The analysis of MSTMOVCF (Multi-segment thoracolumbar mild osteoporotic fractures surgery or conservative treatment) based on ASTLOF (the assessment system of thoracolumbar osteoporotic fracture)

Exploring deep learning radiomics for classifying osteoporotic vertebral fractures in X-ray images

Purpose

To develop and validate a deep learning radiomics (DLR) model that uses X-ray images to predict the classification of osteoporotic vertebral fractures (OVFs).

Material and methods

The study encompassed a cohort of 942 patients, involving examinations of 1076 vertebrae through X-ray, CT, and MRI across three distinct hospitals. The OVFs were categorized as class 0, 1, or 2 based on the Assessment System of Thoracolumbar Osteoporotic Fracture. The dataset was divided randomly into four distinct subsets: a training set comprising 712 samples, an internal validation set with 178 samples, an external validation set containing 111 samples, and a prospective validation set consisting of 75 samples. The ResNet-50 architectural model was used to implement deep transfer learning (DTL), undergoing -pre-training separately on the RadImageNet and ImageNet datasets. Features from DTL and radiomics were extracted and integrated using X-ray images. The optimal fusion feature model was identified through least absolute shrinkage and selection operator logistic regression. Evaluation of the predictive capabilities for OVFs classification involved eight machine learning models, assessed through receiver operating characteristic curves employing the "One-vs-Rest" strategy. The Delong test was applied to compare the predictive performance of the superior RadImageNet model against the ImageNet model.

Results

Following pre-training separately on RadImageNet and ImageNet datasets, feature selection and fusion yielded 17 and 12 fusion features, respectively. Logistic regression emerged as the optimal machine learning algorithm for both DLR models. Across the training set, internal validation set, external validation set, and prospective validation set, the macro-average Area Under the Curve (AUC) based on the RadImageNet dataset surpassed those based on the ImageNet dataset, with statistically significant differences observed (P<0.05). Utilizing the binary "One-vs-Rest" strategy, the model based on the RadImageNet dataset demonstrated superior efficacy in predicting Class 0, achieving an AUC of 0.969 and accuracy of 0.863. Predicting Class 1 yielded an AUC of 0.945 and accuracy of 0.875, while for Class 2, the AUC and accuracy were 0.809 and 0.692, respectively.

Conclusion

The DLR model, based on the RadImageNet dataset, outperformed the ImageNet model in predicting the classification of OVFs, with generalizability confirmed in the prospective validation set.

The treatment effect of posterior lumbar fusion surgery on patients suffering from lumbar disc herniation concurrent with peroneal nerve paralysis

Purpose

The purpose of this study is to investigate the clinical effect of posterior lumbar fusion surgery on patients who suffer from lumbar disc herniation concurrent with peroneal nerve paralysis.

Methods

The patients suffering from peroneal nerve paralysis and undergoing posterior lumbar fusion surgery between January 2012 and December 2019 were retrospectively reviewed. The data of the identified patients were then collected and processed. All patients were followed up post-operatively after discharge from the hospital. The data was analyzed in terms of Oswestry disability index (ODI), visual analogue scale (VAS) score, and relative lower-limb muscle strength.

Results

A total of 87 patients (52 males and 35 females) aged 54 ± 11 years met the inclusion criteria for this study. These patients stayed in hospital for 16 ± 6 days and were followed up for 81 ± 24 months. Data analysis showed that muscle strength of the tibialis anterior and extensor digitorum significantly recovered at the last follow-up with a grade of 3 (median), compared to grade 0 at admission (p < 0.001). Furthermore, the median VAS score decreased to 1 at the last follow-up from 6 at admission (p < 0.001), and the ODI greatly improved with 10% (median) at the last follow-up, while it was 58% at admission (p < 0.001). The ODI improvement rate was 60% on average at the last follow-up. Multivariate regression analysis regarding the ODI and muscle strength improvement rates showed that advanced age was a risk factor for postoperative recovery.

Conclusions

Most of the patients suffering from lumbar disc herniation concurrent with peroneal nerve paralysis can improve after undergoing posterior lumbar fusion surgery, but few can reach full recovery. Advanced age might be a risk factor that affects the prognosis of these patients after surgery.

A novel and convenient method to evaluate bone cement distribution following percutaneous vertebral augmentation

A convenient method to evaluate bone cement distribution following vertebral augmentation is lacking, and therefore so is our understanding of the optimal distribution. To address these questions, we conducted a retrospective study using data from patients with a single-segment vertebral fracture who were treated with vertebral augmentation at our two hospitals. Five evaluation methods based on X-ray film were compared to determine the best evaluation method and the optimal cement distribution. Of the 263 patients included, 49 (18.63%) experienced re-collapse of treated vertebrae and 119 (45.25%) experienced new fractures during follow-up. A 12-score evaluation method (kappa value = 0.652) showed the largest area under the receiver operating characteristic curve for predicting new fractures (0.591) or re-collapse (0.933). In linear regression with the 12-score method, the bone cement distribution showed a negative correlation with the re-collapse of treated vertebra, but it showed a weak correlation with new fracture. The two prediction curves intersected at a score of 10. We conclude that an X-ray-based method for evaluation of bone cement distribution can be convenient and practical, and it can reliably predict risk of new fracture and re-collapse. The 12-score method showed the strongest predictive power, with a score of 10 suggesting optimal bone cement distribution.

Clinical and radiological subsequent fractures after vertebral augmentation for treating osteoporotic vertebral compression fractures: a meta-analysis

PURPOSE

This study aimed to identify all relevant randomized controlled trials (RCT) and prospective non-RCTs to further investigate whether percutaneous vertebral augmentation (PVA) was associated with clinical and radiological subsequent fractures on unoperated levels.

METHODS

We systematically searched PubMed, EMBASE, Cochrane library, Google Scholar, web of science, and ClinicalTrial.gov from the establishment of the database to January 2020. All eligible studies comparing subsequent fractures after PVA with those after conservative treatment (CT) were incorporated. The pooled risk ratio (RR) with its 95% confidence intervals (95% CIs) was used. Heterogeneity, sensitivity, and publication bias analyses were performed.

RESULTS

In all, 32 studies were included in the study: 82/512 patients (16.02%) and 58/433 patients (13.39%) had clinical subsequent fractures in the PVA group and CT group, respectively. No significant differences were observed between the two groups [RR = 1.22, 95% CI 0.70-2.12, P = 0.49]. Further, 175/837 patients (20.91%) in the PVA group and 160/828 patients (19.32%) in the CT group had radiological subsequent fractures. No significant difference was observed between groups [RR = 0.91, 95% CI 0.71-2.12, P = 1.16]. Further, no statistical difference was observed on subgroup analysis between RCTs and non-RCTs or PVP and PKP.

CONCLUSION

Our systematic review revealed that subsequent fractures on unoperated levels were not associated with PVA, regardless of whether they were clinical or radiological subsequent fractures.