Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Xinghua Song has a total of 125 co-authors. Here are the top co-authors by number of shared publications:

  • Haopeng Luan

    13 shared publications
  • Cong Peng

    Department of Spine Surgery, Hanzhong Central Hospital, Hanzhong, China

    12 shared publications
  • Parhat Yasin

    Department of Spine Surgery, The Sixth Affiliated Hospital of Xinjiang Medical University, Urumqi, PR China

    9 shared publications
  • Jiangtao Chen

    First Affiliated Hospital, Xinjiang Medical University, Urumqi, Xinjiang Uyghur Autonomous Region, 830011, China

    6 shared publications
  • Wei Xiang

    Department of Spine Surgery, The Sixth Affiliated Hospital of Xinjiang Medical University, Urumqi, PR China

    6 shared publications
  • Kai Liu

    Department of Trauma and Microreconstructive Surgery, The First Affiliated Hospital of Xinjiang Medical University, Urumqi, China

    5 shared publications

Latest Publications

  • Open Access

    A novel minimally invasive metabolomic strategy for differential diagnosis of spinal infections: blood-derived metabolic signatures as candidate biomarkers distinguishing tuberculous spondylitis from brucellar spondylitis

    by Parhat Yasin, Hongwei Ye, Xinghua Song, Jie Sheng

    Background: Tuberculous spondylitis (TS) and brucellar spondylitis (BS) can present with overlapping manifestations. This study assessed plasma metabolic differences between the two diagnostic groups. Methods: We conducted a retrospective case-control study of 53 patients (36 with TS and 17 with BS) enrolled between January 2021 and December 2024. Diagnoses were established using a composite reference standard integrating compatible clinical and radiological findings with microbiological and/or histopathological evidence. Untargeted ultra-high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS), multivariate analyses and receiver operating characteristic (ROC) analyses were used. Candidate selection and ROC assessment used the same participants. Results: The processed dataset contained 1060 features, of which 190 met the nominal differential-screening criteria (56 with higher and 134 with lower relative peak areas in TS). Lipids and lipid-like molecules accounted for 390 of 1058 classified features (36.86%); 2 features were unclassified. No pathway reached nominal P < 0.05 among 27 tested pathways; linoleic acid metabolism had the lowest P value (0.0767). The highest individual apparent area under the ROC curve (AUC) was 0.922 for the feature annotated as 12-hydroxyjasmonic acid. An unweighted sum of the top nine relative peak areas had an apparent AUC of 0.904 (95% CI 0.825-0.983), apparent sensitivity 0.806 and apparent specificity 0.882. Conclusion: The results identify candidate features for distinguishing the sampled TS and BS groups. They do not establish independent associations, specificity against other pathogens, validated diagnostic accuracy or clinical utility.

  • Open Access

    Explainable machine learning differentiates necrotizing fasciitis and osteomyelitis via routine blood biomarkers

    by Parhat Yasin, Zubaidanmu Aizezi, Shiming Dong, Yasen Yimit, Alimujiang Yusufu, Wei Xiang, Zhoujun Zhu, Haopeng Luan, Xinghua Song

    Necrotizing fasciitis (NF) and osteomyelitis (OM) are severe, limb-threatening infections with overlapping features, making early differentiation challenging. To address this, we developed and validated an explainable machine learning model using routine blood biomarkers from a retrospective, multi-center cohort of 3415 patients (579 NF, 2836 OM). Data from a primary center were used for model development, with data from a second center serving as an independent external testing cohort. Systematic evaluation identified an optimal 10-biomarker LightGBM model that achieved outstanding discrimination on the external cohort, with an AUC of 0.926. Beyond its high accuracy, explainability analyses confirmed the model’s predictions are driven by robust, clinically relevant markers of severe inflammation and metabolic dysfunction, reinforcing its trustworthiness. The final model was deployed as a publicly accessible web tool for real-time risk stratification. This work provides a powerful, externally validated, and explainable AI framework to augment clinical judgment, with strong potential to reduce diagnostic delays and improve outcomes for these devastating infections.


  • Open Access

    Bibliometric and Visualization Analysis of Research on Exosomes as Drug Delivery Systems (2008-2023)

    by Wei Xiang, Qisong Shang, Zhoujun Zhu, Yuanyuan Wu, Xinghua Song

    Objective: Exosomes are unique bio-nanomaterials possessing significant value and potential for drug delivery systems. However, to date, no bibliometric studies in this field have been reported. Our aim is to explore the research hotspots and trends of exosome drug-carrying systems across various medical fields through bibliometric analyses. Methods: Articles and reviews related to “exosome” and “drug delivery” are retrieved from the Web of Science Core Collection. VOSviewer, CiteSpace, Scimago Graphica, and Origin 2021 are employed for bibliometric analyses. Results: A total of 771 articles from 60 countries, such as China and the United States, are included. The number of papers concerning exosomal drug delivery systems has been increasing yearly. The main research institutions are the Chinese Academy of Sciences, Shanghai Jiao Tong University, Huazhong University of Science and Technology, Fudan University, and Sichuan University. The Journal of Controlled Release is the most prevalent and frequently cited journal in this field. These papers are authored by 247 individuals, with Ando, Hidenori having the highest number of publications and Alvarez-Erviti L receiving the most citations. “Extracellular vesicles”, “drug delivery”, “in vitro”, “nanoparticles”, “cells”, “delivery”, and “mesenchymal stem cells” are the principal keywords for this hotspot. Conclusion: This pioneering bibliometric study offers a comprehensive overview of the research trends and advancements in exosomal drug delivery systems in medicine over the past fifteen years.


  • Open Access

    Clinical Risk Stratification and Neutrophil‐Mediated Neuroinflammation in Myeloid Hypertension‐Like Syndrome: An Integrated Clinical and Single‐Cell Transcriptomic Study

    by Qisong Shang, Haopeng Luan, Pengyuan Han, Xiaonan Wang, Junru Zheng, Xing Wang, Xinghua Song

    Background: Myeloid hypertension‐like (MHL) syndrome is a clinically relevant pressure‐associated complication that may occur in patients undergoing endoscopic spinal surgery and can lead to neurovascular injury and inflammatory neurological dysfunction. However, the patient‐level risk profile and immune mechanisms underlying this condition remain insufficiently defined. This study aimed to explore the key immune mechanisms underlying MHL syndrome–induced neurovascular injury using artificial intelligence (AI)–assisted single‐cell transcriptomics, identify specific pathogenic neutrophil subsets, and clarify the regulatory role of the lysine acetyltransferase 8 (KAT8)–neutrophil cytosolic factor 1 (NCF1) axis in neutrophil activation and neutrophil extracellular trap (NET) formation. Method: A retrospective cohort of 244 patients undergoing endoscopic spinal surgery was analyzed to develop and internally validate a multivariable logistic regression model for MHL risk stratification. A nonhuman primate pressure‐associated model was then established to evaluate systemic hemodynamics, spinal microcirculation, histopathological injury, apoptosis, and inflammatory cytokine release. Single‐cell RNA sequencing of peripheral blood and spinal cord tissues was performed to characterize immune cell remodeling, followed by cell–cell communication analysis and gain‐ and loss‐of‐function experiments in a human HL‐60‐based neutrophil‐like cell model to examine the role of the KAT8–NCF1 axis in neutrophil‐like cell activation, oxidative stress, and NET formation. Results: In multivariable analysis, cerebrospinal fluid leakage (CFL) was the strongest predictor of MHL. The model showed good discrimination, with area under the receiver operating characteristic curve (AUC) values of 0.968 and 0.919 in the training and internal validation cohorts, respectively. MHL modeling caused significant neurovascular dysfunction, elevated blood pressure, spinal tissue injury, cell apoptosis, and excessive proinflammatory cytokine release, accompanied by myeloid cell expansion and decreased CD8 + T cells. AI‐driven single‐cell analysis identified a circulation‐enriched proinflammatory neutrophil subset, Neu_KAT8_NCF1, which was highly enriched in oxidative stress and NETs‐related pathways. Mechanistically, KAT8 upregulated NCF1 to promote reactive oxygen species (ROS) accumulation, NET formation, and proinflammatory cytokine secretion, while NCF1 knockdown reversed KAT8‐induced neutrophil hyperactivation. Furthermore, CellChat analysis predicted extensive ligand–receptor communication between Neu_KAT8_NCF1 cells and macrophage and cytotoxic CD8+ T‐cell populations, suggesting their potential involvement in systemic and local immune remodeling. Conclusions: This study identifies Neu_KAT8_NCF1 neutrophils as a disease‐associated circulating neutrophil subset associated with pressure‐related neurovascular injury in MHL syndrome. The KAT8–NCF1 axis may contribute to neuroinflammation by facilitating neutrophil oxidative stress, excessive NET formation, and myeloid‐lymphoid immune dysregulation. These findings provide a clinical and mechanistic framework for identifying high‐risk patients and monitoring perioperative inflammatory responses, although the therapeutic value of targeting this pathway requires further validation using selective modulation strategies.

  • Open Access

    Comprehensive comparative analysis of explainable deep learning model for differentiation of brucellar spondylitis and tuberculous spondylitis through MRI sequences

    by Parhat Yasin, Abudouresuli Tuersun, Anuar Ashir, Yerlan Makhambetov, Jie Sheng, Xinghua Song

    Background: The differentiation of brucellar spondylitis (BS) from tuberculous spondylitis (TS) on magnetic resonance imaging (MRI) is a critical clinical challenge. While deep learning holds promise, the optimal architectural strategy for integrating information from multi-sequence MRI remains unclear. This study systematically compared distinct deep learning architectures to identify a valid and effective integration strategy for this diagnostic problem. Methods: In this retrospective, single-center diagnostic study, we included 235 patients with surgically and pathologically confirmed BS ( n = 82) or TS ( n = 153) from January 2014 to December 2024. We systematically evaluated four distinct architectural strategies for processing sagittal T1-weighted, T2-weighted, and fat-suppressed MRI sequences: (1) baseline models trained on single sequences; (2) a single-branch model that fused sequences as input channels; (3) a heterogeneous multi-branch model using different backbones for each sequence; and (4) a homogeneous multi-branch model using identical backbones. Models were developed on patient-level data splits for training (70%), validation (15%), and internal testing (15%). The primary performance metric was the area under the receiver operating characteristic curve (AUC) on the test set. Statistical significance of performance differences between models was assessed using the DeLong test, with P values adjusted for multiple comparisons using the Benjamini–Hochberg procedure. Results: The single-branch fusion model, which treated the three sequences as channels in a single input, failed to learn, yielding performance equivalent to random chance (test AUC range: 0.474–0.538). In stark contrast, both the single-sequence and multi-branch architectures proved to be effective. The best single-sequence model achieved a test AUC of 0.765 (95% CI 0.759–0.771). The optimal multi-branch model, which successfully integrated all three sequences, achieved a comparable test AUC of 0.764 (95% CI 0.757–0.770). Conclusions: The choice of architecture for integrating multi-sequence MRI data is a critical determinant of model viability. Our findings demonstrate that naive channel wise fusion is an invalid strategy for this task. In contrast, both processing a single MRI sequence and utilizing a multi-branch parallel-processing architecture are valid and effective strategies, achieving comparable diagnostic performance. This study clarifies the architectural principles required for successfully applying deep learning to this multi-modal diagnostic challenge.


  • Open Access

    Decoding the exosomal nucleic acid delivery system axis of macrophage autophagy and immune reprogramming via multi-omics analysis

    by Zhoujun Zhu, Wei Xiang, Pengchao Zhang, Parhat Yasin, Xinghua Song

    Background: MicroRNA-155 (miR-155) is a key regulator of macrophage function, and its abnormal expression is closely associated with the pathogenesis of tuberculosis (TB)—a disease where impaired macrophage autophagy weakens anti-mycobacterial immunity. Exosomes are promising nucleic acid carriers due to their biocompatibility and cell-targeting ability. Here, we constructed exosome-based miR-155 delivery systems (Exo-miR155-ago/Exo-miR155-antago; “ago” = agomir, a miR-155 agonist that enhances its expression; “antago” = antagomir, a miR-155 antagonist that inhibits its expression) to modulate macrophage autophagy and remold anti-TB immune responses. Methods: Exosomes were isolated from the supernatant of bone marrow mesenchymal stem cells using differential centrifugation. The miR155-5p agomir and antagomir were transfected into exosomes via the Exosome Transfection Kit, followed by co-incubation with macrophages. Transcriptomics and proteomics were employed to screen for differentially expressed genes and proteins. Western blot was employed to detect autophagy-related proteins and phosphorylated proteins in signaling pathways (p- denotes phosphorylation, a key post-translational modification regulating protein activity). Techniques including transmission electron microscopy (TEM), Monodansylcadaverine (MDC) staining, and reverse transcription-quantitative polymerase chain reaction (RT-qPCR) were applied to detect the autophagic level of macrophages. Results: Transcriptome sequencing identified 704 differentially expressed genes, with significant enrichment in TNF and NF-κB pathways, differential expression of NF-κB target genes (e.g., autophagy core gene Beclin1), and expression changes in key genes of the energy metabolism-related AMPK/mTOR pathway; proteomic analysis found 164 differentially expressed proteins, including key molecules of the “Pathogen Recognition-TLR4-NF-κB-Autophagy-Related Gene Transcription” pathway (TLR4, p-p65) and core proteins of the AMPK/mTOR pathway (p-AMPK, p-mTOR); functional verification showed the Exo-miR155-ago group had more autophagosomes (TEM), higher autophagic vacuole accumulation (MDC staining), upregulated mRNA/protein of autophagy-related molecules (LC3B, Beclin1), downregulated mRNA/protein of p62 (RT-qPCR/Western blot), activated p-p65 (NF-κB pathway), and increased p-AMPK with decreased p-mTOR (AMPK/mTOR pathway), and all results confirmed Exo-miR155-ago promotes macrophage autophagy via the synergistic effect of the two pathways. Conclusion: This study provides multi-omics evidence for autophagy modulation mediated by the exosomal nucleic acid delivery system, verifies that this system regulates macrophage autophagy by controlling the TLR4-NF-κB pathway and AMPK/mTOR pathway, and clarifies the application potential of this system in tuberculosis (TB) and other macrophage-associated.


  • Open Access

    Explainable machine learning for differential diagnosis of diabetic foot infection and osteomyelitis: a two-center study and clinically applicable web calculator using routine blood biomarkers

    by Parhat Yasin, Shiming Dong, Zubaidanmu Aizezi, Yasen Yimit, Alimujiang Yusufu, Maihemuti Yakufu, Xinghua Song

    Diabetic foot complications, including infections and osteomyelitis, pose significant health risks, with high prevalence and amputation rates. Differentiating diabetic foot infection (DFI) from osteomyelitis (OM) is challenging due to overlapping symptoms and limitations of current diagnostic methods. This study aimed to develop and validate an explainable machine learning (ML) model using routine blood biomarkers to improve differential diagnosis and provide a clinically accessible tool. This retrospective, two-center study included 3,612 patients diagnosed with either DFI (n = 1,699) or OM (n = 1,913). Data from Center 1 (n = 3271) were used for model development (75% training, 25% internal validation), and data from Center 2 (n = 341) served as an independent external validation cohort. A robust feature selection pipeline identified the most predictive routine biomarkers. Multiple machine learning classifiers were trained and evaluated, with the top-performing model selected based on the area under the receiver operating characteristic curve (AUC), Brier score, and other key metrics. Explainable AI (XAI) techniques (SHAP, LIME) were used to ensure model transparency. A web-based calculator was developed for clinical translation. A LightGBM model using only six biomarkers—Age, HbA1c, Creatinine, Albumin, ESR, and Sodium—was selected as the final model. It achieved an AUC of 0.879 (95% CI 0.854–0.902) in internal validation and demonstrated excellent, generalizable performance in the external cohort with an AUC of 0.942 (95% CI 0.936–0.950). The model was well-calibrated and showed significant clinical utility in decision curve analysis. SHAP analysis quantified the specific contribution of each biomarker to individual predictions, enhancing interpretability. The final model was deployed as a user-friendly, publicly accessible web calculator. An externally validated machine learning model based on six routine blood biomarkers can accurately and reliably differentiate DFI from OM. The model demonstrated high discriminative performance and clinical utility. Deployed as a transparent web calculator with integrated explainable AI, this low-cost tool has the potential to aid clinicians in diagnostic decision-making, particularly in resource-limited settings. Not applicable.


  • Open Access

    Dual-center study on AI-driven multi-label deep learning for X-ray screening of knee abnormalities

    by Parhat Yasin, Yasen Yimit, Abuduainijiang Abulimiti, Haopeng Luan, Cong Peng, Maihemuti Yakufu, Xinghua Song

    Knee abnormalities, such as meniscus tears and ligament injuries, are common in clinical practice and pose significant diagnostic challenges. While traditional imaging techniques—X-ray, Computed Tomography (CT) scan, and Magnetic Resonance Imaging (MRI)—are vital for assessment. However, X-rays and CT scans often fail to adequately visualize soft tissue injuries, and MRIs can be costly and time-consuming. To overcome these limitations, we developed an innovative AI-driven approach that allows for the detection of soft tissue abnormalities directly from X-ray images—a capability traditionally reserved for MRI or arthroscopy. We conducted a retrospective study with 4,215 patients from two medical centers, utilizing knee X-ray images annotated by orthopedic surgeons. The YOLOv11 model automated knee localization, while five convolutional neural networks—ResNet152, DenseNet121, MobileNetV3, ShuffleNetV2, and VGG19—were adapted for multi-label classification of eight conditions: meniscus tears (MENI), anterior cruciate ligament tears (ACL), posterior cruciate ligament injuries (PCL), medial collateral ligament injuries (MCL), lateral collateral ligament injuries (LCL), joint effusion (EFFU), bone marrow edema or contusion (CONT), and soft tissue injuries (STI). Data preprocessing involved normalization and Region of Interest (ROI) extraction, with training enhanced by spatial augmentations. Performance was assessed using mean average precision (mAP), F1-scores, and area under the curve (AUC). We also developed a Windows-based PyQt application and a Flask Web application for clinical integration, incorporating explainable AI techniques (GradCAM, ScoreCAM) for interpretability. The YOLOv11 model achieved precise knee localization with a mAP@0.5 of 0.995. In classification, ResNet152 outperformed others, recording a mAP of 90.1% in internal testing and AUCs up to 0.863 (EFFU) in external testing. End-to-end performance on the external set yielded a mAP of 86.1% and F1-scores of 84.0% with ResNet152. The Windows and web applications successfully processed imaging data, aligning with MRI and arthroscopic findings in cases like ACL and meniscus tears. Explainable AI visualizations clarified model decisions, highlighting key regions for complex injuries, such as concurrent ligament and soft tissue damage, enhancing clinical trust. This AI-driven model markedly improved the precision and efficiency of knee abnormality detection through X-ray analysis. By accurately identifying multiple coexisting conditions in a single pass, it offered a scalable tool to enhance diagnostic workflows and patient outcomes, especially in resource-constrained areas.


  • Open Access

    Development and validation of an interpretable nomogram for predicting the risk of the prolonged postoperative length of stay for tuberculous spondylitis: a novel approach for risk stratification

    by Parhat Yasin, Haopeng Luan, Cong Peng, Xinghua Song

    Tuberculous spondylitis (TS) is a challenging health care condition requiring spine surgery, and predicting the probabilities of prolonged postoperative length of stay (PLOS) can aid in effective management, especially with the increasing number of aging patients. This study aimed to develop and validate an interpretable nomogram for risk stratification and prediction of prolonged PLOS for TS patients after surgery, utilizing SHapley Additive exPlanations (SHAP) for model interpretation. This retrospective analysis comprised data of 580 TS patients that were hospitalized between January 2016 and December 2022. Prolonged PLOS was defined as hospitalization exceeding the 75th percentile. Factors associated with an increased risk of prolonged PLOS were identified using the Least Absolute Shrinkage and Selection Operator (LASSO) method and incorporated into a multivariable logistic regression model. SHAP values were generated to explain the contribution of each variable in predicting prolonged PLOS. Based on the identified risk variables, a nomogram was constructed using the SHAP values to represent each factor’s contribution to the final outcome. The nomogram’s effectiveness was assessed using calibration plots, discrimination analysis, and decision curve analysis. Among 580 patients, 127 had prolonged postoperative length of stay (PLOS > 11 days), while 453 had normal stays (≤ 11 days). The developed nomogram incorporated 7 significant risk factors along with their corresponding SHAP values, which are C-reactive protein (CRP), multiple sections, CT-vertebral destruction, MRI-epidural abscess, transfusions, blood loss and postoperative drainage (Drainage volume on the first day after surgery). The calibration graphs showed that the expected and observed probability of prolonged PLOS was in close agreement. Discrimination analysis yielded an area under the curve (AUC) of 0.867 (95% CI: 0.828–0.908) in the training set, indicating good predictive performance. Decision curve analysis confirmed the clinical utility of the nomogram in risk stratification and treatment decision-making. By utilizing SHAP for model interpretation, this study provides an interpretable nomogram for assessing the possibility of extended PLOS in TS patients. The inclusion of SHAP values allows clinicians to understand the contribution of each variable in the prediction, thereby increasing transparency and aiding in the decision-making process. This nomogram has the potential to contribute to improved patient management and optimization of resource allocation. Prospective validation studies are recommended to further evaluate its effectiveness in clinical practice. Not applicable.


  • Open Access

    Machine Learning-Based Interpretable Screening for Osteoporosis in Tuberculosis Spondylitis Patients Using Blood Test Data: Development and External Validation of a Novel Web-Based Risk Calculator with Explainable Artificial Intelligence (XAI)

    by Parhat Yasin, Liwen Ding, Mardan Mamat, Wei Guo, Xinghua Song

    Background: Tuberculosis spondylitis (TS), also known as Pott’s disease, is the most common destructive form of musculoskeletal tuberculosis and poses significant clinical challenges, particularly when complicated by osteoporosis. Osteoporosis exacerbates surgical outcomes and increases the risk of complications, making its accurate prediction crucial for effective patient management. Methods: This retrospective study included 906 TS patients from two medical centers between January 2016 and November 2022. We collected demographic information and blood test data from routine examinations. To address class imbalance, the synthetic minority oversampling technique (SMOTE) was applied. Feature selection was performed using LASSO, Boruta, and Recursive Feature Elimination (RFE) to identify key predictors of osteoporosis. Multiple machine learning (ML) algorithms, including logistic regression, random forest, and XGBoost, were trained and optimized using nested cross-validation and hyperparameter tuning. The optimal model was further refined through threshold tuning to enhance performance metrics. Model interpretability was achieved using SHapley Additive exPlanations (SHAP), and an online web application was developed for real-time clinical use. Results: Out of 906 patients, 60 were diagnosed with osteoporosis based on Dual-energy X-ray absorptiometry (DXA) measurements. Feature selection identified hemoglobin (HB), estimated glomerular filtration rate (eGFR), and cystatin C (CYS_C) as significant predictors. The logistic regression model exhibited the highest performance with an area under the receiver operating characteristic curve (AUC) of 0.826, which was externally validated with an AUC of 0.796. Threshold tuning optimized the decision threshold to 0.32, improving the F1-score and balancing sensitivity and specificity. SHAP analysis highlighted the critical roles of HB, eGFR, and CYS_C in osteoporosis prediction. The developed web application facilitates the model’s integration into clinical workflows, enabling healthcare professionals to make informed decisions at the bedside. Conclusion: This study successfully developed and validated an ML-based tool for predicting osteoporosis in TS patients using readily available clinical data. The model demonstrated robust predictive performance and was effectively integrated into a user-friendly online application, offering a practical solution to enhance surgical decision-making and improve patient outcomes in real-time clinical settings.