Zhengfei Yang

Zhengfei Yang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Zhengfei Yang has a total of 224 co-authors. Here are the top co-authors by number of shared publications:

  • Siqi Liu

    Department of Emergency Medicine, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China

    10 shared publications
  • Wanchun Tang

    Weil Institute of Emergency and Critical Care Research, Virginia Commonwealth University, Richmond, Virginia

    10 shared publications
  • Qin Ling

    Weil Institute of Emergency and Critical Care Research, School of Medicine, Virginia Commonwealth University, Richmond, VA, 23284, USA

    8 shared publications
  • Wanchun Tang

    Weil Institute of Critical Care Medicine, Rancho Mirage, CA, United States; Keck School of Medicine of the University of Southern California, Los Angeles, CA, United States

    8 shared publications
  • Jinzhao Zhang

    Department of Emergency Medicine, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China

    7 shared publications
  • Lian Liang

    Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China

    7 shared publications

Latest Publications

  • Open Access

    Optimisation and validation of capture mNGS for predicting antimicrobial resistance

    by Qiao Lin, Xu Mei, Huimin Zheng, Jie Meng, Fusheng He, Bin Yang, Xiao Ru, Mengting Su, Danlan Wang, Ni Tan, et al.


  • Spatio-temporal graph attention network for rehabilitation movement classification

    by Qiyu Yang, Zehui Zhang, Yi Huang, Zhengfei Yang

    The automated assessment of home-based rehabilitation benefits from skeleton-based methods, which provide a robust and privacy-preserving means of monitoring patient movements. While such approaches effectively capture local body pose and spatial joint configurations, they often fall short in modeling the complete temporal coherence of an entire exercise sequence. To address this limitation, we construct a novel cascaded framework that combines a Spatio-Temporal Graph Convolutional Network (ST-GCN) for capturing local joint dynamics with a Transformer-based attention mechanism to model global, long-range dependencies. Evaluated on the UI-PRMD and KIMORE datasets under a rigorous 5-fold cross-validation protocol with independent test sets, STGA-Net achieved mean accuracies of 92.71% and 95.24%, respectively, surpassing existing state-of-the-art methods. These results demonstrate its strong potential as a quantitative and privacy-preserving tool for personalized rehabilitation assessment in both clinical and home settings.

  • Open Access

    Intravoxel incoherent motion MRI for early detection and assessment of renal injury after cardiac arrest and resuscitation in a rat model

    by Jie Zhang, Na Zhan, Haoyi Ye, Siqi Liu, Yunke Tan, Jinzhao Zhang, Teng Huang, Zhihua Wu, Zhengfei Yang, Zhifeng Liu

    Objective: To investigate the utility of intravoxel incoherent motion (IVIM) magnetic resonance imaging (MRI) for early detection of renal injury 24 h after cardiac arrest and cardiopulmonary resuscitation. Materials and methods: Twenty-four male Sprague–Dawley rats were randomized into cardiac arrest ( n = 15) and sham-operated ( n = 9) groups. Five cardiac arrest rats died during the procedure, resulting in 10 cardiac arrest and 9 sham rats for analysis. IVIM-MRI and renal parameters were assessed 24 h after cardiac arrest and cardiopulmonary resuscitation. Diffusion-related parameters, including apparent diffusion coefficient (ADC), true diffusion coefficient (D), pseudo-diffusion coefficient (D*), perfusion fraction (f), and exponential ADC (eADC), were obtained. The primary imaging endpoint was defined as D. Following completion of MRI, renal function, histopathology, and expression of apoptosis markers (Bcl-2, Caspase-3), water channel protein aquaporin-3, and tight junction proteins (zonula occludens-1, occludin) were subsequently evaluated. Results: Despite preserved conventional renal function, IVIM MRI detected early renal alterations 24 h after cardiac arrest. Compared with the sham group, eADC, D, ADC, D*, and f values were all significantly reduced. Histopathological analysis revealed pronounced tubular epithelial injury, increased apoptosis (Bcl-2, p < 0.001; Caspase-3, p = 0.010), and reduced expression of aquaporin-3 ( p = 0.039), zonula occludens-1 ( p < 0.001), and occludin ( p < 0.001). Conclusion: IVIM MRI enables early, noninvasive detection of renal injury 24 h after cardiac arrest and CPR, reflecting tubular injury, tight junction disruption, and perfusion deficits, potentially facilitating timely intervention and improving prognosis. Relevance statement: This approach may facilitate individualized therapy and improve post-cardiac arrest syndrome outcome. Key Points: The long-term outcomes of cardiac arrest remain poor due to post-cardiac arrest syndrome. IVIM MRI enables early, noninvasive detection of renal injury after cardiac arrest and cardiopulmonary resuscitation. This approach may facilitate individualized therapy and improve post-cardiac arrest syndrome. Graphical

  • Open Access

    Performance and Effectiveness of DeepSeek in Assisting Diagnosis of Acutely and Critically Ill Patients: A Single‐Center Retrospective Study

    by Teng Huang, Jinzhao Zhang, Junpeng Tang, Pengfei Wang, Fanrong Lin, Kaidi Tao, Siqi Liu, Zhengfei Yang


  • Multimodal MR Imaging Reveals the Mechanisms of Post‐Cardiac‐Arrest Brain edema: Ferroptosis‐Mediated BBB Disruption and AQP4 Dysfunction

    by Yunke Tan, HaoYi Ye, Qiulin Ge, Peng Wang, Zhifeng Liu, Zhengfei Yang

    Background: The role of ferroptosis in Cardiac arrest (CA)‐induced cerebral edema remains unclear. Purpose: To investigate whether ferroptosis contributes to blood–brain barrier (BBB) disruption and aquaporin‐4 (AQP4) dysfunction following CA. Study Type: Prospective. Animal Model: Asphyxia‐induced CA rat model. Forty two rats were used and assigned to the CA (24) and the sham (18) group. Field Strength/Sequence: T2‐weighted anatomical imaging with 2D turbo spin‐echo sequence, QSM with 3D GRE sequence, IVIM with 2D RESOLVE EPI sequence, 1 H‐MRS with SVS‐PRESS sequence. Assessment: Multiparametric MRI was performed 24 h after return of spontaneous circulation (ROSC). Imaging findings were validated using histology, immunohistochemistry, Western blot, and transmission electron microscopy. Statistical Tests: Unpaired two‐tailed Student's T ‐test was used. A p value less than 0.05 was considered statistically significant. Results: CA led to marked neurological deficits, although no obvious abnormalities were observed on T 2 ‐weighted MRI. QSM revealed marked iron accumulation in the hippocampus, which was associated with elevated ROS, lipid peroxidation, and altered expression of ferroptosis‐related markers (downregulation of GPX4 and upregulation of ACSL4). Electron microscopy confirmed mitochondrial changes characteristic of ferroptosis. Ferroptosis‐induced lipid peroxidation resulted in degradation of tight‐junction proteins (ZO‐1 and occludin), BBB leakage (elevated serum S100β), and loss of AQP4 polarity. IVIM showed a selective reduction in true tissue diffusivity (IVIM D ), indicating cytotoxic edema. 1 H‐MRS revealed decreased N ‐acetylaspartate, increased lipid peaks, and reduced myo‐inositol levels, consistent with neuronal death and astrocytic swelling. Data Conclusion: This study provides evidence for a sequential cascade after CA and ROSC in which iron overload is associated with ferroptosis, which is linked to disruption of the blood–brain barrier. This barrier disruption coincides with AQP4 depolarization, inducing cytotoxic and vasogenic edema, which, along with depolarization, is accompanied by neuronal death. Multimodal MRI noninvasively captures this process, offering an early detection and monitoring platform for ferroptosis‐related brain injury and underscoring its potential as a translational tool for neuroprotective interventions. Evidence Level: 1. Technical Efficacy: Stage 1.


  • Open Access

    How does OA affect sepsis? An observational study and Mendelian randomization mediation analysis

    by Dawei Xu, Junjie Chen, Fanrong Lin, Qingqi Meng, Yi He, Jianrong Huang, Jos Runhaar, Zhengfei Yang

  • Open Access

    Longitudinal association between chronic diseases and fall risk among middle-aged and older adults in China: the mediating role of activity limitations

    by Derong Huang, Yingjie Fu, Wen Yang, Biying Xu, Zhengfei Yang

    The World Health Organization identifies falls as the second leading cause of accidental injury deaths worldwide, following road traffic injuries [1]. As such, falls have emerged as a critical global public health issue [2], posing serious threats to the physical and mental well-being of middle-aged and older adults. The problem is particularly pronounced in […]


  • Open Access

    Hospital length of stay and the role of ICU transfer of care in postoperative oral cancer patients

    by Junpeng Tang, Yan Huang, Jinzhao Zhang, Fanrong Lin, Siqi Liu, Weipeng Liu, Jingying Hou, Zhengfei Yang

    The decision to transfer postoperative oral cancer patients to the intensive care unit (ICU) lacks standardized criteria and remains contentious. This study aimed to determine whether ICU transfer independently prolongs hospital length of stay (LOS) and identify the primary determinants of LOS in these patients. A retrospective single-center cohort study was conducted in 674 patients who underwent oral cancer surgery. Patients were categorized into the ICU and general ward groups. Data on demographics (age, sex, BMI, smoking, drinking, and betel chewing), comorbidities (hypertension, diabetes and other chronic diseases), intraoperative factors (tracheotomy, duration of operation, lymph node dissection, flap transplantation, and blood transfusion), and clinical scoring systems (SOFA, APACHE-II, CAPRINI and NRS 2002), were all analysed. Linear regression analyses were performed to identify independent determinants of LOS. Patients in the ICU group were older and had higher rates of hypertension and diabetes (p < 0.05). Between-group comparisons exhibited a higher incidence of intraoperative blood transfusion in the ICU group (p = 0.001), whereas flap transplantation rate was lower (p < 0.001). Comparisons of postoperative clinical scores showed that the ICU group had significantly higher scores of SOFA (p < 0.001), APACHE-II (p < 0.05), and CAPRINI (p < 0.05). Additionally, patients with a SOFA score > 2 or an APACHE-II score ≥ 15 accounted for a higher proportion in this group. Subsequent analysis revealed that there was a longer unadjusted LOS in the ICU group (p < 0.001). However, ICU transfer was not an independent predictor of LOS after adjustment for confounders (p > 0.05). Independent determinants of prolonged LOS included tracheotomy (p < 0.001), intraoperative blood transfusion (p < 0.001), and bilateral lymph node dissection (p < 0.001). Age-stratified analyses indicated that ICU admission independently prolonged LOS in older adults (≥ 65 years, p < 0.05), but not in younger adults (< 65 years, p > 0.05). ICU transfer is not an independent determinant of the overall LOS of postoperative oral cancer patients, instead, surgical complexity and lymph node dissection play a more critical role. In older adults, ICU transfer represents a key modifiable factor associated with prolonged LOS. These findings support a risk-stratified and age-based triage protocol to optimize ICU utilization and improve healthcare efficacy. The online version contains supplementary material available at 10.1186/s12903-026-07997-1.