Search

Objectives The incidence and adverse events of postoperative blood transfusion in spinal tuberculosis (TB) have attracted increasing attention. Our purpose was to develop a prediction model to evaluate blood transfusion risk after spinal fusion (SF) for spinal TB. Methods Nomogram and machine learning algorithms, support vector machine (SVM), decision tree (DT), multilayer perceptron (MLP), Naive Bayesian (NB), k-nearest neighbors (K-NN) and random forest (RF), were constructed to identified predictors of blood transfusion from all spinal TB cases treated by SF in our department between May 2010 and April 2020. The prediction performance of the models was evaluated by 10-fold cross-validation. We calculated the average AUC and the maximum AUC, then demonstrated the ROC curve with maximum AUC. Results The collected cohort ultimately was consisted of 152 patients, where 56 required allogeneic blood transfusions. The predictors were surgical duration, preoperative Hb, preoperative ABL, preoperative MCHC, number of fused vertebrae, IBL, and anticoagulant history. We obtained the average AUC of nomogram (0.75), SVM (0.62), k-NM (0.65), DT (0.56), NB (0.74), MLP (0.56) and RF (0.72). An interactive web calculator based on this model has been provided (https://drwenleli.shinyapps.io/STTapp/). Conclusions We confirmed seven independent risk factors affecting blood transfusion and diagramed them with the nomogram and web calculator.

Background This study aimed to predict C5 palsy (C5P) after posterior laminectomy and fusion (PLF) with cervical myelopathy (CM) from routinely available variables using a support vector machine (SVM) method. Methods We conducted a retrospective investigation based on 184 consecutive patients with CM after PLF, and data were collected from March 2013 to December 2019. Clinical and imaging variables were obtained and imported into univariable and multivariable logistic regression analyses to identify risk factors for C5P. According to published reports and clinical experience, a series of variables was selected to develop an SVM machine learning model to predict C5P. The accuracy (ACC), area under the receiver operating characteristic curve (AUC), and confusion matrices were used to evaluate the performance of the prediction model. Results Among the 184 consecutive patients, C5P occurred in 26 patients (14.13%). Multivariate analyses demonstrated the following 4 independent factors associated with C5P: abnormal electromyogram (odds ratio [OR] = 7.861), JOA recovery rate (OR = 1.412), modified Pavlov ratio (OR = 0.009), and presence of C4–C5 foraminal stenosis (OR = 15.492). The SVM model achieved an area under the receiver operating characteristic curve (AUC) of 0.923 and an ACC of 0.918. Additionally, the confusion matrix showed the classification results of the discriminant analysis. Conclusions The designed SVM model presented satisfactory performance in predicting C5P from routinely available variables. However, future external validation is needed.

Background Percutaneous pedicle screw fixation (PPSF) is the primary approach for single-segment thoracolumbar burst fractures (TLBF). The healing angle at the thoracolumbar junction is one of the most significant criteria for evaluating the efficacy of PPSF. Therefore, the purpose of this study was to analyze the predictors associated with the poor postoperative alignment of the thoracolumbar region from routine variables using a support vector machine (SVM) model. Methods We retrospectively analyzed patients with TLBF operated at our academic institute between March 1, 2014 and December 31, 2019. Stepwise logistic regression analysis was performed to assess potential statistical differences between all clinical and radiological variables and the adverse events. Based on multivariate logistic results, a series of independent risk factors were fed into the SVM model. Meanwhile, the feature importance of radiologic outcome for each parameter was explored. The predictive performance of the SVM classifier was evaluated using the area under the receiver operating characteristic curve (AUC), accuracy (ACC) and confusion matrices with 10-fold cross-validation, respectively. Results In the recruited 150 TLBFs, unfavorable radiological outcomes were observed in 53 patients (35.33%). The relationship between osteoporosis (p = 0.036), preoperative Cobb angle (p = 0.001), immediate postoperative Cobb angle (p = 0.029), surgically corrected Cobb angle (p = 0.001), intervertebral disc injury (Score 2 p = 0.001, Score 3 p = 0.001), interpedicular distance (IPD) (p = 0.001), vertebral body compression rate (VBCR) (p = 0.010) and adverse events was confirmed by univariate regression. Thereafter, independent risk factors including preoperative Cobb angle, the disc status and IPD and independent protective factors surgical correction angle were identified by multivariable logistic regression. The established SVM classifier demonstrated favorable predictive performance with the best AUC = 0.93, average AUC = 0.88, and average ACC = 0.87. The variables associated with radiological outcomes, in order of correlation strength, were intervertebral disc injury (42%), surgically corrected Cobb angle (25%), preoperative Cobb angle (18%), and IPD (15%). The confusion matrix reveals the classification results of the discriminant analysis. Conclusions Critical radiographic indicators and surgical purposes were confirmed to be associated with an unfavorable radiographic outcome of TLBF. This SVM model demonstrated good predictive ability for endpoints in terms of adverse events in patients after PPSF surgery.

BackgroundThis study aims to predict the lymphatic metastasis in Ewing's sarcoma (ES) patients by nomogram. The risk of lymphatic metastasis in patients with ES was predicted by the built model, which provided guidance for the clinical diagnosis and treatment planning.MethodsA total of 929 patients diagnosed with ES were enrolled from the year of 2010 to 2016 in the Surveillance, Epidemiology, and End Results (SEER) database. The nomogram was established to determine predictive factors of lymphatic metastasis according to univariate and multivariate logistic regression analysis. The validation of the model performed using multicenter data (n = 51). Receiver operating characteristics (ROC) curves and calibration plots were used to evaluate the prediction accuracy of the nomogram. Decision curve analysis (DCA) was implemented to illustrate the practicability of the nomogram clinical application. Based on the nomogram, we established a web calculator to visualize the risk of lymphatic metastases. We further plotted Kaplan-Meier overall survival (OS) curves to compare the survival time of patients with and without lymphatic metastasis.ResultsIn this study, the nomogram was established based on six significant factors (survival time, race, T stage, M stage, surgery, and lung metastasis), which were identified for lymphatic metastasis in ES patients. The model showed significant diagnostic accuracy with the value of the area under the curve (AUC) was 0.743 (95%CI: 0.714–0.771) for SEER internal validation and 0.763 (95%CI: 0.623–0.871) for multicenter data external validation. The calibration plot and DCA indicated that the model had vital clinical application value.ConclusionIn this study, we constructed and developed a nomogram with risk factors to predict lymphatic metastasis in ES patients and validated accuracy of itself. We found T stage (Tx OR = 2.540, 95%CI = 1.433–4.503, P < 0.01), M stage (M1, OR = 2.061, 95%CI = 1.189–3.573, P < 0.05) and survival time (OR = 0.982, 95%CI = 0.972–0.992, P < 0.001) were important independent factors for lymphatic metastasis in ES patients. Furthermore, survival time in patients with lymphatic metastasis or unclear situation (P < 0.0001) was significantly lower. It can help clinicians make better decisions to provide more accurate prognosis and treatment for ES patients.

Purpose The objective of this study is to construct a computer aided diagnosis system for normal people and pneumoconiosis using X-raysand deep learning algorithms. Materials and methods 1760 anonymous digital X-ray images of real patients between January 2017 and June 2020 were collected for this experiment. In order to concentrate the feature extraction ability of the model more on the lung region and restrain the influence of external background factors, a two-stage pipeline from coarse to fine was established. First, the U-Net model was used to extract the lung regions on each sides of the collection images. Second, the ResNet-34 model with transfer learning strategy was implemented to learn the image features extracted in the lung region to achieve accurate classification of pneumoconiosis patients and normal people. Results Among the 1760 cases collected, the accuracy and the area under curve of the classification model were 92.46% and 89% respectively. Conclusion The successful application of deep learning in the diagnosis of pneumoconiosis further demonstrates the potential of medical artificial intelligence and proves the effectiveness of our proposed algorithm. However, when we further classified pneumoconiosis patients and normal subjects into four categories, we found that the overall accuracy decreased to 70.1%. We will use the CT modality in future studies to provide more details of lung regions.

Background This study aimed to develop and validate an individualized nomogram to predict the risk of positive hidden blood loss (HBL) in patients with single-level thoracolumbar burst fracture (TBF) during the perioperative period. Methods We conducted a retrospective investigation including 150 consecutive patients with TBL, and the corresponding patient data was extracted from March 2013 to March 2019. The independent risk factors for positive HBL were screened using univariate and multivariate logistic regression analyses. According to published literature and clinical experience, a series of variables were selected to develop a nomogram prediction model for positive HBL. The area under the receiver operating characteristic curves (AUC), C-index, calibration plot, and decision curve analysis (DCA) were used to evaluate the performance of the prediction model. Bootstrapping validation was performed to evaluate the performance of the model. Results Among the 150 consecutive patients, 62 patients were positive for HBL (38.0%). The multivariate logistic regression analysis showed that the six risk factors of age, length of surgical incision, duration of operation, percentage of vertebral height restoration (P1%), preoperative total cholesterol, and preoperative fibrinogen were independent risk factors of positive HBL. The C-index was 0.831 (95% CI 0.740–0.889) and 0.845 in bootstrapping validation, respectively. The calibration curve showed that the predicted probability of the model was consistent with the actual probability. Decision curve analysis (DCA) showed that the nomogram had clinical utility. Conclusion Overall, we explored the relationship between the positive HBL requirement and predictors. The individualized prediction model for patients with single-level TBF can accurately assess the risk of positive HBL and facilitate clinical decision making. However, external validation will be needed in the future.

Background The prognosis of lung metastasis (LM) in patients with chondrosarcoma was poor. The aim of this study was to construct a prognostic nomogram to predict the risk of LM, which was imperative and helpful for clinical diagnosis and treatment. Methods Data of all chondrosarcoma patients diagnosed between 2010 and 2016 was queried from the Surveillance, Epidemiology, and End Results (SEER) database. In this retrospective study, a total of 944 patients were enrolled and randomly splitting into training sets (n = 644) and validation cohorts(n = 280) at a ratio of 7:3. Univariate and multivariable logistic regression analyses were performed to identify the prognostic nomogram. The predictive ability of the nomogram model was assessed by calibration plots and receiver operating characteristics (ROCs) curve, while decision curve analysis (DCA) and clinical impact curve (CIC) were applied to measure predictive accuracy and clinical practice. Moreover, the nomogram was validated by the internal cohort. Results Five independent risk factors including age, sex, marital, tumor size, and lymph node involvement were identified by univariate and multivariable logistic regression. Calibration plots indicated great discrimination power of nomogram, while DCA and CIC presented that the nomogram had great clinical utility. In addition, receiver operating characteristics (ROCs) curve provided a predictive ability in the training sets (AUC = 0.789, 95% confidence interval [CI] 0.789–0.808) and the validation cohorts (AUC = 0.796, 95% confidence interval [CI] 0.744–0.841). Conclusion In our study, the nomogram accurately predicted risk factors of LM in patients with chondrosarcoma, which may guide surgeons and oncologists to optimize individual treatment and make a better clinical decisions. Trial registration JOSR-D-20-02045, 29 Dec 2020.

Videos are used widely as the media platforms for human beings to touch the physical change of the world. However, we always receive the mixed sound from the multiple sound objects, and cannot distinguish and localize the sounds as the separate entities in videos. In order to solve this problem, a model named the Deep Multi-Modal Attention Network (DMMAN), is established to model the unconstrained video datasets for further finishing the sound source separation and event localization tasks in this paper. Based on the multi-modal separator and multi-modal matching classifier module, our model focuses on the sound separation and modal synchronization problems using two stage fusion of the sound and visual features. To link the multi-modal separator and multi-modal matching classifier modules, the regression and classification losses are employed to build the loss function of the DMMAN. The estimated spectrum masks and attention synchronization scores calculated by the DMMAN can be easily generalized to the sound source and event localization tasks. The quantitative experimental results show the DMMAN not only separates the high quality of the sound sources evaluated by Signal-to-Distortion Ratio and Signal-to-Interference Ratio metrics, but also is suitable for the mixed sound scenes that are never heard jointly. Meanwhile, DMMAN achieves better classification accuracy than other contrast baselines for the event localization tasks.