Automated Vertebral Segmentation and Measurement of Vertebral Compression Ratio Based on Deep Learning in X-Ray Images

AbstractVertebral compression fracture is a deformity of vertebral bodies found on lateral spine images. To diagnose vertebral compression fracture, accurate measurement of vertebral compression ratio is required. Therefore, rapid and accurate segmentation of vertebra is important for measuring the vertebral compression ratio. In this study, we used 339 data of lateral thoracic and lumbar vertebra images for training and testing a deep learning model for segmentation. The result of segmentation by the model was compared with the manual measurement, which is performed by a specialist. As a result, the average sensitivity of the dataset was 0.937, specificity was 0.995, accuracy was 0.992, and dice similarity coefficient was 0.929, area under the curve of receiver operating characteristic curve was 0.987, and the precision recall curve was 0.916. The result of correlation analysis shows no statistical difference between the manually measured vertebral compression ratio and the vertebral compression ratio using the data segmented by the model in which the correlation coefficient was 0.929. In addition, the Bland–Altman plot shows good equivalence in which VCR values are in the area within average ± 1.96. In conclusion, vertebra segmentation based on deep learning is expected to be helpful for the measurement of vertebral compression ratio.

Download Full-text

Vertebral compression fracture rate following stereotactic ablative body radiotherapy for spine oligometastases: a UK experience

Journal of Radiotherapy in Practice ◽

10.1017/s1460396921000224 ◽

2021 ◽

pp. 1-5

Author(s):

Sameed Hussain ◽

Anjali Zarkar ◽

Ahmed Elmodir ◽

Daniel Ford ◽

Sundus Yahya ◽

...

Keyword(s):

Vertebral Compression Fracture ◽

Demographic Data ◽

Spinal Metastases ◽

Compression Fracture ◽

Spinal Instability ◽

Fracture Rate ◽

Strict Adherence ◽

Stereotactic Ablative Body Radiotherapy ◽

Vertebral Bodies ◽

Pre Treatment

Abstract Aim: Stereotactic ablative body radiotherapy (SABR) for spine metastases is associated with a risk of vertebral compression fracture (VCF). The aim of this study was to determine the rate of VCF at one UK institution and evaluate the use of the Spinal Instability Neoplastic Score (SINS) to predict these. Materials and methods: A retrospective analysis of all patients who underwent SABR for spinal metastases between 2014 and 2018 at one UK institution was performed. Basic demographic data were collected, and SINS prior to SABR was calculated. The primary outcome was VCF rate. Secondary outcomes included time to VCF and need for surgical intervention following VCF. Results: A total of 48 oligometastases were treated with a median follow-up of 20·5 months. A maximum of two vertebral bodies were treated. The median baseline SINS was calculated as 3. The median dose was 26 Gy in three fractions. Two patients were reported to have VCF and both were successfully conservatively managed. Findings: SABR for spine oligometastases is being performed safely with low VCF rates which are comparable with those in international publications. This may be as a result of strict adherence to criteria for delivery of SABR with low pre-treatment SINS.

Download Full-text

An Electronic Component Recognition Algorithm Based on Deep Learning with a Faster SqueezeNet

Mathematical Problems in Engineering ◽

10.1155/2020/2940286 ◽

2020 ◽

Vol 2020 ◽

pp. 1-11

Author(s):

Yuanyuan Xu ◽

Genke Yang ◽

Jiliang Luo ◽

Jianan He

Keyword(s):

Deep Learning ◽

Characteristic Curve ◽

Area Under The Curve ◽

Principal Component ◽

Recognition Algorithm ◽

Electronic Component ◽

Support Vector ◽

Learning Networks ◽

Component Recognition ◽

And Performance

Electronic component recognition plays an important role in industrial production, electronic manufacturing, and testing. In order to address the problem of the low recognition recall and accuracy of traditional image recognition technologies (such as principal component analysis (PCA) and support vector machine (SVM)), this paper selects multiple deep learning networks for testing and optimizes the SqueezeNet network. The paper then presents an electronic component recognition algorithm based on the Faster SqueezeNet network. This structure can reduce the size of network parameters and computational complexity without deteriorating the performance of the network. The results show that the proposed algorithm performs well, where the Receiver Operating Characteristic Curve (ROC) and Area Under the Curve (AUC), capacitor and inductor, reach 1.0. When the FPR is less than or equal 10 − 6 level, the TPR is greater than or equal to 0.99; its reasoning time is about 2.67 ms, achieving the industrial application level in terms of time consumption and performance.

Download Full-text

Fusion High-Resolution Network for Diagnosing ChestX-ray Images

Electronics ◽

10.3390/electronics9010190 ◽

2020 ◽

Vol 9 (1) ◽

pp. 190 ◽

Cited By ~ 3

Author(s):

Zhiwei Huang ◽

Jinzhao Lin ◽

Liming Xu ◽

Huiqian Wang ◽

Tong Bai ◽

...

Keyword(s):

Neural Networks ◽

Deep Learning ◽

High Resolution ◽

Convolutional Neural Networks ◽

Characteristic Curve ◽

Area Under The Curve ◽

Disease Classification ◽

Deep Convolutional Neural Networks ◽

Learning Methods ◽

Global And Local

The application of deep convolutional neural networks (CNN) in the field of medical image processing has attracted extensive attention and demonstrated remarkable progress. An increasing number of deep learning methods have been devoted to classifying ChestX-ray (CXR) images, and most of the existing deep learning methods are based on classic pretrained models, trained by global ChestX-ray images. In this paper, we are interested in diagnosing ChestX-ray images using our proposed Fusion High-Resolution Network (FHRNet). The FHRNet concatenates the global average pooling layers of the global and local feature extractors—it consists of three branch convolutional neural networks and is fine-tuned for thorax disease classification. Compared with the results of other available methods, our experimental results showed that the proposed model yields a better disease classification performance for the ChestX-ray 14 dataset, according to the receiver operating characteristic curve and area-under-the-curve score. An ablation study further confirmed the effectiveness of the global and local branch networks in improving the classification accuracy of thorax diseases.

Download Full-text

Evaluation of Risk Factors for Vertebral Compression Fracture after Carbon-Ion Radiotherapy for Primary Spinal and Paraspinal Sarcoma

BioMed Research International ◽

10.1155/2017/9467402 ◽

2017 ◽

Vol 2017 ◽

pp. 1-7 ◽

Cited By ~ 3

Author(s):

Yoshihiro Matsumoto ◽

Makoto Shinoto ◽

Makoto Endo ◽

Nokitaka Setsu ◽

Keiichiro Iida ◽

...

Keyword(s):

Risk Factors ◽

Operating Characteristic ◽

Vertebral Compression Fracture ◽

Characteristic Curve ◽

Compression Fracture ◽

Carbon Ion Radiotherapy ◽

Vertebral Compression Fractures ◽

Spinal Instability ◽

Carbon Ion ◽

Spine Surgeon

Background and Purpose. Carbon-ion radiotherapy (C-ion RT) was effective therapy for inoperable spinal and paraspinal sarcomas. However, a significant adverse event following radiotherapies is vertebral compression fractures (VCFs). In this study, we investigated the incidence of and risk factors for post-C-ion RT VCFs in patients with spinal or paraspinal sarcomas. Material and Methods. Thirty consecutive patients with spinal or paraspinal sarcomas treated with C-ion RT were retrospectively reviewed. Various clinical parameters and the Spinal Instability Neoplastic Score (SINS) were used to evaluate the risk factors for post-C-ion RT VCFs. Results. The overall incidence of VCFs was 23% (median time: 7 months). Patients with VCFs showed a markedly higher SINS score (median value, 9 points) than those without VCF (5 points). The area under the receiver operating characteristic curve for the SINS score was 0.88, and the optimum SINS cut-off score was 8 points. The cumulative incidence of VCFs at 1 year was 9% for patients with a SINS score under 8 points, versus 80% for those with a SINS score of 8 points or higher (p<0.0001). Conclusions. In patients with a SINS score of 8 points or higher, referral to a spine surgeon for stabilization and multidisciplinary discussion is appropriate.

Download Full-text

Radiological and clinical effect of balloon kyphoplasty for the management of rheumatoid arthritis patients with osteoporotic vertebral compression fracture

10.21203/rs.3.rs-47606/v1 ◽

2020 ◽

Author(s):

Ji Guo ◽

Weifeng Zhai ◽

Licheng Wei ◽

Jianpo Zhang ◽

Lang Jin ◽

...

Keyword(s):

Rheumatoid Arthritis ◽

Vertebral Compression Fracture ◽

Compression Fracture ◽

Balloon Kyphoplasty ◽

Cement Leakage ◽

Osteoporotic Vertebral Compression Fracture ◽

Control Group ◽

Compression Rate ◽

Kyphotic Angle ◽

Vertebral Bodies

Abstract Objective: This study was conducted to investigate the outcome of percutaneous balloon kyphoplasty (KP) for the treatment of osteoporotic vertebral compression fracture(OVCF) in patients with rheumatoid arthritis (RA) and analyze the influence of erythrocyte sedimentation rate (ESR), c-reactive protein (CRP), injected cement volume and duration of taking glucocorticoid on the outcome of KP procedure. Methods: A total of 39 RA patients (63 vertebral bodies) and 38 patients (50 vertebral bodies) without RA received KP management for OVCF. Changes in vertebral compression rate, local kyphotic angle, conditions of bone cement leakage, visual analogue scale (VAS) and Oswestry disability index (ODI) scores were evaluated for radiological and clinical outcomes of KP procedure. In addition, 39 OVCF patients with RA were divided into different groups according to the value of ESR, CRP, injected cement volume and duration of taking glucocorticoid to evaluate their influence on the outcomes of KP procedure.Results: The KP procedure significantly improved the compression rate, local kyphotic angle, VAS and ODI scores in both RA group and control group. The compression rate increased 11.56±3.8% in RA group which is significantly larger than the control group(p<0.05). The change of local kyphotic angle in RA group was 3.77±1.9, which is also larger than that in control group(p<0.05). Whereas, the changes of VAS and ODI scores were not significantly different between the two groups. Besides, radiological and clinical outcomes were not significantly different among the groups of different ESR, CRP, injected cement volume and duration of taking glucocorticoid no matter before or 1 year after the KP procedure, but 44% RA patients who take glucocorticoid for over 10 years had cement leakage after the KP procedure which is significantly higher than the group of RA patients with less than 10 years glucocorticoid use(p<0.05). In addition, 7 intradiscal cement leakage occurred in patients take glucocorticoid over 10 years where as no intradiscal leakage showed up in its control group(p<0.01).Conclusion: KP procedure was effective for OVCF patients with or without RA, for restoring vertebral body height, reducing local kyphotic angle, relieving pain and recovering spinal function. Compared to the control group, RA patients received more improvement in compression rate and local kyphotic angle after the operation. Intradiscal leakage occurred more in patients who take glucocorticoid for over 10 years.

Download Full-text

Efficacy of Postural Reduction in Osteoporotic Vertebral Compression Fractures Followed by Percutaneous Vertebroplasty

Neurosurgery ◽

10.1227/01.neu.0000204313.36531.79 ◽

2006 ◽

Vol 58 (4) ◽

pp. 695-700 ◽

Cited By ~ 32

Author(s):

Dong-Kyu Chin ◽

Young-Soo Kim ◽

Yong-Eun Cho ◽

Jun-Jae Shin

Keyword(s):

Cobb Angle ◽

Compression Ratio ◽

Vertebral Compression Fracture ◽

Percutaneous Vertebroplasty ◽

Compression Fracture ◽

Vertebral Compression Fractures ◽

Significant Relation ◽

Osteoporotic Vertebral Fractures ◽

The Mean ◽

Postural Reduction

Abstract OBJECTIVE: Vertebroplasty in the symptomatic osteoporotic vertebral fracture has become increasingly popular. However, there have been some limitations in restoring the height of the collapsed vertebrae and in preventing the leaking of cement. In the severely collapsed vertebrae of more than two thirds of their original height, vertebroplasty is regarded as a contraindication. We tried postural reduction using a soft pillow under the compressed level. This study was undertaken to investigate the effectiveness of the combination of postural reduction and vertebroplasty for re-expansion and stabilization of the osteoporotic vertebral fractures. METHODS: A total of 75 patients with single level vertebral compression fracture were treated with postural reduction followed by vertebroplasty. In 30 patients, the vertebral body was severely collapsed more than two-thirds of its original height. We calculated the compression ratio (anterior height/posterior height) and measured the Cobb angle. We analyzed the degree of re-expansion according to the onset duration. RESULTS: The mean compression ratio was 0.60 ± 0.15 initially and increased to 0.75 ± 0.17 after vertebroplasty. The mean Cobb angle was 16.14 ± 11.29° and corrected to 10.71 ± 12.08°. The degree of re-expansion showed significant relation with the onset duration. Twenty-eight of 30 (93%) severely collapsed vertebrae re-expanded after postural reduction, which made vertebroplasty possible. CONCLUSION: This new method of vertebroplasty leads to significant restoration of height and correction of kyphosis. The re-expansion was closely related with onset duration. In cases of severely collapsed vertebrae which is able to be re-expanded by postural reduction, vertebroplasty could be applied safely.

Download Full-text

Percutaneous Vertebroplasty: Efficacy of Unipedicular Vertebroplasty as Compared to Bipedicular Vertebroplasty

Indian Journal of Radiology and Imaging ◽

10.1055/s-0041-1739375 ◽

2022 ◽

Author(s):

Mahendra Kumar Dwivedi ◽

Vikrant Bhende ◽

Dnyaneshwar Narayanrao Panchbhaiyye ◽

Madhura Vijay Bayaskar

Keyword(s):

Vertebral Compression Fracture ◽

Percutaneous Vertebroplasty ◽

Compression Fracture ◽

P Value ◽

Significant Difference ◽

Insignificant Difference ◽

Vas Scores ◽

Vertebral Bodies ◽

The Mean ◽

Standard Techniques

Abstract Introduction Percutaneous vertebroplasty has been used for treatment of intractable painful fractures of vertebral bodies. With the help of refined procedures and standard techniques, the interventional radiologist can now offer help to orthopedics and neurosurgeons in these cases, which include treatment of vertebral compression fracture. Vertebroplasty is aimed at reducing the pain induced by collapse. Vertebroplasty is the standard mode of treatment for vertebral collapse, and in our study, bipedicular vertebroplasty was compared with unipedicular approach as bipedicular vertebroplasty is the routinely used approach. Aim To compare efficacy of unipedicular percutaneous vertebroplasty with that of bipedicular percutaneous vertebroplasty. Material and Methods A total of 52 vertebroplasties were done over a period of 2 years. Out of 52 patients, 28 patients underwent unipedicular vertebroplasty and 24 patients underwent bipedicular vertebroplasty. Visual analogue scale (VAS) scores were used to assess the pain prior to vertebroplasty and after vertebroplasty. Efficacy of the two procedures were assessed by comparing VAS scores. Results There was no statistically significant difference observed in the preprocedure and postprocedure VAS scores (p-value < 0.0001, < 0.0001, respectively). The mean procedure time was lesser in unipedicular vertebroplasty (41.9 ± 3.90) than bipedicular vertebroplasty (54.5 ± 3.4). Conclusion Unipedicular vertebroplasty is as effective as bipedicular vertebroplasty, as there is insignificant difference in postprocedure VAS scores between the unipedicular and bipedicular vertebroplasty.

Download Full-text

Do the dynamic stress mobility radiographs predict the postoperative vertebral height restoration, kyphosis correction, and cement volume injected after vertebroplasty for osteoporotic thoracolumbar vertebral fractures with intravertebral cleft?

Journal of Orthopaedic Surgery ◽

10.1177/2309499018806700 ◽

2018 ◽

Vol 26 (3) ◽

pp. 230949901880670 ◽

Cited By ~ 1

Author(s):

Chee Kidd Chiu ◽

Kulathunga Arachchige Lisitha ◽

Dahlia Munchar Elias ◽

Voon Wei Yong ◽

Chris Yin Wei Chan ◽

...

Keyword(s):

Vertebral Compression Fracture ◽

Compression Fracture ◽

Vertebral Height ◽

Kyphotic Angle ◽

Lateral Spine ◽

Stress Radiographs ◽

Kyphosis Correction ◽

Dynamic Mobility ◽

Spine Radiograph ◽

The Mean

Background: This prospective clinical–radiological study was conducted to determine whether the dynamic mobility stress radiographs can predict the postoperative vertebral height restoration, kyphosis correction, and cement volume injected after vertebroplasty. Methods: Patients included had the diagnosis of significant back pain caused by osteoporotic vertebral compression fracture secondary to trivial injury. All the patients underwent routine preoperative sitting lateral spine radiograph, supine stress lateral spine radiograph, and supine anteroposterior spine radiograph. The radiological parameters recorded were anterior vertebral height (AVH), middle vertebral height (MVH), posterior vertebral height (PVH), MVH level below, wedge endplate angle (WEPA), and regional kyphotic angle (RKA). The supine stress versus sitting difference (SSD) for all the above parameters were calculated. Results: A total of 28 patients (4 males; 24 females) with the mean age of 75.6 ± 7.7 years were recruited into this study. The mean cement volume injected was 5.5 ± 1.8 ml. There was no difference between supine stress and postoperative radiographs for AVH ( p = 0.507), PVH ( p = 0.913) and WEPA ( p = 0.379). The MVH ( p = 0.026) and RKA ( p = 0.005) were significantly less in the supine stress radiographs compared to postoperative radiographs. There was significant correlation ( p < 0.05) between supine stress and postoperative AVH, MVH, PVH, WEPA, and RKA. The SSD for AVH, PVH, WEPA, and RKA did not have significant correlation with the cement volume ( p > 0.05). Only the SSD-MVH had significant correlation with cement volume, but the correlation was weak ( r = 0.39, p = 0.04). Conclusions: Dynamic mobility stress radiographs can predict the postoperative vertebral height restoration and kyphosis correction after vertebroplasty for thoracolumbar osteoporotic fracture with intravertebral clefts. However, it did not reliably predict the amount of cement volume injected as it was affected by other factors.

Download Full-text

Fully automated pelvic bone segmentation in multiparameteric MRI using a 3D convolutional neural network

Insights into Imaging ◽

10.1186/s13244-021-01044-z ◽

2021 ◽

Vol 12 (1) ◽

Author(s):

Xiang Liu ◽

Chao Han ◽

He Wang ◽

Jingyun Wu ◽

Yingpu Cui ◽

...

Keyword(s):

Neural Network ◽

Deep Learning ◽

Convolutional Neural Network ◽

External Validation ◽

Lumbar Vertebra ◽

Qualitative Evaluation ◽

Initial Step ◽

Pelvic Bone ◽

Dice Similarity Coefficient ◽

Bone Segmentation

Abstract Background Accurate segmentation of pelvic bones is an initial step to achieve accurate detection and localisation of pelvic bone metastases. This study presents a deep learning-based approach for automated segmentation of normal pelvic bony structures in multiparametric magnetic resonance imaging (mpMRI) using a 3D convolutional neural network (CNN). Methods This retrospective study included 264 pelvic mpMRI data obtained between 2018 and 2019. The manual annotations of pelvic bony structures (which included lumbar vertebra, sacrococcyx, ilium, acetabulum, femoral head, femoral neck, ischium, and pubis) on diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC) images were used to create reference standards. A 3D U-Net CNN was employed for automatic pelvic bone segmentation. Additionally, 60 mpMRI data from 2020 were included and used to evaluate the model externally. Results The CNN achieved a high Dice similarity coefficient (DSC) average in both testing (0.80 [DWI images] and 0.85 [ADC images]) and external (0.79 [DWI images] and 0.84 [ADC images]) validation sets. Pelvic bone volumes measured with manual and CNN-predicted segmentations were highly correlated (R2 value of 0.84–0.97) and in close agreement (mean bias of 2.6–4.5 cm3). A SCORE system was designed to qualitatively evaluate the model for which both testing and external validation sets achieved high scores in terms of both qualitative evaluation and concordance between two readers (ICC = 0.904; 95% confidence interval: 0.871–0.929). Conclusions A deep learning-based method can achieve automated pelvic bone segmentation on DWI and ADC images with suitable quantitative and qualitative performance.

Download Full-text