Namki Hong, Sang Wouk Cho, Young Han Lee, Chang Oh Kim, Hyeon Chang Kim, Yumie Rhee, William D Leslie, Steven R Cummings, Kyoung Min Kim
{"title":"Deep learning-based identification of vertebral fracture and osteoporosis in lateral spine radiographs and DXA VFA to predict incident fracture.","authors":"Namki Hong, Sang Wouk Cho, Young Han Lee, Chang Oh Kim, Hyeon Chang Kim, Yumie Rhee, William D Leslie, Steven R Cummings, Kyoung Min Kim","doi":"10.1093/jbmr/zjaf050","DOIUrl":null,"url":null,"abstract":"<p><p>Deep learning (DL) identification of vertebral fractures and osteoporosis in lateral spine radiographs and DXA vertebral fracture assessment (VFA) images may improve fracture risk assessment in older adults. In 26 299 lateral spine radiographs from 9276 individuals attending a tertiary-level institution (60% train set; 20% validation set; 20% test set; VERTE-X cohort), DL models were developed to detect prevalent vertebral fracture and osteoporosis. The pre-trained DL models from lateral spine radiographs were then fine-tuned in 30% of a DXA VFA dataset (KURE cohort), with performance evaluated in the remaining 70% test set. The area under the receiver operating characteristics curve (AUROC) for DL models to detect prevalent vertebral fracture and osteoporosis was 0.926 (95% CI 0.908-0.955) and 0.848 (95% CI 0.827-0.869) from VERTE-X spine radiographs, respectively, and 0.924 (95% CI 0.905-0.942) and 0.867 (95% CI 0.853-0.881) from KURE DXA VFA images, respectively. A total of 13.3% and 13.6% of individuals sustained an incident fracture during a median follow-up of 5.4 years and 6.4 years in the VERTE-X test set (n = 1852) and KURE test set (n = 2456), respectively. Incident fracture risk was significantly greater among individuals with DL-detected vertebral fracture (hazard ratios [HR] 3.23 [95% CI 2.51-5.17] and 2.11 [95% CI 1.62-2.74] for the VERTE-X and KURE test sets) or DL-detected osteoporosis (HR 2.62 [95% CI 1.90-3.63] and 2.14 [95% CI 1.72-2.66]), which remained significant after adjustment for clinical risk factors and femoral neck BMD. DL scores improved incident fracture discrimination and net benefit when combined with clinical risk factors. In summary, DL-detected prevalent vertebral fracture and osteoporosis in lateral spine radiographs and DXA VFA images enhanced fracture risk prediction in older adults.</p>","PeriodicalId":185,"journal":{"name":"Journal of Bone and Mineral Research","volume":" ","pages":""},"PeriodicalIF":5.1000,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Bone and Mineral Research","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1093/jbmr/zjaf050","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENDOCRINOLOGY & METABOLISM","Score":null,"Total":0}
引用次数: 0
Abstract
Deep learning (DL) identification of vertebral fractures and osteoporosis in lateral spine radiographs and DXA vertebral fracture assessment (VFA) images may improve fracture risk assessment in older adults. In 26 299 lateral spine radiographs from 9276 individuals attending a tertiary-level institution (60% train set; 20% validation set; 20% test set; VERTE-X cohort), DL models were developed to detect prevalent vertebral fracture and osteoporosis. The pre-trained DL models from lateral spine radiographs were then fine-tuned in 30% of a DXA VFA dataset (KURE cohort), with performance evaluated in the remaining 70% test set. The area under the receiver operating characteristics curve (AUROC) for DL models to detect prevalent vertebral fracture and osteoporosis was 0.926 (95% CI 0.908-0.955) and 0.848 (95% CI 0.827-0.869) from VERTE-X spine radiographs, respectively, and 0.924 (95% CI 0.905-0.942) and 0.867 (95% CI 0.853-0.881) from KURE DXA VFA images, respectively. A total of 13.3% and 13.6% of individuals sustained an incident fracture during a median follow-up of 5.4 years and 6.4 years in the VERTE-X test set (n = 1852) and KURE test set (n = 2456), respectively. Incident fracture risk was significantly greater among individuals with DL-detected vertebral fracture (hazard ratios [HR] 3.23 [95% CI 2.51-5.17] and 2.11 [95% CI 1.62-2.74] for the VERTE-X and KURE test sets) or DL-detected osteoporosis (HR 2.62 [95% CI 1.90-3.63] and 2.14 [95% CI 1.72-2.66]), which remained significant after adjustment for clinical risk factors and femoral neck BMD. DL scores improved incident fracture discrimination and net benefit when combined with clinical risk factors. In summary, DL-detected prevalent vertebral fracture and osteoporosis in lateral spine radiographs and DXA VFA images enhanced fracture risk prediction in older adults.
期刊介绍:
The Journal of Bone and Mineral Research (JBMR) publishes highly impactful original manuscripts, reviews, and special articles on basic, translational and clinical investigations relevant to the musculoskeletal system and mineral metabolism. Specifically, the journal is interested in original research on the biology and physiology of skeletal tissues, interdisciplinary research spanning the musculoskeletal and other systems, including but not limited to immunology, hematology, energy metabolism, cancer biology, and neurology, and systems biology topics using large scale “-omics” approaches. The journal welcomes clinical research on the pathophysiology, treatment and prevention of osteoporosis and fractures, as well as sarcopenia, disorders of bone and mineral metabolism, and rare or genetically determined bone diseases.