1.Predicting Postoperative Progression of Ossification of the Posterior Longitudinal Ligament in the Cervical Spine Using Interpretable Radiomics Models
Siyuan QIN ; Ruomu QU ; Ke LIU ; Ruixin YAN ; Weili ZHAO ; Jun XU ; Enlong ZHANG ; Feifei ZHOU ; Ning LANG
Neurospine 2025;22(1):144-156
Objective:
This study investigates the potential of radiomics to predict postoperative progression of ossification of the posterior longitudinal ligament (OPLL) after posterior cervical spine surgery.
Methods:
This retrospective study included 473 patients diagnosed with OPLL at Peking University Third Hospital between October 2006 and September 2022. Patients underwent posterior spinal surgery and had at least 2 computed tomography (CT) examinations spaced at least 1 year apart. OPLL progression was defined as an annual growth rate exceeding 7.5%. Radiomic features were extracted from preoperative CT images of the OPLL lesions, followed by feature selection using correlation coefficient analysis and least absolute shrinkage and selection operator, and dimensionality reduction using principal component analysis. Univariable analysis identified significant clinical variables for constructing the clinical model. Logistic regression models, including the Rad-score model, clinical model, and combined model, were developed to predict OPLL progression.
Results:
Of the 473 patients, 191 (40.4%) experienced OPLL progression. On the testing set, the combined model, which incorporated the Rad-score and clinical variables (area under the receiver operating characteristic curve [AUC] = 0.751), outperformed both the radiomics-only model (AUC = 0.693) and the clinical model (AUC = 0.620). Calibration curves demonstrated good agreement between predicted probabilities and observed outcomes, and decision curve analysis confirmed the clinical utility of the combined model. SHAP (SHapley Additive exPlanations) analysis indicated that the Rad-score and age were key contributors to the model’s predictions, enhancing clinical interpretability.
Conclusion
Radiomics, combined with clinical variables, provides a valuable predictive tool for assessing the risk of postoperative progression in cervical OPLL, supporting more personalized treatment strategies. Prospective, multicenter validation is needed to confirm the utility of the model in broader clinical settings.
2.Predicting Postoperative Progression of Ossification of the Posterior Longitudinal Ligament in the Cervical Spine Using Interpretable Radiomics Models
Siyuan QIN ; Ruomu QU ; Ke LIU ; Ruixin YAN ; Weili ZHAO ; Jun XU ; Enlong ZHANG ; Feifei ZHOU ; Ning LANG
Neurospine 2025;22(1):144-156
Objective:
This study investigates the potential of radiomics to predict postoperative progression of ossification of the posterior longitudinal ligament (OPLL) after posterior cervical spine surgery.
Methods:
This retrospective study included 473 patients diagnosed with OPLL at Peking University Third Hospital between October 2006 and September 2022. Patients underwent posterior spinal surgery and had at least 2 computed tomography (CT) examinations spaced at least 1 year apart. OPLL progression was defined as an annual growth rate exceeding 7.5%. Radiomic features were extracted from preoperative CT images of the OPLL lesions, followed by feature selection using correlation coefficient analysis and least absolute shrinkage and selection operator, and dimensionality reduction using principal component analysis. Univariable analysis identified significant clinical variables for constructing the clinical model. Logistic regression models, including the Rad-score model, clinical model, and combined model, were developed to predict OPLL progression.
Results:
Of the 473 patients, 191 (40.4%) experienced OPLL progression. On the testing set, the combined model, which incorporated the Rad-score and clinical variables (area under the receiver operating characteristic curve [AUC] = 0.751), outperformed both the radiomics-only model (AUC = 0.693) and the clinical model (AUC = 0.620). Calibration curves demonstrated good agreement between predicted probabilities and observed outcomes, and decision curve analysis confirmed the clinical utility of the combined model. SHAP (SHapley Additive exPlanations) analysis indicated that the Rad-score and age were key contributors to the model’s predictions, enhancing clinical interpretability.
Conclusion
Radiomics, combined with clinical variables, provides a valuable predictive tool for assessing the risk of postoperative progression in cervical OPLL, supporting more personalized treatment strategies. Prospective, multicenter validation is needed to confirm the utility of the model in broader clinical settings.
3.Predicting Postoperative Progression of Ossification of the Posterior Longitudinal Ligament in the Cervical Spine Using Interpretable Radiomics Models
Siyuan QIN ; Ruomu QU ; Ke LIU ; Ruixin YAN ; Weili ZHAO ; Jun XU ; Enlong ZHANG ; Feifei ZHOU ; Ning LANG
Neurospine 2025;22(1):144-156
Objective:
This study investigates the potential of radiomics to predict postoperative progression of ossification of the posterior longitudinal ligament (OPLL) after posterior cervical spine surgery.
Methods:
This retrospective study included 473 patients diagnosed with OPLL at Peking University Third Hospital between October 2006 and September 2022. Patients underwent posterior spinal surgery and had at least 2 computed tomography (CT) examinations spaced at least 1 year apart. OPLL progression was defined as an annual growth rate exceeding 7.5%. Radiomic features were extracted from preoperative CT images of the OPLL lesions, followed by feature selection using correlation coefficient analysis and least absolute shrinkage and selection operator, and dimensionality reduction using principal component analysis. Univariable analysis identified significant clinical variables for constructing the clinical model. Logistic regression models, including the Rad-score model, clinical model, and combined model, were developed to predict OPLL progression.
Results:
Of the 473 patients, 191 (40.4%) experienced OPLL progression. On the testing set, the combined model, which incorporated the Rad-score and clinical variables (area under the receiver operating characteristic curve [AUC] = 0.751), outperformed both the radiomics-only model (AUC = 0.693) and the clinical model (AUC = 0.620). Calibration curves demonstrated good agreement between predicted probabilities and observed outcomes, and decision curve analysis confirmed the clinical utility of the combined model. SHAP (SHapley Additive exPlanations) analysis indicated that the Rad-score and age were key contributors to the model’s predictions, enhancing clinical interpretability.
Conclusion
Radiomics, combined with clinical variables, provides a valuable predictive tool for assessing the risk of postoperative progression in cervical OPLL, supporting more personalized treatment strategies. Prospective, multicenter validation is needed to confirm the utility of the model in broader clinical settings.
4.Predicting Postoperative Progression of Ossification of the Posterior Longitudinal Ligament in the Cervical Spine Using Interpretable Radiomics Models
Siyuan QIN ; Ruomu QU ; Ke LIU ; Ruixin YAN ; Weili ZHAO ; Jun XU ; Enlong ZHANG ; Feifei ZHOU ; Ning LANG
Neurospine 2025;22(1):144-156
Objective:
This study investigates the potential of radiomics to predict postoperative progression of ossification of the posterior longitudinal ligament (OPLL) after posterior cervical spine surgery.
Methods:
This retrospective study included 473 patients diagnosed with OPLL at Peking University Third Hospital between October 2006 and September 2022. Patients underwent posterior spinal surgery and had at least 2 computed tomography (CT) examinations spaced at least 1 year apart. OPLL progression was defined as an annual growth rate exceeding 7.5%. Radiomic features were extracted from preoperative CT images of the OPLL lesions, followed by feature selection using correlation coefficient analysis and least absolute shrinkage and selection operator, and dimensionality reduction using principal component analysis. Univariable analysis identified significant clinical variables for constructing the clinical model. Logistic regression models, including the Rad-score model, clinical model, and combined model, were developed to predict OPLL progression.
Results:
Of the 473 patients, 191 (40.4%) experienced OPLL progression. On the testing set, the combined model, which incorporated the Rad-score and clinical variables (area under the receiver operating characteristic curve [AUC] = 0.751), outperformed both the radiomics-only model (AUC = 0.693) and the clinical model (AUC = 0.620). Calibration curves demonstrated good agreement between predicted probabilities and observed outcomes, and decision curve analysis confirmed the clinical utility of the combined model. SHAP (SHapley Additive exPlanations) analysis indicated that the Rad-score and age were key contributors to the model’s predictions, enhancing clinical interpretability.
Conclusion
Radiomics, combined with clinical variables, provides a valuable predictive tool for assessing the risk of postoperative progression in cervical OPLL, supporting more personalized treatment strategies. Prospective, multicenter validation is needed to confirm the utility of the model in broader clinical settings.
5.Predicting Postoperative Progression of Ossification of the Posterior Longitudinal Ligament in the Cervical Spine Using Interpretable Radiomics Models
Siyuan QIN ; Ruomu QU ; Ke LIU ; Ruixin YAN ; Weili ZHAO ; Jun XU ; Enlong ZHANG ; Feifei ZHOU ; Ning LANG
Neurospine 2025;22(1):144-156
Objective:
This study investigates the potential of radiomics to predict postoperative progression of ossification of the posterior longitudinal ligament (OPLL) after posterior cervical spine surgery.
Methods:
This retrospective study included 473 patients diagnosed with OPLL at Peking University Third Hospital between October 2006 and September 2022. Patients underwent posterior spinal surgery and had at least 2 computed tomography (CT) examinations spaced at least 1 year apart. OPLL progression was defined as an annual growth rate exceeding 7.5%. Radiomic features were extracted from preoperative CT images of the OPLL lesions, followed by feature selection using correlation coefficient analysis and least absolute shrinkage and selection operator, and dimensionality reduction using principal component analysis. Univariable analysis identified significant clinical variables for constructing the clinical model. Logistic regression models, including the Rad-score model, clinical model, and combined model, were developed to predict OPLL progression.
Results:
Of the 473 patients, 191 (40.4%) experienced OPLL progression. On the testing set, the combined model, which incorporated the Rad-score and clinical variables (area under the receiver operating characteristic curve [AUC] = 0.751), outperformed both the radiomics-only model (AUC = 0.693) and the clinical model (AUC = 0.620). Calibration curves demonstrated good agreement between predicted probabilities and observed outcomes, and decision curve analysis confirmed the clinical utility of the combined model. SHAP (SHapley Additive exPlanations) analysis indicated that the Rad-score and age were key contributors to the model’s predictions, enhancing clinical interpretability.
Conclusion
Radiomics, combined with clinical variables, provides a valuable predictive tool for assessing the risk of postoperative progression in cervical OPLL, supporting more personalized treatment strategies. Prospective, multicenter validation is needed to confirm the utility of the model in broader clinical settings.
6.Clinical effects of total laparoscopic technique in the treatment of pelvic and acetabular fractures
Siyuan QING ; Shaobo ZHANG ; Ruizhe ZHAO ; Jie LI ; Guqi HONG ; Yu ZHANG ; Jun HU ; Xiaodong QIN
Chinese Journal of Orthopaedics 2024;44(7):463-469
Objective:To investigate the clinical efficacy of total laparoscopic technique in the treatment of pelvic and acetabular fractures.Methods:A total of 7 patients with pelvic and acetabular fractures admitted to the First Affiliated Hospital of Nanjing Medical University from April 2017 to August 2023 were retrospectively analyzed. There were 5 males and 2 females, aged 43.9±9.9 years (range, 27-56 years). There was 1 pelvic fracture (Tile B3) and 6 acetabular fractures, Letournel typing transverse in 2 cases, double column in 1 case, transverse combined anterior column in 1 case, anterior column in 1 case, and "T" shape in 1 case. The causes of injury included traffic accident in 4 cases, falling from height in 1 case, falling in 1 case, and heavy object injury in 1 case. The operation time, intraoperative blood loss, iatrogenic injuries of blood vessels, nerves, bladder and rectum, and incidence of postoperative complications were recorded. Matta imaging criteria was used to evaluate the quality of fracture reduction, and the hip function was evaluated by modified Merle d'Aubigné-Postel score.Results:All 7 patients were successfully operated and followed up for 12.0 (7.0, 72.0) months (range, 7-77 months). The operation time was 251.1±80.2 min (range, 150.0-370.0 min), and the intraoperative blood loss was 100.0 (50.0, 300.0) ml (range, 50-350 ml). There was no nerve, blood vessel, bladder or rectum injury during operation. Postoperative pelvic X-ray and CT showed no internal fixation into the joint. Anatomic reduction was achieved in 5 cases and satisfactory reduction in 2 cases. One patient with a transverse acetabular fracture had a slight separation and displacement of the posterior column, but the acetabular articular surface was flat as a whole without obvious steps and was not fixed. One patient with pelvic fracture had a comminuted anterior pelvic ring with severe displacement and did not achieve anatomical repositioning intraoperatively, but the anterior pelvic ring was a non-intra-articular fracture, and the fracture ends were immobilized with bilateral plates, and the morphology and stability of the anterior pelvic ring recovered well after the operation. The Merle d'Aubigné - Postel score at the final follow-up was 18.0 (17.0, 18.0) points (range, 17.0-18.0 points), with 4 excellent and 3 good cases. All 7 patients achieved clinical healing with a healing time of 2.5 (2.5, 3.0) months (range, 1.0-3.0 months). No delayed union or nonunion occurred. At the last follow-up, there were no complications such as internal fixation fracture, incision and internal fixation infection, and deep vein thrombosis.Conclusion:Total laparoscopic technique for the treatment of pelvic and acetabular fractures has less intraoperative blood loss, less surgical trauma, and satisfactory clinical efficacy. It can be used as a new choice of minimally invasive treatment of pelvic and acetabular fractures.
7.Anatomical characteristics of femoral intercondylar notch of knee joint for predicting non-contact anterior cruciate ligament tear
Yupeng ZHU ; Jun XU ; Qizheng WANG ; Yongye CHEN ; Siyuan QIN ; Ruixin YAN ; Peijin XIN ; Ning LANG
Chinese Journal of Medical Imaging Technology 2024;40(6):902-906
Objective To observe the value of anatomical characteristics of femoral intercondylar notch of knee joint for predicting non-contact anterior cruciate ligament tear(NC-ACLT).Methods MRI data of knee joint of 55 patients with NC-ACLT(NC-ACLT group)and 55 controls(control group)were retrospectively analyzed.The parameters of intercondylar notch,including depth,width,depth/width ratio,opening width,opening width index,area and width of the femoral condyle's outer edge at the same level were measured between groups,and the types of intercondylar notch(type A,U and W)were recorded.Univariate and multivariate logistic regression analysis were used to screen the independent impact factors of NC-ACLT.Receiver operating characteristic curves were drawn,and the area under the curves(AUC)were calculated to evaluate the efficacy of each intercondylar notch parameter for predicting NC-ACLT.Results The depth and depth/width ratio of intercondylar notch in NC-ACLT group were both higher,while the opening width and opening width index of intercondylar notch in NC-ACLT group were both lower than those in control group(all P<0.05).Multivariate logistic regression analysis revealed that the depth of intercondylar notch was an independent impact factors of NC-ACLT(P<0.05).Taken 29.55 mm in depth of intercondylar notch,1.45 in depth/width ratio of intercondylar notch,21.15 mm in opening width of intercondylar notch and 0.29 in opening width index as the optimal cut-off value,respectively,the sensitivity of the above parameters for predicting NC-ACLT was 74.55%,58.18%,67.27%and 67.27%,the specificity was 69.09%,80.00%,61.82%and 78.18%,and the AUC was 0.720,0.713,0.652 and 0.710,respectively.Conclusion The anatomical characteristics of femoral intercondylar notch of knee joint could be used to predict NC-ACLT.The depth,depth/width ratio,opening width and opening width index of intercondylar notch could be used as predictive indicators.
8.Oral squamous cell carcinomas: state of the field and emerging directions.
Yunhan TAN ; Zhihan WANG ; Mengtong XU ; Bowen LI ; Zhao HUANG ; Siyuan QIN ; Edouard C NICE ; Jing TANG ; Canhua HUANG
International Journal of Oral Science 2023;15(1):44-44
Oral squamous cell carcinoma (OSCC) develops on the mucosal epithelium of the oral cavity. It accounts for approximately 90% of oral malignancies and impairs appearance, pronunciation, swallowing, and flavor perception. In 2020, 377,713 OSCC cases were reported globally. According to the Global Cancer Observatory (GCO), the incidence of OSCC will rise by approximately 40% by 2040, accompanied by a growth in mortality. Persistent exposure to various risk factors, including tobacco, alcohol, betel quid (BQ), and human papillomavirus (HPV), will lead to the development of oral potentially malignant disorders (OPMDs), which are oral mucosal lesions with an increased risk of developing into OSCC. Complex and multifactorial, the oncogenesis process involves genetic alteration, epigenetic modification, and a dysregulated tumor microenvironment. Although various therapeutic interventions, such as chemotherapy, radiation, immunotherapy, and nanomedicine, have been proposed to prevent or treat OSCC and OPMDs, understanding the mechanism of malignancies will facilitate the identification of therapeutic and prognostic factors, thereby improving the efficacy of treatment for OSCC patients. This review summarizes the mechanisms involved in OSCC. Moreover, the current therapeutic interventions and prognostic methods for OSCC and OPMDs are discussed to facilitate comprehension and provide several prospective outlooks for the fields.
Humans
;
Carcinoma, Squamous Cell/therapy*
;
Squamous Cell Carcinoma of Head and Neck
;
Mouth Neoplasms/therapy*
;
Head and Neck Neoplasms
;
Tumor Microenvironment
9.Efficacy comparison between modified two-window iliofemoral approach and ilioinguinal approach for reduction and internal fixation of composite acetabular fracture
Guqi HONG ; Siyuan QING ; Jun HU ; Jiahu FANG ; Tianrun LYU ; Qun CHEN ; Xiaodong QIN
Chinese Journal of Trauma 2023;39(2):145-152
Objective:To compare the effect of reduction and internal fixation of composite acetabular fracture with the modified two-window iliofemoral approach and ilioinguinal approach.Methods:A retrospective cohort study was used to analyze the clinical data of 160 patients with composite acetabular fracture admitted to First Affiliated Hospital of Nanjing Medical University from January 2016 to August 2021, including 117 males and 43 females, aged 15-78 years [(44.1±16.0)years]. According to the Letournel classification system, there were 101 patients with both-column fracture, 5 with anterior wall/column combined with posterior semi-transverse fracture and 24 with T-shaped fracture. A total of 80 patients were treated using the modified iliofemoral incision combined with limited Pfannstiel incision (modified two-window iliofemoral approach group) and the other 80 patients were treated using the ilioinguinal approach (ilioinguinal approach group). The fracture healing was observed. The operation time and intraoperative bleeding volume were compared between the two groups. The quality of fracture reduction was evaluated by Matta scoring standard at 1 day and 6 months after operation. The modified Merle d′Aubigne & Postel scoring standard was used to evaluate the function of the affected hip joint at the last follow-up. The incidence of complications such as neurovascular injury, iatrogenic bladder injury, heterotopic ossification and femoral head necrosis were compared between the two groups.Results:All patients were followed up for 12-78 months [(43.3±17.9)months], with bony union of the fracture. The operation time and intraoperative bleeding volume in modified two-window iliofemoral approach group were 150.0 (123.8, 180.0)minutes and 600.0 (500.0, 787.5)ml when compared to 190.0 (150.0, 240.0)minutes and 700.0 (562.5, 887.5)ml in ilioinguinal approach group (all P<0.01). There was no significant difference between the two groups in the quality of fracture reduction at 1 day and 6 months after operation, function of hip joint at the last follow-up and incidence of complications (all P>0.05). Conclusions:For reduction and internal fixation of composite acetabular fracture, the modified two-window iliofemoral approach has advantages over the ilioinguinal approach in reducing operation time and intraoperative bleeding, although both methods yield similar results in fracture reduction quality, postoperative hip function and complication rate.
10.Repurposing econazole as a pharmacological autophagy inhibitor to treat pancreatic ductal adenocarcinoma.
Ningna WENG ; Siyuan QIN ; Jiayang LIU ; Xing HUANG ; Jingwen JIANG ; Li ZHOU ; Zhe ZHANG ; Na XIE ; Kui WANG ; Ping JIN ; Maochao LUO ; Liyuan PENG ; Edouard C NICE ; Ajay GOEL ; Suxia HAN ; Canhua HUANG ; Qing ZHU
Acta Pharmaceutica Sinica B 2022;12(7):3085-3102
Pancreatic ductal adenocarcinoma (PDAC) is characterized by the highest mortality among carcinomas. The pathogenesis of PDAC requires elevated autophagy, inhibition of which using hydroxychloroquine has shown promise. However, current realization is impeded by its suboptimal use and unpredictable toxicity. Attempts to identify novel autophagy-modulating agents from already approved drugs offer a rapid and accessible approach. Here, using a patient-derived organoid model, we performed a comparative analysis of therapeutic responses among various antimalarial/fungal/parasitic/viral agents, through which econazole (ECON), an antifungal compound, emerged as the top candidate. Further testing in cell-line and xenograft models of PDAC validated this activity, which occurred as a direct consequence of dysfunctional autophagy. More specifically, ECON boosted autophagy initiation but blocked lysosome biogenesis. RNA sequencing analysis revealed that this autophagic induction was largely attributed to the altered expression of activation transcription factor 3 (ATF3). Increased nuclear import of ATF3 and its transcriptional repression of inhibitor of differentiation-1 (ID-1) led to inactivation of the AKT/mammalian target of rapamycin (mTOR) pathway, thus giving rise to autophagosome accumulation in PDAC cells. The magnitude of the increase in autophagosomes was sufficient to elicit ER stress-mediated apoptosis. Furthermore, ECON, as an autophagy inhibitor, exhibited synergistic effects with trametinib on PDAC. This study provides direct preclinical and experimental evidence for the therapeutic efficacy of ECON in PDAC treatment and reveals a mechanism whereby ECON inhibits PDAC growth.

Result Analysis
Print
Save
E-mail