Prediction of subsequent osteoporotic vertebral compression fracture on CT radiography via deep learning
Abstract Combination of computed tomography (CT) radiography and deep learning to predict subsequent osteoporotic vertebral compression fracture (OVCF) has not been reported. To do so, we analyzed retrospectively CT images from 103 patients who experienced twice OVCF in Tongji Hospital from 2011 to...
| Main Authors: | , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2022-12-01
|
| Series: | View |
| Subjects: | |
| Online Access: | https://doi.org/10.1002/VIW.20220012 |
| _version_ | 1811197713280139264 |
|---|---|
| author | Xiao Hu Yanjing Zhu Yadong Qian Ruiqi Huang Shuai Yin Zhili Zeng Ning Xie Bin Ma Yan Yu Qing Zhao Zhourui Wu Jianjie Wang Wei Xu Yilong Ren Chen Li Rongrong Zhu Liming Cheng |
| author_facet | Xiao Hu Yanjing Zhu Yadong Qian Ruiqi Huang Shuai Yin Zhili Zeng Ning Xie Bin Ma Yan Yu Qing Zhao Zhourui Wu Jianjie Wang Wei Xu Yilong Ren Chen Li Rongrong Zhu Liming Cheng |
| author_sort | Xiao Hu |
| collection | DOAJ |
| description | Abstract Combination of computed tomography (CT) radiography and deep learning to predict subsequent osteoporotic vertebral compression fracture (OVCF) has not been reported. To do so, we analyzed retrospectively CT images from 103 patients who experienced twice OVCF in Tongji Hospital from 2011 to 2022. Meanwhile, CT images from 70 age‐matched osteoporotic patients without vertebral fracture were used as the negative control. Convolutional neural network was used for classification and the Adam optimizer combining the momentum and exponentially weighted moving average gradients methods were used to update the weights of the networks. In the prediction model, we split 80% data of each type of the patient as the training group, while the other 20% was held as the independent testing group. We found that the number of subsequent fracture in women is higher than that in men (81 vs. 22). Additionally, the incidence rate of adjacent vertebral fracture is higher than that of remote vertebral fracture (64.1 vs. 35.9%), while the onset time of the former was 11.9 ± 12.8 months, significantly less than 22.3 ± 18.2 months of the latter (p < .001). For the prediction of subsequent fracture, our model attained .839 of accuracy and .883 of receiver operating characteristic–area under curve on the whole testing dataset. Furthermore, our model gained .867 and .719 of accuracy on the single‐class testing dataset separated from the former, .817 of accuracy on the independent test. In conclusion, we managed to generate a deep learning‐based model, which is able to predict subsequent OVCF in a precise and unbiased way just using CT images. |
| first_indexed | 2024-04-12T01:18:56Z |
| format | Article |
| id | doaj.art-a33c4a14df324fcb9154cafb05b758a0 |
| institution | Directory Open Access Journal |
| issn | 2688-3988 2688-268X |
| language | English |
| last_indexed | 2024-04-12T01:18:56Z |
| publishDate | 2022-12-01 |
| publisher | Wiley |
| record_format | Article |
| series | View |
| spelling | doaj.art-a33c4a14df324fcb9154cafb05b758a02022-12-22T03:53:52ZengWileyView2688-39882688-268X2022-12-0136n/an/a10.1002/VIW.20220012Prediction of subsequent osteoporotic vertebral compression fracture on CT radiography via deep learningXiao Hu0Yanjing Zhu1Yadong Qian2Ruiqi Huang3Shuai Yin4Zhili Zeng5Ning Xie6Bin Ma7Yan Yu8Qing Zhao9Zhourui Wu10Jianjie Wang11Wei Xu12Yilong Ren13Chen Li14Rongrong Zhu15Liming Cheng16Key Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaDivision of Spine Department of Orthopaedics Tongji Hospital Tongji University School of Medicine Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaKey Laboratory of Spine and Spinal Cord Injury Repair and Regeneration of Ministry of Education, Orthopaedic Department of Tongji Hospital, School of Medicine, School of Life Sciences and Technology Tongji University Shanghai ChinaAbstract Combination of computed tomography (CT) radiography and deep learning to predict subsequent osteoporotic vertebral compression fracture (OVCF) has not been reported. To do so, we analyzed retrospectively CT images from 103 patients who experienced twice OVCF in Tongji Hospital from 2011 to 2022. Meanwhile, CT images from 70 age‐matched osteoporotic patients without vertebral fracture were used as the negative control. Convolutional neural network was used for classification and the Adam optimizer combining the momentum and exponentially weighted moving average gradients methods were used to update the weights of the networks. In the prediction model, we split 80% data of each type of the patient as the training group, while the other 20% was held as the independent testing group. We found that the number of subsequent fracture in women is higher than that in men (81 vs. 22). Additionally, the incidence rate of adjacent vertebral fracture is higher than that of remote vertebral fracture (64.1 vs. 35.9%), while the onset time of the former was 11.9 ± 12.8 months, significantly less than 22.3 ± 18.2 months of the latter (p < .001). For the prediction of subsequent fracture, our model attained .839 of accuracy and .883 of receiver operating characteristic–area under curve on the whole testing dataset. Furthermore, our model gained .867 and .719 of accuracy on the single‐class testing dataset separated from the former, .817 of accuracy on the independent test. In conclusion, we managed to generate a deep learning‐based model, which is able to predict subsequent OVCF in a precise and unbiased way just using CT images.https://doi.org/10.1002/VIW.20220012computed tomographyconvolutional neural networkdeep learningosteoporotic vertebral compression fractureprediction model |
| spellingShingle | Xiao Hu Yanjing Zhu Yadong Qian Ruiqi Huang Shuai Yin Zhili Zeng Ning Xie Bin Ma Yan Yu Qing Zhao Zhourui Wu Jianjie Wang Wei Xu Yilong Ren Chen Li Rongrong Zhu Liming Cheng Prediction of subsequent osteoporotic vertebral compression fracture on CT radiography via deep learning View computed tomography convolutional neural network deep learning osteoporotic vertebral compression fracture prediction model |
| title | Prediction of subsequent osteoporotic vertebral compression fracture on CT radiography via deep learning |
| title_full | Prediction of subsequent osteoporotic vertebral compression fracture on CT radiography via deep learning |
| title_fullStr | Prediction of subsequent osteoporotic vertebral compression fracture on CT radiography via deep learning |
| title_full_unstemmed | Prediction of subsequent osteoporotic vertebral compression fracture on CT radiography via deep learning |
| title_short | Prediction of subsequent osteoporotic vertebral compression fracture on CT radiography via deep learning |
| title_sort | prediction of subsequent osteoporotic vertebral compression fracture on ct radiography via deep learning |
| topic | computed tomography convolutional neural network deep learning osteoporotic vertebral compression fracture prediction model |
| url | https://doi.org/10.1002/VIW.20220012 |
| work_keys_str_mv | AT xiaohu predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT yanjingzhu predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT yadongqian predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT ruiqihuang predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT shuaiyin predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT zhilizeng predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT ningxie predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT binma predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT yanyu predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT qingzhao predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT zhouruiwu predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT jianjiewang predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT weixu predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT yilongren predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT chenli predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT rongrongzhu predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning AT limingcheng predictionofsubsequentosteoporoticvertebralcompressionfractureonctradiographyviadeeplearning |