{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T15:09:03Z","timestamp":1785510543649,"version":"3.56.0"},"reference-count":50,"publisher":"MDPI AG","issue":"3","license":[{"start":{"date-parts":[[2026,2,1]],"date-time":"2026-02-01T00:00:00Z","timestamp":1769904000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Diagnostics"],"abstract":"<jats:p>Background: Osteoporosis (OP) is characterized by reduced bone mineral density and increased fracture risk. Many spinal surgery patients have undiagnosed OP due to the lack of preoperative screening, leading to postoperative complications. Magnetic resonance imaging (MRI), a routine, non-invasive tool for spinal assessment, offers potential for opportunistic OP detection. This study aimed to develop deep learning models to identify OP using lumbar MRI. Methods: We retrospectively enrolled 218 patients (\u226550 years) who underwent both lumbar MRI and dual-energy X-ray absorptiometry (DXA). After segmentation of vertebral bodies from T1- and T2-weighted MRI images, 738 images per sequence were extracted. Separate convolutional neural network (CNN) models were trained for each sequence. Model performance was evaluated using receiver operating characteristic curves and area under the curve (AUC). Results: Among tested classifiers, EfficientNet b4 showed the best performance. For the T1-weighted model, it achieved an AUC of 82%, with a sensitivity of 85% and specificity of 79%. For the T2-weighted model, the AUC was 83%, with a sensitivity of 86% and specificity of 80%. These results were superior to those of InceptionResNet v2 and ResNet-50 for both sequences. Conclusions: The AI models provided reliable OP classification without additional imaging or radiation. AI-based analysis of standard lumbar MRI sequences can accurately identify OP. These models may assist in early detection of undiagnosed OP in surgical candidates, enabling timely treatment and perioperative strategies to improve outcomes and reduce healthcare burden.<\/jats:p>","DOI":"10.3390\/diagnostics16030423","type":"journal-article","created":{"date-parts":[[2026,2,2]],"date-time":"2026-02-02T12:49:44Z","timestamp":1770036584000},"page":"423","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["Lumbar MRI-Based Deep Learning for Osteoporosis Prediction"],"prefix":"10.3390","volume":"16","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-9179-3496","authenticated-orcid":false,"given":"Ue-Cheung","family":"Ho","sequence":"first","affiliation":[{"name":"Division of Neurosurgery, Department of Surgery, National Taiwan University Hospital, Taipei 100, Taiwan"},{"name":"Graduate Institute of Clinical Medicine, College of Medicine, National Taiwan University, Taipei 100, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hsueh-Yi","family":"Lu","sequence":"additional","affiliation":[{"name":"Department of Industrial Engineering and Management, National Yunlin University of Science and Technology, Yunlin 640, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1405-8700","authenticated-orcid":false,"given":"Lu-Ting","family":"Kuo","sequence":"additional","affiliation":[{"name":"Division of Neurosurgery, Department of Surgery, National Taiwan University Hospital, Taipei 100, Taiwan"},{"name":"Division of Neurosurgery, Department of Surgery, National Taiwan University Hospital, Yunlin Branch, Yunlin 640, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2026,2,1]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"1080","DOI":"10.1016\/S0140-6736(21)02646-5","article-title":"Drug therapy for osteoporosis in older adults","volume":"399","author":"Reid","year":"2022","journal-title":"Lancet"},{"key":"ref_2","first-page":"937","article-title":"A systematic review of the outcomes of osteoporotic fracture patients after hospital discharge: Morbidity, subsequent fractures, and mortality","volume":"10","author":"Nazrun","year":"2014","journal-title":"Ther. Clin. Risk Manag."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"364","DOI":"10.1016\/S0140-6736(18)32112-3","article-title":"Osteoporosis","volume":"393","author":"Compston","year":"2019","journal-title":"Lancet"},{"key":"ref_4","first-page":"105","article-title":"The epidemiology of osteoporosis","volume":"133","author":"Clynes","year":"2020","journal-title":"Br. Med. Bull."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"333","DOI":"10.1038\/s41574-020-0339-7","article-title":"A broader strategy for osteoporosis interventions","volume":"16","author":"Reid","year":"2020","journal-title":"Nat. Rev. Endocrinol."},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Glinkowski, W., Narloch, J., Krasuski, K., and \u015aliwczy\u0144ski, A. (2019). The increase of osteoporotic hip fractures and associated one-year mortality in Poland: 2008\u20132015. J. Clin. Med., 8.","DOI":"10.3390\/jcm8091487"},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"203","DOI":"10.1186\/s13018-019-1226-6","article-title":"Mortality after osteoporotic hip fracture: Incidence, trends, and associated factors","volume":"14","author":"Mazzucchelli","year":"2019","journal-title":"J. Orthop. Surg. Res."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"1219","DOI":"10.1007\/s00198-007-0370-8","article-title":"Prevalence of osteoporosis in patients requiring spine surgery: Incidence and significance of osteoporosis in spine disease","volume":"18","author":"Chin","year":"2007","journal-title":"Osteoporos. Int."},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Fan, Z.Q., Yan, X.A., Li, B.F., Shen, E., Xu, X., Wang, H., and Zhuang, Y. (2023). Prevalence of osteoporosis in spinal surgery patients older than 50 years: A systematic review and meta-analysis. PLoS ONE, 18.","DOI":"10.1371\/journal.pone.0286110"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"200","DOI":"10.22603\/ssrr.2022-0198","article-title":"Systematic literature review and meta-analysis on the clinical outcomes of spine surgeries in patients with concurrent osteoporosis","volume":"7","author":"Ogiri","year":"2023","journal-title":"Spine Surg. Relat. Res."},{"key":"ref_11","first-page":"21","article-title":"Image enhancement techniques using highpass and lowpass filters","volume":"109","author":"Makandar","year":"2015","journal-title":"Int. J. Comput. Appl."},{"key":"ref_12","unstructured":"Tan, M., and Le, Q.V. (2019). EfficientNet: Rethinking model scaling for convolutional neural networks. International Conference of Machine Learning, PMLR."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"1586","DOI":"10.1109\/TKDE.2019.2912815","article-title":"Reliable accuracy estimates from k\u2013fold cross validation","volume":"32","author":"Wong","year":"2020","journal-title":"IEEE Trans. Knowl. Data Eng."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"101607","DOI":"10.1016\/j.bonr.2022.101607","article-title":"Dual-energy x-ray absorptiometry (DXA) to measure bone mineral density (BMD) for diagnosis of osteoporosis\u2014Experimental data from artificial vertebrae confirms significant dependence on bone size","volume":"17","author":"Golding","year":"2022","journal-title":"Bone Rep."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"2074","DOI":"10.1007\/s00330-014-3584-0","article-title":"Opportunistic screening for osteoporosis on routine computed tomography? An external validation study","volume":"25","author":"Buckens","year":"2015","journal-title":"Eur. Radiol."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"1387","DOI":"10.1007\/s00198-014-3013-x","article-title":"Reliability and validity of lower extremity computed tomography as a screening tool for osteoporosis","volume":"26","author":"Lee","year":"2015","journal-title":"Osteoporos. Int."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"147","DOI":"10.1007\/s00198-015-3224-9","article-title":"Effect of IV contrast on lumbar trabecular attenuation at routine abdominal CT: Correlation with DXA and implications for opportunistic osteoporosis screening","volume":"27","author":"Pickhardt","year":"2016","journal-title":"Osteoporos. Int."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"316","DOI":"10.1016\/S0001-2998(87)80024-7","article-title":"Quantitative computed tomography in assessment of osteoporosis","volume":"17","author":"Genant","year":"1987","journal-title":"Semin. Nucl. Med."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"1859","DOI":"10.1007\/s00198-020-05670-z","article-title":"Is dual-energy absorptiometry accurate in the assessment of bone status of patients with chronic kidney disease?","volume":"32","author":"Kim","year":"2021","journal-title":"Osteoporos. Int."},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Kinsella, S., Murphy, K., Breen, M., O\u2019Neill, S., McLaughlin, P., Coyle, J., Bogue, C., O\u2019Neill, F., Moore, N., and McGarrigle, A. (2015). Comparison of single CT scan assessment of bone mineral density, vascular calcification and fat mass with standard clinical measurements in renal transplant subjects: The ABC HeART study. BMC Nephrol., 16.","DOI":"10.1186\/s12882-015-0182-6"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"537","DOI":"10.1016\/j.rcl.2022.02.003","article-title":"Osteoporosis imaging","volume":"60","author":"Martel","year":"2022","journal-title":"Radiol. Clin. N. Am."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"1866","DOI":"10.1016\/j.spinee.2022.06.003","article-title":"Patient physiology influences the MRI-based vertebral bone quality score","volume":"22","author":"Aynaszyan","year":"2022","journal-title":"Spine J."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"2951","DOI":"10.1007\/s00330-015-3699-y","article-title":"A new diagnostic score to detect osteoporosis in patients undergoing lumbar spine MRI","volume":"25","author":"Bandirali","year":"2015","journal-title":"Eur. Radiol."},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"1005","DOI":"10.17219\/acem\/94150","article-title":"Relationship between quantitative parameters of lumbar vertebral perfusion and bone mineral density (BMD) in postmenopausal women","volume":"28","author":"Huang","year":"2019","journal-title":"Adv. Clin. Exp. Med."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"974","DOI":"10.1002\/jmri.22489","article-title":"Quantification of vertebral bone marrow fat content using 3 Tesla MR spectroscopy: Reproducibility, vertebral variation, and applications in osteoporosis","volume":"33","author":"Li","year":"2011","journal-title":"J. Magn. Reson. Imaging"},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"279","DOI":"10.1002\/jmri.20367","article-title":"Osteoporosis is associated with increased marrow fat content and decreased marrow fat unsaturation: A proton MR spectroscopy study","volume":"22","author":"Yeung","year":"2005","journal-title":"J. Magn. Reson. Imaging"},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"1265","DOI":"10.1007\/s00198-019-04924-9","article-title":"Texture analysis of vertebral bone marrow using chemical shift encoding-based water-fat MRI: A feasibility study","volume":"30","author":"Burian","year":"2019","journal-title":"Osteoporos. Int."},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"304","DOI":"10.1016\/j.jmr.2010.09.013","article-title":"Qualitative and quantitative ultrashort echo time (UTE) imaging of cortical bone","volume":"207","author":"Du","year":"2010","journal-title":"J. Magn. Reson."},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"1739","DOI":"10.1002\/mrm.27531","article-title":"On the sensitivity of quantitative susceptibility mapping for measuring trabecular bone density","volume":"81","author":"Diefenbach","year":"2019","journal-title":"Magn. Reson. Med."},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"106726","DOI":"10.1016\/j.compbiomed.2023.106726","article-title":"A review of deep learning-based multiple-lesion recognition from medical images: Classification, detection and segmentation","volume":"157","author":"Jiang","year":"2023","journal-title":"Comput. Biol. Med."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"61","DOI":"10.1016\/j.patrec.2020.07.042","article-title":"Comparative analysis of image classification algorithms based on traditional machine learning and deep learning","volume":"141","author":"Wang","year":"2021","journal-title":"Pattern Recognit. Lett."},{"key":"ref_32","doi-asserted-by":"crossref","unstructured":"AlSaeed, D., and Omar, S.F. (2022). Brain MRI analysis for Alzheimer\u2019s disease diagnosis using CNN-based feature extraction and machine learning. Sensors, 22.","DOI":"10.3390\/s22082911"},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"101645","DOI":"10.1016\/j.nicl.2018.101645","article-title":"Alzheimer\u2019s Disease Neuroimaging Initiative. Automated classification of Alzheimer\u2019s disease and mild cognitive impairment using a single MRI and deep neural networks","volume":"21","author":"Basaia","year":"2019","journal-title":"NeuroImage Clin."},{"key":"ref_34","doi-asserted-by":"crossref","unstructured":"Ou Yang, W.Y., Lai, C.C., Tsou, M.T., and Hwang, L.C. (2021). Development of machine learning models for prediction of osteoporosis from clinical health examination data. Int. J. Environ. Res. Public Health, 18.","DOI":"10.3390\/ijerph18147635"},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"e162","DOI":"10.3346\/jkms.2023.38.e162","article-title":"A prediction model for osteoporosis risk using a machine-learning approach and its validation in a large cohort","volume":"38","author":"Wu","year":"2023","journal-title":"J. Korean Med. Sci."},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"126","DOI":"10.1007\/s00223-020-00700-8","article-title":"Age-related increases in marrow fat volumes have regional impacts on bone cell numbers and structure","volume":"107","author":"Chen","year":"2020","journal-title":"Calcif. Tissue Int."},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"165","DOI":"10.1023\/A:1011513223894","article-title":"Adipocyte tissue volume in bone marrow is increased with aging and in patients with osteoporosis","volume":"2","author":"Justesen","year":"2001","journal-title":"Biogerontology"},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"945","DOI":"10.1148\/radiol.2363041425","article-title":"Vertebral bone mineral density, marrow perfusion, and fat content in healthy men and men with osteoporosis: Dynamic contrast-enhanced MR imaging and MR spectroscopy","volume":"236","author":"Griffith","year":"2005","journal-title":"Radiology"},{"key":"ref_39","doi-asserted-by":"crossref","first-page":"831","DOI":"10.1148\/radiol.2413051858","article-title":"Vertebral marrow fat content and diffusion and perfusion indexes in women with varying bone density: MR evaluation","volume":"241","author":"Griffith","year":"2006","journal-title":"Radiology"},{"key":"ref_40","doi-asserted-by":"crossref","first-page":"317","DOI":"10.1111\/j.1365-2796.2012.02564.x","article-title":"New insights into osteoporosis: The bone-fat connection","volume":"272","author":"Kawai","year":"2012","journal-title":"J. Intern. Med."},{"key":"ref_41","doi-asserted-by":"crossref","first-page":"414","DOI":"10.1210\/jcem.82.2.3735","article-title":"Body weight, body fat distribution, and hormonal replacement therapy in early postmenopausal women","volume":"82","author":"Gambacciani","year":"1997","journal-title":"J. Clin. Endocrinol. Metab."},{"key":"ref_42","first-page":"633","article-title":"Correlation between quantitative and semiquantitative magnetic resonance imaging and histopathology findings in dermatomyositis","volume":"37","author":"Milisenda","year":"2019","journal-title":"Clin. Exp. Rheumatol."},{"key":"ref_43","doi-asserted-by":"crossref","first-page":"245","DOI":"10.1016\/j.ejrad.2004.01.021","article-title":"Assessment of the skeletal status by MR relaxometry techniques of the lumbar spine: Comparison with dual X-ray absorptiometry","volume":"50","author":"Maris","year":"2004","journal-title":"Eur. J. Radiol."},{"key":"ref_44","doi-asserted-by":"crossref","first-page":"631","DOI":"10.1148\/radiology.196.3.7644622","article-title":"Osteoporosis: Clinical assessment with quantitative MR imaging in diagnosis","volume":"196","author":"Wehrli","year":"1995","journal-title":"Radiology"},{"key":"ref_45","doi-asserted-by":"crossref","first-page":"535","DOI":"10.1007\/s00256-018-3057-7","article-title":"Quantification of shoulder muscle intramuscular fatty infiltration on T1-weighted MRI: A viable alternative to the Goutallier classification system","volume":"48","author":"Davis","year":"2019","journal-title":"Skelet. Radiol."},{"key":"ref_46","doi-asserted-by":"crossref","first-page":"1483","DOI":"10.1148\/rg.2017170056","article-title":"CT texture analysis: Definitions, applications, biologic correlates, and challenges","volume":"37","author":"Lubner","year":"2017","journal-title":"RadioGraphics"},{"key":"ref_47","doi-asserted-by":"crossref","first-page":"18330","DOI":"10.1038\/s41598-022-23184-y","article-title":"Artificial intelligence used to diagnose osteoporosis from risk factors in clinical data and proposing sports protocols","volume":"12","author":"Fasihi","year":"2022","journal-title":"Sci. Rep."},{"key":"ref_48","doi-asserted-by":"crossref","first-page":"819","DOI":"10.1007\/s00198-016-3828-8","article-title":"Clinical fracture risk evaluated by hierarchical agglomerative clustering","volume":"28","author":"Kruse","year":"2017","journal-title":"Osteoporos. Int."},{"key":"ref_49","doi-asserted-by":"crossref","first-page":"169","DOI":"10.1007\/s11657-020-00802-8","article-title":"Application of machine learning approaches for osteoporosis risk prediction in postmenopausal women","volume":"15","author":"Shim","year":"2020","journal-title":"Arch. Osteoporos."},{"key":"ref_50","doi-asserted-by":"crossref","first-page":"100125","DOI":"10.1016\/j.health.2022.100125","article-title":"A performance analysis of dimensionality reduction algorithms in machine learning models for cancer prediction","volume":"3","author":"Kabir","year":"2023","journal-title":"Healthc. Anal."}],"container-title":["Diagnostics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2075-4418\/16\/3\/423\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,2]],"date-time":"2026-02-02T13:01:00Z","timestamp":1770037260000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2075-4418\/16\/3\/423"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,2,1]]},"references-count":50,"journal-issue":{"issue":"3","published-online":{"date-parts":[[2026,2]]}},"alternative-id":["diagnostics16030423"],"URL":"https:\/\/doi.org\/10.3390\/diagnostics16030423","relation":{},"ISSN":["2075-4418"],"issn-type":[{"value":"2075-4418","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,2,1]]}}}