Osteoporosis and Sarcopenia最新文献

{"title":"Can artificial intelligence-derived bone mineral density outperform traditional morphometric measurements in opportunistic osteoporosis screening using pelvic radiographs? A multicenter comparative diagnostic study","authors":"Chia-Che Lee,&nbsp;Xi-Qian Heng,&nbsp;Hung-Kuan Yen,&nbsp;I-Hsin Chen,&nbsp;Tzu-Hao Tseng,&nbsp;Hong-Da Lu,&nbsp;Hung-Chun Lee,&nbsp;Ruey-Feng Chang,&nbsp;Chen-Yu Wang,&nbsp;Shau-Huai Fu","doi":"10.1016/j.afos.2026.09.001","DOIUrl":"10.1016/j.afos.2026.09.001","url":null,"abstract":"<div><h3>Objectives</h3><div>To investigate whether an artificial intelligence (AI) model could demonstrate high diagnostic accuracy in identifying dual-energy X-ray absorptiometry (DXA) confirmed osteoporosis, outperform traditional human-read metrics, and deliver substantial clinical net benefit.</div></div><div><h3>Methods</h3><div>This retrospective multicenter cohort study included 382 patients with hip fractures who had both preoperative pelvic radiographs and postoperative DXA scans. An AI model analyzed the intact contralateral proximal femur to predict both bone mineral density (BMD) values and T-scores. Diagnostic performance was evaluated against DXA-confirmed osteoporosis (T-score ≤−2.5) and compared with manual cortical thickness index (CTI) and femoral neck width (FNW) measurements using Spearman rank-order correlation analysis, receiver operating characteristic (ROC) curves, the DeLong test, and decision curve analysis (DCA).</div></div><div><h3>Results</h3><div>The AI model showed a strong correlation with DXA-derived BMD values for both the femoral neck (<em>ρ</em> = 0.76, R² = 0.55, mean absolute error (MAE) = 0.06 g/cm²) and the total hip (<em>ρ</em> = 0.69, R² = 0.51, MAE = 0.08 g/cm²). Correspondingly, the AI-predicted T-scores correlated well with DXA-derived T-scores for the femoral neck (<em>ρ</em> = 0.76, R² = 0.55, MAE = 0.47) and the total hip (<em>ρ</em> = 0.69, R² = 0.51, MAE = 0.65). The AI model demonstrated excellent discriminative performance, achieving an area under the ROC curve (AUC) of 0.88 (95% confidence interval [CI] 0.83–0.93) for the femoral neck and 0.87 (95% CI 0.82–0.92) for the total hip, both significantly higher than CTI (0.72 [95% CI 0.65–0.79]; P &lt; 0.001) and FNW (0.53 [95% CI 0.44–0.62]; P &lt; 0.001). DCA illustrated that the AI models provided greater clinical net benefit than manual measurements, remaining clinically advantageous up to a threshold probability exceeding 0.90.</div></div><div><h3>Conclusions</h3><div>Deep learning AI offers a superior, automated alternative to traditional manual morphometric measurements for identifying osteoporosis on routine pelvic radiographs in patients with hip fractures, potentially facilitating timely secondary fracture prevention.</div></div>","PeriodicalId":19701,"journal":{"name":"Osteoporosis and Sarcopenia","volume":"12 3","pages":"Pages 140-149"},"PeriodicalIF":2.4,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149056401","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"From underdiagnosis to scalable screening: Applications of artificial intelligence for vertebral compression fracture detection – a systematic review and meta-analysis of diagnostic performance","authors":"Hsuan-Yu Chen,&nbsp;Linjun Yang,&nbsp;Gwen Wilson,&nbsp;Milos Brkljac,&nbsp;Monty S. Khela,&nbsp;Kellen L. Mulford,&nbsp;Cody C. Wyles","doi":"10.1016/j.afos.2026.08.001","DOIUrl":"10.1016/j.afos.2026.08.001","url":null,"abstract":"<div><h3>Objectives</h3><div>Vertebral compression fractures (VCFs) are common and often missed on routine imaging. We systematically reviewed the diagnostic performance of artificial intelligence (AI) for VCF detection.</div></div><div><h3>Methods</h3><div>A systematic review and meta-analysis were conducted following PRISMA 2020 and PROSPERO registration by searching Ovid Medline, Embase, and Central from January 1, 2020, to November 18, 2025, for studies using AI models for VCF detection and diagnosis. Risk of bias was assessed with QUADAS-2. Studies with complete 2 × 2 data were pooled using random-effects meta-analysis, with subgroup analyses by modality.</div></div><div><h3>Results</h3><div>Fifty-one studies were included. Thirty-seven contributed to meta-analysis with around 259 190 participants and 45,346 VCF patients. Original radiology reports missed &gt; 50% of incidental VCFs in many cohorts (up to 81%). The overall sensitivity of AI-assisted VCF detection was 85.0% (95% CI 84.2%–85.7%) and specificity 94.6% (95% CI 92.8%–95.9%). The positive and negative likelihood ratios were 15.4 and 0.16. At 22% pre-test probability, post-test probability was 77.7% after a positive AI result and 4.1% after a negative result. CT-based AI models performed best (sensitivity: 88.0%, specificity: 97.5%), while X-ray-based AI models showed lower sensitivity (79.2%) with high specificity (95.1%).</div></div><div><h3>Conclusions</h3><div>AI models demonstrate high diagnostic accuracy for VCF detection, particularly in CT imaging. However, improved detection alone does not ensure better outcomes, as substantial treatment gaps persist. The clinical impact of AI depends on rigorous prospective external validation and integration into structured osteoporosis care pathways, linking automated identification to timely evaluation and secondary fracture prevention.</div></div>","PeriodicalId":19701,"journal":{"name":"Osteoporosis and Sarcopenia","volume":"12 3","pages":"Pages 3-15"},"PeriodicalIF":2.4,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149028129","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Osteoporosis screening and management framework for community-based primary care in Hong Kong: A Delphi consensus study","authors":"Ching-Lung Cheung,&nbsp;Hing-Sun Chan,&nbsp;Pak-Kin Chan,&nbsp;Alvina Wing-Sze Chau,&nbsp;Cindy Ming-Wah Choi,&nbsp;Dicky Ka-Chun Chow,&nbsp;Dilys Yan-Wing Chow,&nbsp;Hoi-Yan Chung,&nbsp;Bonnie So-Kuen Lam,&nbsp;Wing-Wo Lam,&nbsp;Bonnie Leung,&nbsp;Yuen-Lok Ma,&nbsp;Janus Chi-Keung Wong,&nbsp;Sze-Hung Wong,&nbsp;Kenny Wai-Ming Wong,&nbsp;William Ho-Ching Yuen","doi":"10.1016/j.afos.2026.08.005","DOIUrl":"10.1016/j.afos.2026.08.005","url":null,"abstract":"<div><h3>Objectives</h3><div>Osteoporosis poses a significant public health challenge as it leads to an increased incidence of fractures. This can result in considerable disability and substantial financial burdens. With the rapidly aging population in Hong Kong, the incidence and impact of osteoporotic fractures are expected to rise, underscoring the need for a coordinated, community-based care model to identify at-risk older adults and offer timely management.</div></div><div><h3>Methods</h3><div>This study adopted a Delphi methodology to develop consensus statements and a structured framework for community osteoporosis screening and management in Hong Kong. A multidisciplinary panel of 15 experts with experience in osteoporosis care in the primary care setting participated in three iterative rounds of surveys that collected anonymized ratings and feedback.</div></div><div><h3>Results</h3><div>A total of 31 consensus statements were established covering risk assessment, diagnosis, referral, and the care pathway of osteoporosis in the primary care setting. The statements provide practical recommendations for a systematic framework that streamlines workflow and delineates roles and responsibilities for primary care professionals and carers. This guidance is intended for primary care services, including community pharmacies, district health centers, and elderly service centers engaged in osteoporosis screening and management of individuals diagnosed with the condition.</div></div><div><h3>Conclusions</h3><div>The consensus recommendations advocate for the development of a sustainable, community-driven service framework that promotes routine risk assessment, timely initiation of treatment and coordinated long-term management. Implementation of this model has the potential to improve patient outcomes, reduce fracture-related morbidity, and alleviate the overall healthcare burden of osteoporosis in Hong Kong.</div></div>","PeriodicalId":19701,"journal":{"name":"Osteoporosis and Sarcopenia","volume":"12 3","pages":"Pages 122-131"},"PeriodicalIF":2.4,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149056399","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Upward changes in serum 25-hydroxyvitamin D levels in South Korea: the Korea National Health and Nutrition Examination Survey studies IX (2022–2023) versus VI (2013–2014)","authors":"Ju-Hyun Park,&nbsp;Na-Hyeon Lee,&nbsp;Yoon-Sok Chung,&nbsp;Choon Hee Chung,&nbsp;Han Seok Choi","doi":"10.1016/j.afos.2026.08.002","DOIUrl":"10.1016/j.afos.2026.08.002","url":null,"abstract":"<div><h3>Objectives</h3><div>To compare serum 25-hydroxyvitamin D (25[OH]D) concentrations and the prevalence of vitamin D deficiency between the Korea National Health and Nutrition Examination Survey (KNHANES) VI and IX, and to evaluate demographic, seasonal, and lifestyle factors associated with vitamin D status in South Korea.</div></div><div><h3>Methods</h3><div>Data from KNHANES IX (2022–2023), in which serum 25(OH)D concentrations were measured by liquid chromatography–tandem mass spectrometry, were compared with data from KNHANES VI (2013–2014), in which radioimmunoassay was used. Information on dietary supplement intake (≥2 weeks per year) was obtained through a standardized questionnaire. Vitamin D status was defined according to serum 25(OH)D concentrations as deficiency (&lt;20 ng/mL), adequacy (≥20 and &lt;50 ng/mL), and excess (≥50 ng/mL).</div></div><div><h3>Results</h3><div>Serum 25(OH)D levels significantly increased across both sexes and all age groups. The odds of vitamin D deficiency were significantly lower in KNHANES IX compared to VI, with adjusted odds ratios of 0.358 for men and 0.214 for women. The proportion of subjects taking dietary supplements increased significantly from 43.2% to 66.3%. Body mass index was significantly associated with higher odds of vitamin D deficiency exclusively among women in KNHANES IX. Winter remained persistently associated with higher odds of deficiency.</div></div><div><h3>Conclusions</h3><div>Over the past decade, South Korea has experienced an upward shift in population-level vitamin D concentrations. These changes may be associated with increased intake of dietary supplements, but because direct standardization of analytical methods has not been achieved, the relative contributions of behavioral changes and analytical method-related differences cannot be quantified.</div></div>","PeriodicalId":19701,"journal":{"name":"Osteoporosis and Sarcopenia","volume":"12 3","pages":"Pages 132-139"},"PeriodicalIF":2.4,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149056400","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning model for simultaneous vertebral level localization and fracture detection on DXA vertebral fracture assessment images","authors":"Sang Wouk Cho,&nbsp;Dongin Moon,&nbsp;Chang Oh Kim,&nbsp;Hyeon Chang Kim,&nbsp;Yumie Rhee,&nbsp;Namki Hong","doi":"10.1016/j.afos.2026.09.002","DOIUrl":"10.1016/j.afos.2026.09.002","url":null,"abstract":"<div><h3>Objectives</h3><div>While vertebral fracture assessment (VFA) on dual-energy X-ray absorptiometry (DXA) images is widely used to detect osteoporotic vertebral fractures, existing deep learning approaches typically perform either fracture detection or vertebral localization, not both. We developed VERTE-X L, a deep learning model that simultaneously localizes vertebrae (T7–L5) and detects fractures on DXA VFA images, and evaluated its performance in a community-dwelling older adult cohort.</div></div><div><h3>Methods</h3><div>We used DXA VFA images from 1000 randomly sampled participants (age ≥65 years) in the Korean Urban Rural Elderly (KURE) cohort, split into training (N = 800) and test (N = 200) sets. VERTE-X L employs a DINOv2-L encoder, a localization decoder for per-vertebra coordinate prediction, and a consistent rank logits (CORAL) ordinal regression module for fracture detection. Performance was evaluated at image and vertebra levels and compared with previously reported detection-only models (VERTE-X and VERTE-X model fine-tuned on VFA data) and localization-only models (VerteNet and Multi-Context Hybrid CNN Transformer).</div></div><div><h3>Results</h3><div>On the held-out test set, VERTE-X L achieved an image-level area under the receiver operating characteristic curve (AUROC) of 0.952 (sensitivity 89.7%, specificity 94.2%, net present value [NPV] 98.2%) and vertebra-level AUROC of 0.957 (NPV 99.7%). Localization success (within 15 pixels) was 94.7% for fracture vertebrae. VERTE-X L outperformed VERTE-X in detection (AUROC 0.952 vs 0.913, P = 0.034) and outperformed VerteNet in localization (mean radial error 10.0 vs. 20.3 px, P &lt; 0.001).</div></div><div><h3>Conclusions</h3><div>The proposed method enables simultaneous fracture detection and vertebral localization on DXA VFA images.</div></div>","PeriodicalId":19701,"journal":{"name":"Osteoporosis and Sarcopenia","volume":"12 3","pages":"Pages 16-22"},"PeriodicalIF":2.4,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149028130","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Fall risk in rheumatoid arthritis revisited: A follow-up study focusing on sarcopenia and glucocorticoid use","authors":"Naoki Nakagawa,&nbsp;Keita Nagira,&nbsp;Masako Hayashibara,&nbsp;Takashi Wada,&nbsp;Yuki Kitsuda,&nbsp;Yuta Takasu,&nbsp;Chikako Takeda,&nbsp;Mari Osaki,&nbsp;Hiroshi Hagino,&nbsp;Hideki Nagashima","doi":"10.1016/j.afos.2026.09.003","DOIUrl":"10.1016/j.afos.2026.09.003","url":null,"abstract":"<div><h3>Objectives</h3><div>To re-evaluate fall incidence and its determinants in a contemporary cohort of ambulatory women with rheumatoid arthritis (RA) in a follow-up study based on our previously published cohort data.</div></div><div><h3>Methods</h3><div>We prospectively evaluated 124 ambulatory female patients aged ≥50 years with RA at a single institution in 2022–2023, monitoring falls for 1 year. Outcomes included patient characteristics, sarcopenia, disease activity, and physical function. Multivariable logistic regression and Firth-penalized logistic regression were used in sensitivity analysis to assess the interaction between sarcopenia and glucocorticoid (GC) use.</div></div><div><h3>Results</h3><div>The annual fall rate in the 2022–2023 cohort was lower than in our previous cohort (27.4% vs. 50.0%). History of falls (odds ratio [OR] 2.76, 95% confidence interval [CI] 1.02–7.45, P = 0.045) and hypnotics/anxiolytics use (OR 3.92, 95% CI 1.01–15.20, P = 0.048) were independent risk factors for falls. A significant interaction between sarcopenia and GC use was observed in the initial analysis but was no longer statistically significant in sensitivity analyses.</div></div><div><h3>Conclusions</h3><div>Compared with our previously reported cohort, the contemporary RA cohort showed a lower fall rate and a different pattern of associated risk factors. Further research is needed to explore the potential interaction between sarcopenia and glucocorticoid use. The observed interaction should be considered exploratory.</div></div>","PeriodicalId":19701,"journal":{"name":"Osteoporosis and Sarcopenia","volume":"12 3","pages":"Pages 150-155"},"PeriodicalIF":2.4,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149056397","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Consensus recommendations on age-related sarcopenia in South Asia: A nutrition-focused approach","authors":"Bhakti Samant,&nbsp;Anita Jatana,&nbsp;Shivshankar Timmanpyati,&nbsp;Archana Sinha,&nbsp;Amreen Sheikh,&nbsp;Daphnee Lovesley,&nbsp;Datta Patel,&nbsp;Kiran Dalal,&nbsp;Ritika Samaddar,&nbsp;Saima Faruqui,&nbsp;Sheetal Ghali,&nbsp;Shifali Mehraa,&nbsp;Simmi Shah,&nbsp;Amal Firouse,&nbsp;Atul Mishra","doi":"10.1016/j.afos.2026.08.003","DOIUrl":"10.1016/j.afos.2026.08.003","url":null,"abstract":"<div><div>Age-related sarcopenia, defined as the progressive decline in muscle function, strength, and mass with age, is a growing health concern in South Asian countries (Bangladesh, India, and Sri Lanka) due to their rapidly aging populations. It tends to appear earlier, is frequently overlooked, and worsens after age 60. Despite its prevalence, standardized tools for screening and assessing sarcopenia and dietary status are lacking. Since diet quality is a modifiable risk factor, nutritional strategies play a key role in prevention and management. However, international dietary guidelines are often unsuitable for South Asians due to regional dietary differences and variations within countries. As sarcopenia onset occurs earlier in this population, there is an urgent need to implement standardized screening, diagnosis, and nutrition assessments during dietetic consultations. To address this gap, a panel of experienced dietitians from Bangladesh, India, and Sri Lanka developed this South Asian consensus to provide evidence-based recommendations for identifying and managing age-related sarcopenia. The panel emphasized the use of locally available, affordable, and nutritionally adequate foods to support practical, region-specific interventions.</div></div>","PeriodicalId":19701,"journal":{"name":"Osteoporosis and Sarcopenia","volume":"12 3","pages":"Pages 105-121"},"PeriodicalIF":2.4,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149056386","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning–based quantification of muscle volume and attenuation and their associations with preinjury ambulatory status in patients with hip fractures","authors":"Tomoki Asano,&nbsp;Keisuke Uemura,&nbsp;Sotaro Kono,&nbsp;Ryo Higuchi,&nbsp;Yi Gu,&nbsp;Hirokazu Mae,&nbsp;Tetsuro Tani,&nbsp;Kazuma Takashima,&nbsp;Tatsuhiko Kutsuna,&nbsp;Yoshito Otake,&nbsp;Masaki Takao,&nbsp;Seiji Okada,&nbsp;Hidetoshi Hamada","doi":"10.1016/j.afos.2026.05.001","DOIUrl":"10.1016/j.afos.2026.05.001","url":null,"abstract":"<div><h3>Objectives</h3><div>Use a deep learning model on computerized tomography images to explore how muscle volume and attenuation relate to preinjury ambulatory levels in patients with intertrochanteric or femoral neck fractures.</div></div><div><h3>Methods</h3><div>This retrospective study involved 170 female patients with intertrochanteric or femoral neck fractures who had preoperative hip computed tomography (CT) scans. Using a deep learning–based model, muscles on the unfractured side were automatically segmented, and the muscle volume and attenuation were measured for four groups: gluteal muscles, hip adductors, quadriceps, and hamstrings. Preinjury walking ability was categorized into three groups based on the Koval index: Group 1 (Koval index 1–3; walking outdoors), Group 2 (Koval index 4–6; walking indoors), and Group 3 (Koval index 7; chairbound or bedridden). Finally, associations between preinjury ambulatory status and muscle parameters, patients’ age, body mass index (BMI), hip bone mineral density, geriatric nutritional risk index, and Charlson comorbidity index (CCI) were analyzed using ordinal logistic regression models.</div></div><div><h3>Results</h3><div>Ambulatory status was associated with muscle attenuation, CCI, and BMI across all four muscle groups. Other factors did not reach statistical significance. Patients with poorer ambulatory function had lower muscle attenuation, higher CCI, and lower BMI.</div></div><div><h3>Conclusions</h3><div>Muscle attenuation on CT images correlated with preinjury ambulatory status, whereas muscle volume did not. This indicates that evaluating muscle quality via CT may more accurately reflect functional decline in older adults than assessing muscle quantity alone.</div></div>","PeriodicalId":19701,"journal":{"name":"Osteoporosis and Sarcopenia","volume":"12 3","pages":"Pages 40-47"},"PeriodicalIF":2.4,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149028133","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"From detection to action: Transparency, accuracy, and integration for clinical artificial intelligence in musculoskeletal aging","authors":"Ta-Wei Tai,&nbsp;Dung Van Tran","doi":"10.1016/j.afos.2026.08.004","DOIUrl":"10.1016/j.afos.2026.08.004","url":null,"abstract":"","PeriodicalId":19701,"journal":{"name":"Osteoporosis and Sarcopenia","volume":"12 3","pages":"Pages 1-2"},"PeriodicalIF":2.4,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149028760","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Federated learning for privacy-preserving opportunistic screening of lumbar-spine DXA-defined osteoporotic-level bone mineral density using chest radiographs: A multi-center independent validation study","authors":"Bo-Ru Lin,&nbsp;Ming-Yueh Huang,&nbsp;Wei-Chung Sun,&nbsp;Po-Wei Wu,&nbsp;Paul Chang,&nbsp;Ethan Tu,&nbsp;Wei-Chieh Hung","doi":"10.1016/j.afos.2026.06.001","DOIUrl":"10.1016/j.afos.2026.06.001","url":null,"abstract":"<div><h3>Objectives</h3><div>To evaluate the performance of an artificial intelligence (AI) system trained via federated learning for opportunistic screening of lumbar-spine DXA-defined osteoporotic-level bone mineral density (T-score ≤ −2.5) using routine chest X-rays (CXRs) in a multi-center independent validation setting.</div></div><div><h3>Methods</h3><div>A deep learning model was developed via privacy-preserving federated learning on 5943 paired CXR and dual-energy X-ray absorptiometry (DXA) examinations and validated on an independent dual-center cohort of 3007 adults undergoing same-day CXR and DXA. Ground truth was a lumbar spine DXA T-score ≤ −2.5. The primary metric was AUROC, alongside sensitivity and specificity. Image-only and clinical-only ablation models were evaluated on the same cohort to disentangle imaging from demographic contributions.</div></div><div><h3>Results</h3><div>Among 3007 validation cases, 200 (6.7%) had a T-score ≤ −2.5. The AI achieved an AUROC of 0.942 (95% CI, 0.923–0.962), with sensitivity of 91.0% and specificity of 82.7%, and performance was preserved across study sites, sexes, and the clinically relevant older stratum (≥60 years). In ablation analysis, the image-only model achieved AUROC 0.931 (95% CI 0.912–0.950) and the clinical-only baseline 0.780 (0.732–0.827); non-overlapping 95% CIs (image-only higher) support an independent imaging contribution beyond demographics.</div></div><div><h3>Conclusions</h3><div>This federated learning-based AI demonstrates high discrimination for identifying lumbar-spine DXA-defined osteoporotic-level bone mineral density (T-score ≤ −2.5) from standard CXRs and shows temporal transportability across two institutions without centralizing raw imaging data. Positioned as a triage adjunct to confirmatory DXA, it offers a scalable, privacy-preserving opportunistic screening solution.</div></div>","PeriodicalId":19701,"journal":{"name":"Osteoporosis and Sarcopenia","volume":"12 3","pages":"Pages 23-31"},"PeriodicalIF":2.4,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149028131","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}