/s; 95% CI 1.85; - 1.52; n = 274; I2 50%) compared to healthy tendons with very low CoE. Sensitivity analysis on stiffness still showed high heterogeneity. CONCLUSION: Pathological tendons may have reduced SWE velocity compared to controls, but the evidence is very uncertain. Future robust high-quality longitudinal studies and clear technical indications on the use of this tool are needed. PROTOCOL: PROSPERO identifier: CRD42023405410 CLINICAL RELEVANCE STATEMENT: SWE is a relatively recent modality that may increase sensitivity and diagnostic accuracy of conventional ultrasound imaging promoting early detection of tendinopathy. Non-negligible heterogeneity has been observed in included studies, so our findings may encourage the conduct of future high-quality longitudinal studies which can provide clear technical indications on the use of this promising tool in tendon imaging.

Dual-energy CT in musculoskeletal imaging: technical considerations and clinical applications.

Dual-energy CT stands out as a robust and innovative imaging modality, which has shown impressive advancements and increasing applications in musculoskeletal imaging. It allows to obtain detailed images with novel insights that were once the exclusive prerogative of magnetic resonance imaging. Attenuation data obtained by using different energy spectra enable to provide unique information about tissue characterization in addition to the well-established strengths of CT in the evaluation of bony structures. To understand clearly the potential of this imaging modality, radiologists must be aware of the technical complexity of this imaging tool, the different ways to acquire images and the several algorithms that can be applied in daily clinical practice and for research. Concerning musculoskeletal imaging, dual-energy CT has gained more and more space for evaluating crystal arthropathy, bone marrow edema, and soft tissue structures, including tendons and ligaments. This article aims to analyze and discuss the role of dual-energy CT in musculoskeletal imaging, exploring technical aspects, applications and clinical implications and possible perspectives of this technique.

Sleep Characteristics, Magnetic Resonance Imaging- and Dual-Energy X-Ray Absorptiometry-Based Morphological Parameters, Muscle Strength, and Risk of Falls in Sedentary Older Subjects.

BACKGROUND: The aims of this cross-sectional study were to (a) assess actigraphy-based sleep parameters (total sample and gender differences), (b) assess differences in morphological parameters and physical function between short- versus normal-sleepers and poor- versus good-sleepers, and (c) assess the possible correlations between sleep variables and morphological and physical function parameters in older subjects. METHODS: This study enrolled 42 healthy older participants (60-80 years). Participants completed the following clinical evaluations: (1) whole-body dual-energy X-ray absorptiometry to assess the appendicular skeletal muscle mass index; (2) magnetic resonance imaging acquisition to determine the cross-sectional muscle area of thigh muscles and intermuscular adipose tissue; (3) risk of fall assessment through the mini-Balance Evaluation Systems Test; (4) strength assessment: (a) chair stand test and (b) handgrip strength test; (5) sleep monitoring by actigraphy to assess total sleep time, sleep efficiency, wake after sleep onset, sleep latency, fragmentation index, mobile time, and subjective sleep quality. RESULTS: 31.0% of subjects were short-sleepers (total sleep time < 6 hr), 19.1% were poor-sleepers (sleep efficiency < 85%), and gender differences were detected in mobile time (males: 15.8 ± 6.0 and females: 13.4 ± 6.8; p < .001) and fragmentation index (males: 35.3 ± 14.3 and females: 29.6 ± 14.6; p < .001); no significant differences were observed between groups (short- vs. normal-sleepers and poor- vs. good-sleepers) in morphological and physical function variables; correlation analysis showed that sleep latency negatively correlated with Mini-Balance Evaluation Systems Test (r = -.352; p = .022) and a positive correlation was detected between cross-sectional muscle area and mobile time (r = .349, p = .023). CONCLUSION: No differences were observed in morphological and function parameters between good- versus poor-sleepers, those subjects with worse sleep onset latency (i.e., longer time to fall asleep) registered higher for risk of fall. The potential role of sleep in the physiological mechanisms of muscular aging must be explored through cross-sectional cohort studies with a larger population.

Artificial intelligence for radiographic imaging detection of caries lesions: a systematic review.

BACKGROUND: The aim of this systematic review is to evaluate the diagnostic performance of Artificial Intelligence (AI) models designed for the detection of caries lesion (CL). MATERIALS AND METHODS: An electronic literature search was conducted on PubMed, Web of Science, SCOPUS, LILACS and Embase databases for retrospective, prospective and cross-sectional studies published until January 2023, using the following keywords: artificial intelligence (AI), machine learning (ML), deep learning (DL), artificial neural networks (ANN), convolutional neural networks (CNN), deep convolutional neural networks (DCNN), radiology, detection, diagnosis and dental caries (DC). The quality assessment was performed using the guidelines of QUADAS-2. RESULTS: Twenty articles that met the selection criteria were evaluated. Five studies were performed on periapical radiographs, nine on bitewings, and six on orthopantomography. The number of imaging examinations included ranged from 15 to 2900. Four studies investigated ANN models, fifteen CNN models, and two DCNN models. Twelve were retrospective studies, six cross-sectional and two prospective. The following diagnostic performance was achieved in detecting CL: sensitivity from 0.44 to 0.86, specificity from 0.85 to 0.98, precision from 0.50 to 0.94, PPV (Positive Predictive Value) 0.86, NPV (Negative Predictive Value) 0.95, accuracy from 0.73 to 0.98, area under the curve (AUC) from 0.84 to 0.98, intersection over union of 0.3-0.4 and 0.78, Dice coefficient 0.66 and 0.88, F1-score from 0.64 to 0.92. According to the QUADAS-2 evaluation, most studies exhibited a low risk of bias. CONCLUSION: AI-based models have demonstrated good diagnostic performance, potentially being an important aid in CL detection. Some limitations of these studies are related to the size and heterogeneity of the datasets. Future studies need to rely on comparable, large, and clinically meaningful datasets. PROTOCOL: PROSPERO identifier: CRD42023470708.

DXA-based bone strain index in normocalcemic primary hyperparathyroidism.

The trabecular and cortical bone assessed by bone strain index seems not to be significantly affected in NHPT. INTRODUCTION: The natural history and bone involvement of normocalcemic hyperparathyroidism (NHPT) are not fully clarified yet. The bone strain index (BSI) is a deformation index based on the finite element method and can be applied to DXA scans. In this study, we aim to assess BSI in subjects with NHPT. METHOD: A case-control study included 170 subjects: 40 subjects with NHPT, 50 subjects with primary hypercalcemic hyperparathyroidism (PHPT), and 80 controls (age- and sex-matched with the NPTH group). RESULTS: Lumbar spine (LS) bone mineral density (BMD), femoral neck (FN) BMD, total hip (TH) BMD, and TBS were similar between NHPT and both PHPT and controls. FN-BSI was lower in NHPT compared to PHPT (1.52 ± 0.31 vs 1.72 ± 0.42 p = 0.031) while there were no differences between NHPT and controls. TH-BSI was lower in NHPT compared to PHPT (1.36 ± 0.23 vs 1.52 ± 0.34, p = 0.030), while there were no differences between NHPT and controls. LS-BSI was not different between NHPT and both PHPT and controls. CONCLUSION: The trabecular and cortical bones assessed by BSI seem not to be significantly impaired in NHPT. Further prospective studies are needed to confirm these findings and to give an insight into the natural history of NHPT to improve knowledge and management of this condition.

Imaging of Anatomical Variants Around the Knee.

Several anatomical variants have been described in the knee. These variants may involve intra- and extra-articular structures, such as menisci, ligaments, plicae, bony structures, muscles, and tendons. They have a variable prevalence, are generally asymptomatic, and are usually discovered incidentally in knee magnetic resonance imaging examinations. A thorough knowledge of these findings is essential to avoid overestimating and overinvestigating normal findings. This article reviews most anatomical variants around the knee, describing how to avoid misinterpretation.

Micro-fragmented adipose tissue (mFAT) associated with arthroscopic debridement provides functional improvement in knee osteoarthritis: a randomized controlled trial.

PURPOSE: Current conservative treatments for knee OA provide limited benefits, with symptoms relief for a short amount of time. Regenerative medicine approaches such as the use of microfragmented adipose tissue (mFAT) showed promising results in terms of durable effects and the possibility to enhance tissue healing and counteract the progression of the pathology. Nevertheless, up to today, the large part of clinical data about mFAT use refers to uncontrolled studies, especially in the surgical setting. The purpose of this study was to evaluate the effectiveness of mFAT applied in association with arthroscopic debridement (AD) for the treatment of knee OA, in terms of symptoms relief and tissue healing. METHODS: This study is a prospective, randomized controlled clinical trial. 78 patients affected by knee OA grade 3-4 according to KL classification were randomly assigned to AD or AD + mFAT treatment groups. Clinical, radiological and serological assessments were performed at 6 months after treatment. Additional clinical evaluation was performed at the end of the study with an average follow-up of 26.1 ± 9.5 months. VAS, KOOS, WOMAC and SF-12 were also collected at both timepoints, KSS only at 6 months. RESULTS: Treatment with AD + mFAT improved functional scores at both 6 months (KOOS-PS: + 11.7 ± 20.2 vs + 24.4 ± 22.5, in AD and AD + mFAT, respectively, p = 0.024; KSS: + 14.9 ± 15.9 vs + 24.8 ± 23.5, in AD and AD + mFAT, respectively, p = 0.046) and 24-month follow-ups (KOOS-PS Functional subscale: - 2.0 ± 3.5 vs - 4.7 ± 4.2, in AD and AD + mFAT, respectively, p = 0.012). Lower T2-mapping scores were obtained in AD + mFAT-treated group in medial and lateral condyle compartments (p < 0.001). Slight increase was observed in the levels of a serum biomarker of cartilage deposition (PIIINP) in both groups at 6-month follow-up (p = 0.037). CONCLUSION: mFAT improves functional outcome and MRI appearance when used in association with AD, therefore supporting its use in the treatment of knee OA in an arthroscopic setting.

MRI-based artificial intelligence to predict infection following total hip arthroplasty failure.

PURPOSE: To investigate whether artificial intelligence (AI) can differentiate septic from non-septic total hip arthroplasty (THA) failure based on preoperative MRI features. MATERIALS AND METHODS: We included 173 patients (98 females, age: 67 ± 12 years) subjected to first-time THA revision surgery after preoperative pelvis MRI. We divided the patients into a training/validation/internal testing cohort (n = 117) and a temporally independent external-testing cohort (n = 56). MRI features were used to train, validate and test a machine learning algorithm based on support vector machine (SVM) to predict THA infection on the training-internal validation cohort with a nested fivefold validation approach. Machine learning performance was evaluated on independent data from the external-testing cohort. RESULTS: MRI features were significantly more frequently observed in THA infection (P < 0.001), except bone destruction, periarticular soft-tissue mass, and fibrous membrane (P > 0.005). Considering all MRI features in the training/validation/internal-testing cohort, SVM classifier reached 92% sensitivity, 62% specificity, 79% PPV, 83% NPV, 82% accuracy, and 81% AUC in predicting THA infection, with bone edema, extracapsular edema, and synovitis having been the best predictors. After being tested on the external-testing cohort, the classifier showed 92% sensitivity, 79% specificity, 89% PPV, 83% NPV, 88% accuracy, and 89% AUC in predicting THA infection. SVM classifier showed 81% sensitivity, 76% specificity, 66% PPV, 88% NPV, 80% accuracy, and 74% AUC in predicting THA infection in the training/validation/internal-testing cohort based on the only presence of periprosthetic bone marrow edema on MRI, while it showed 68% sensitivity, 89% specificity, 93% PPV, 60% NPV, 75% accuracy, and 79% AUC in the external-testing cohort. CONCLUSION: AI using SVM classifier showed promising results in predicting THA infection based on MRI features. This model might support radiologists in identifying THA infection.

MRI radiomics-based machine learning for classification of deep-seated lipoma and atypical lipomatous tumor of the extremities.

PURPOSE: To determine diagnostic performance of MRI radiomics-based machine learning for classification of deep-seated lipoma and atypical lipomatous tumor (ALT) of the extremities. MATERIAL AND METHODS: This retrospective study was performed at three tertiary sarcoma centers and included 150 patients with surgically treated and histology-proven lesions. The training-validation cohort consisted of 114 patients from centers 1 and 2 (n = 64 lipoma, n = 50 ALT). The external test cohort consisted of 36 patients from center 3 (n = 24 lipoma, n = 12 ALT). 3D segmentation was manually performed on T1- and T2-weighted MRI. After extraction and selection of radiomic features, three machine learning classifiers were trained and validated using nested fivefold cross-validation. The best-performing classifier according to previous analysis was evaluated and compared to an experienced musculoskeletal radiologist in the external test cohort. RESULTS: Eight features passed feature selection and were incorporated into the machine learning models. After training and validation (74% ROC-AUC), the best-performing classifier (Random Forest) showed 92% sensitivity and 33% specificity in the external test cohort with no statistical difference compared to the radiologist (p = 0.474). CONCLUSION: MRI radiomics-based machine learning may classify deep-seated lipoma and ALT of the extremities with high sensitivity and negative predictive value, thus potentially serving as a non-invasive screening tool to reduce unnecessary referral to tertiary tumor centers.