Our Nations bone health: Prevention and treatment of Cervical hip fracture caused by Osteoporosis

Bayramova, Afa

doi:10.17352/pjor.000013

Cited by 2 publications

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

A Computed Tomography–Based Fracture Prediction Model With Images of Vertebral Bones and Muscles by Employing Deep Learning: Development and Validation Study

Kong,

Cho,

Park

et al. 2024

J Med Internet Res

View full text Add to dashboard Cite

Background: Along with the progressively aging populations, the use of opportunistic computed tomography (CT) scanning is increasing, which could be a valuable method to acquire information on both muscles and bones. Objective:We aimed to develop and externally validate opportunistic CT-based fracture prediction model using images of bones and muscles by employing a deep learning method. Methods:The model was developed based on a retrospective longitudinal cohort study of 1,359 patients with abdominal CT images at Seoul National University Bundang Hospital between 2010 and 2019. The model was externally validated in 495 patients from Seoul National University Boramae Hospital. The primary outcome of the study was the incidence of vertebral fracture events within 5 years of follow-up. The image model was developed using an attention convolutional neural networkrecurrent neural network model for images of the bone and paravertebral muscles. Results:The mean age was 72 years, and 69.4% were females. In the development set, the areas under the receiver operator curve (AUROC) for predicting vertebral fractures were superior in images of the bone and paravertebral muscle than in those of the bone-only (0.736±0.003 vs. 0.688±0.001; p<0.001). In the validation cohort, the AUROC for the images of the bone-only and the images of the bone and paravertebral muscle were 0.698±0.001 and 0.729±0.002, respectively; p<0.001). For the clinical models using age, sex, body mass index, use of steroids, smoking status, and secondary osteoporosis, AUROC values for the developmental and validation cohorts were 0.635±0.002 and 0.698±0.021, respectively, significantly lower than those of the image model using bone and muscle (p<0.001). In conclusion, the deep learning model of the convolutional neural networkrecurrent neural network structure based on CT images of the muscles and bones could better predict the risk of vertebral fractures than the clinical models. Conclusions:The model using the images of bone and muscle showed better performance than that using the images of the boneonly. Opportunistic CT screening may contribute to identifying patients with a high fracture risk in the future.

show abstract

A Computed Tomography–Based Fracture Prediction Model With Images of Vertebral Bones and Muscles by Employing Deep Learning: Development and Validation Study

Kong,

Cho,

Park

et al. 2024

J Med Internet Res

View full text Add to dashboard Cite

show abstract

A Computed Tomography–Based Fracture Prediction Model With Images of Vertebral Bones and Muscles by Employing Deep Learning: Development and Validation Study (Preprint)

Kong¹,

Cho²,

Park³

et al. 2023

Preprint

View full text Add to dashboard Cite

BACKGROUND Along with the progressively aging populations, the use of opportunistic computed tomography (CT) scanning is increasing, which could be a valuable method to acquire information on both muscles and bones. OBJECTIVE We aimed to develop and externally validate opportunistic CT-based fracture prediction model using images of bones and muscles by employing a deep learning method. METHODS The model was developed based on a retrospective longitudinal cohort study of 1,359 patients with abdominal CT images at Seoul National University Bundang Hospital between 2010 and 2019. The model was externally validated in 495 patients from Seoul National University Boramae Hospital. The primary outcome of the study was the incidence of vertebral fracture events within 5 years of follow-up. The image model was developed using an attention convolutional neural network-recurrent neural network model for images of the bone and paravertebral muscles. RESULTS The mean age was 72 years, and 69.4% were females. In the development set, the areas under the receiver operator curve (AUROC) for predicting vertebral fractures were superior in images of the bone and paravertebral muscle than in those of the bone-only (0.736±0.003 vs. 0.688±0.001; p<0.001). In the validation cohort, the AUROC for the images of the bone-only and the images of the bone and paravertebral muscle were 0.698±0.001 and 0.729±0.002, respectively; p<0.001). For the clinical models using age, sex, body mass index, use of steroids, smoking status, and secondary osteoporosis, AUROC values for the developmental and validation cohorts were 0.635±0.002 and 0.698±0.021, respectively, significantly lower than those of the image model using bone and muscle (p<0.001). In conclusion, the deep learning model of the convolutional neural network-recurrent neural network structure based on CT images of the muscles and bones could better predict the risk of vertebral fractures than the clinical models. CONCLUSIONS The model using the images of bone and muscle showed better performance than that using the images of the bone-only. Opportunistic CT screening may contribute to identifying patients with a high fracture risk in the future.

show abstract

Our Nations bone health: Prevention and treatment of Cervical hip fracture caused by Osteoporosis

Cited by 2 publications

References 9 publications

A Computed Tomography–Based Fracture Prediction Model With Images of Vertebral Bones and Muscles by Employing Deep Learning: Development and Validation Study

A Computed Tomography–Based Fracture Prediction Model With Images of Vertebral Bones and Muscles by Employing Deep Learning: Development and Validation Study

A Computed Tomography–Based Fracture Prediction Model With Images of Vertebral Bones and Muscles by Employing Deep Learning: Development and Validation Study (Preprint)

Contact Info

Product

Resources

About