Risk factors and risk thresholds for adverse outcomes after PEID in patients with lumbar disc herniation: a retrospective cohort study

Risk factors and risk thresholds for adverse outcomes after PEID in patients with lumbar disc herniation: a retrospective cohort study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Risk factors and risk thresholds for adverse outcomes after PEID in patients with lumbar disc herniation: a retrospective cohort study Xuan Deng, Chengrui Peng, Xiuqian Wang, Li He, Hu Qian, Jun Ao This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8684995/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 4 You are reading this latest preprint version Abstract Background Percutaneous endoscopic interlaminar discectomy (PEID) is a common surgical technique for lumbar disc herniation (LDH), but the risk factors for adverse outcomes remain controversial. This study aims to develop and validate a predictive model based on machine learning algorithms to identify key clinical indicators influencing adverse outcomes after PEID. Methods This retrospective study included 414 LDH patients who underwent single-level PEID between October 2018 and June 2024. Data were divided into training (n = 290) and validation (n = 124) sets. Six machine learning algorithms were used for feature selection, identifying core indicators. Models were constructed based on these indicators and evaluated for predictive performance. Result Five core indicators were identified: Modic Changes (MC), Basal Width Of The Herniated Disc (BWHD), Body Mass Index (BMI), Ratio Of Disc Herniation (RDH), and Interspinous Ligament Injury (ILI). The XGB model performed best, with an AUC of 0.809 in the training set and 0.718 in the validation set. Risk thresholds for BWHD, BMI, and RDH were 1.715mm, 23.346kg/m², and 37.2%, respectively. Conclusion MC, ILI, BWHD, BMI, and RDH are risk factors for postoperative adverse outcomes in PEID patients. The model provides useful clinical guidance, with validated risk thresholds for key indicators. PEID adverse outcomes machine learning predictive model risk threshold Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Lumbar disc herniation (LDH) is a prevalent degenerative spinal disorder and a leading cause of low back pain and sciatica, substantially impairing patients’ quality of life, work capacity, and social functioning.[ 1 , 2 ] With advances in minimally invasive spinal surgery, percutaneous endoscopic interlaminar discectomy (PEID) has become an important treatment option for LDH owing to its minimal tissue trauma, rapid recovery, and shortened hospital stay.[ 3 ] Nevertheless, despite its overall favorable efficacy, a subset of patients still experience adverse outcomes, including persistent symptoms, recurrence, or reoperation, which continue to pose clinical challenges.[ 4 , 5 ] Recent studies have explored potential factors associated with adverse outcomes after PEID, focusing on clinical characteristics and imaging parameters such as baseline patient status, intervertebral disc degeneration, spinal stability, and intraoperative factors.[ 6 , 7 ] However, most existing studies have primarily identified associations rather than defining actionable risk thresholds, limiting their utility in guiding precise perioperative decision-making. Moreover, traditional regression-based approaches are often insufficient for modeling the complex nonlinear relationships inherent in multidimensional clinical data. As a result, robust and clinically applicable predictive tools for adverse outcomes after PEID remain scarce. With the rapid development of artificial intelligence and medical big data, machine learning has demonstrated considerable advantages in clinical prediction by efficiently extracting key features, modeling nonlinear relationships, and optimizing predictive performance.[ 8 ] Algorithms such as support vector machine (SVM) are well suited for high-dimensional nonlinear classification, while ensemble methods including extreme gradient boosting (XGB) and gradient boosting machine (GBM) have further enhanced predictive accuracy in clinically oriented models. Increasing evidence indicates that machine learning outperforms conventional statistical methods in orthopedics, imaging diagnostics, and prediction of reoperation or recurrence.[ 9 – 11 ] However, despite these advances, machine learning–based studies specifically addressing adverse outcomes after PEID remain limited, and no widely accepted predictive models have been established for this clinical scenario. Therefore, the present study enrolled patients with LDH who underwent single-level PEID between October 2018 and June 2024. By integrating multiple machine learning algorithms for feature selection and model construction, this study aimed to: (1) identify key clinical and imaging indicators associated with adverse postoperative outcomes after PEID; (2) compare the predictive performance of different machine learning models to determine the optimal approach; and (3) for the first time, define clinically meaningful risk thresholds for key continuous variables. This approach extends beyond traditional correlational analyses and provides a quantitative, clinically applicable framework to support personalized risk stratification and postoperative management in patients undergoing PEID. Material and Methods Patients population A retrospective cohort of patients with LDH who underwent PEID at our institution between October 2018 and June 2024 was enrolled in this study. All patients had imaging-confirmed LDH and received surgery after failure of conservative treatment. Inclusion criteria were: (1) a definitive diagnosis of single-segment LDH based on clinical and imaging findings; (2) treatment with PEID; (3) availability of high-quality preoperative MRI suitable for all measurements; and (4) a minimum follow-up of 18 months with complete clinical data. Exclusion criteria comprised: (1) multilevel LDH requiring simultaneous surgical intervention; (2) severe spinal stenosis or spondylolisthesis requiring open surgery; (3) follow-up < 18 months; (4) incomplete clinical or imaging data; and (5) non-degenerative lumbar conditions, including malignancy or spinal infection. Adverse outcomes were defined as clinical scenarios including, but not limited to, inadequate relief of symptoms following PEID, pain recurrence, postoperative functional limitations, surgical complications, and the need for revision surgery due to suboptimal treatment efficacy.[ 12 , 13 ] Data Collection and Processing Data collected from the hospital database primarily comprised clinical data and imaging information. Clinical variables included age, sex, smoking history, BMI, admission blood glucose level, disease onset duration, length of hospital stay, operative time, intraoperative blood loss, presence of hypertension, presence of diabetes mellitus, history of trauma, and smoking status. All imaging parameters were measured at the surgical segment, with T2-weighted images used for MRI assessments. Key imaging metrics included intervertebral disc height, vertebral body height, disc height index (DHI), BWHD, sROM, MC type, ILI, type of disc herniation, and Pfirrmann classification.[ 14 ] The RDH was defined as the ratio of BWHD to vertebral body width at the midline ( Fig. 1 ) . Machine Learning–Based Feature Selection and Model Development All enrolled patients were randomly divided into a training set (n = 290, including 65 adverse outcomes) and a validation set (n = 124, including 28 adverse outcomes) at a 7:3 ratio. In the training set, six machine learning algorithms (SVM, XGB, GBM, LASSO, KNN, and NNET) were applied for feature selection and model construction, with performance assessed using AUC, accuracy, sensitivity, specificity, kappa, and F1-score. The top three models based on AUC were selected, and the overlapping features were identified as core indicators. Models were then rebuilt using these core indicators with ten-fold cross-validation, and the algorithm with the highest mean AUC was chosen as the optimal model. Decision curve analysis evaluated clinical utility, while SHAP values and a nomogram were used to interpret feature contributions. Finally, the optimal model was validated in the validation set using the same evaluation metrics. Identification and Validation of Risk Thresholds Risk thresholds for continuous variables were determined in the training cohort using receiver operating characteristic (ROC) curve analysis. Optimal cutoff values were identified based on the Youden index (J = sensitivity + specificity − 1). The selected thresholds were subsequently validated in the validation cohort to assess their robustness and discriminatory ability. Statistical Analysis Statistical analyses were performed using SPSS Statistics (version 22.0; IBM Corp., Chicago, IL, USA) and R software (version 4.3.2). Continuous variables were compared using the independent t-test or Mann–Whitney U test, as appropriate, while categorical variables were analyzed using the χ² test. Machine learning analyses were conducted in R using the “glmnet,” “rms,” “rmda,” and “pROC” packages. All tests were two-sided, and a P value < 0.05 was considered statistically significant. Results Baseline Patient Data The flow chart of this retrospective study is presented in Fig. 2 . A total of 414 patients with LDH were included, of whom 93 (22.5%) experienced adverse outcomes and 321 (77.5%) did not. No significant differences were observed between the two groups regarding baseline characteristics such as age or smoking history (P > 0.05). In contrast, patients with adverse outcomes exhibited significantly higher BWHD, RDH, and BMI values, as well as a higher prevalence of MC and ILI (all P < 0.05). Detailed baseline characteristics are presented in Table 1 . Table 1 Results of difference analysis Variable Controls (N = 321) Cases (N = 93) Total (N = 414) P Sex 0.112 male 204 68 272 female 117 25 142 Age(year) 45.80 ± 13.68 47.17 ± 12.97 46.11 ± 13.52 0.40 Blood Glucose(mmol/L) 5.07 ± 1.57 5.22 ± 1.46 5.10 ± 1.54 0.18 BMI(kg/m²) 24.31 ± 3.71 24.74 ± 3.49 24.41 ± 3.66 0.27 Hypertension 0.15 No 280 75 355 Yes 41 18 59 Diabetes Mellitus 0.08 No 300 81 381 Yes 21 12 33 Smoking 0.44 No 259 71 330 Yes 62 22 84 LHS 7.44 ± 4.07 7.95 ± 4.00 7.56 ± 4.05 0.30 Disease Duration(day) 26.83 ± 45.57 29.98 ± 39.93 27.54 ± 44.34 0.06 Blood Loss(ml) 18.04 ± 11.07 19.35 ± 8.12 18.34 ± 10.48 0.03 Operation Time(min) 101.95 ± 36.79 103.17 ± 40.02 102.22 ± 37.50 0.90 Pfirrmann Classification 0.75 Ⅰ 2 0 2 Ⅱ 9 3 12 Ⅲ 86 27 113 Ⅳ 184 48 232 Ⅴ 40 15 55 IDH 11.62 ± 2.46 11.69 ± 2.30 11.64 ± 2.42 0.80 VBH 29.97 ± 3.32 29.88 ± 3.37 29.95 ± 3.327 0.83 DHI 0.39 ± 0.08 0.39 ± 0.08 0.39 ± 0.08 0.63 BWHD 1.63 ± 0.65 2.06 ± 0.79 1.72 ± 0.70 0.00 RDH 0.33 ± 0.11 0.37 ± 0.11 0.34 ± 0.11 0.01 SS(°) 33.11 ± 10.07 34.40 ± 10.54 33.40 ± 10.18 0.28 ALL 0.50 L2 2 1 3 L2/3 16 6 22 L3 18 9 27 L3/4 66 18 84 L4 88 32 120 L4/5 67 15 82 L5 47 9 56 L5/S1 17 3 20 MC 0.00 No Modic changes 291 73 364 Grade I 18 14 32 Grade II 11 5 16 Grade III 1 1 2 Left-right Distribution 0.00 Left side 153 50 203 Right side 121 31 152 Central 51 12 63 Far lateral 5 0 5 Lee Classification 0.83 Entrance zone 277 81 358 Mid zone 33 8 41 Exit zone 11 4 15 Lateral Cobb Angle 24.24 ± 12.63 24.75 ± 12.06 24.36 ± 12.49 0.73 Lumbar Spine ROM on X-ray 22.82 ± 11.89 21.44 ± 12.99 22.51 ± 12.15 0.33 Lumbar Extension Cobb Angle on X-ray 32.15 ± 12.77 32.98 ± 13.03 32.34 ± 12.82 0.58 Lumbar Flexion Cobb Angle on X-ray 9.39 ± 13.14 11.53 ± 14.68 9.87 ± 13.51 0.18 Responsible Segment ROM 5.87 ± 5.22 6.279 ± 3.88 5.96 ± 4.95s 0.19 MMFI 17.10 ± 10.61 15.80 ± 7.80 16.81 ± 10.10 0.76 AAS 0.59 No 259 72 331 Yes 62 21 83 Zygapophyseal Joint Hypertrophy 0.69 No 293 83 376 Yes 28 10 38 Calcification 0.59 No 259 72 331 Yes 62 21 83 ILI 0.00 No 294 68 362 Yes 27 25 52 * BMI :Body Mass Index LHS :Length of Hospital Stay FHS :First Herniated Segment IDH :Intervertebral Disc Height VBH :Vertebral Body Height DHI :Disc Height Index BWHD :Basal Width of Herniated Disc RDH :Ratio Of Disc Herniation SS :Sacral Slope ALL :Apex of Lumbar Lordosis MC :Modic Changes ROM :Range of Motion AAS :Abdominal Aortic Sclerosis ILI :Interspinous Ligament Injury MMFI :Multifidus Muscle Fat Infiltration Rate FPT :first protruding type Performance of Adverse Outcome Prediction Models in the Training and Validation Cohorts Six machine learning algorithms were used for feature selection, and the three algorithms with the highest AUC values—LASSO (AUC = 0.681), XGB (AUC = 0.701), and GBM (AUC = 0.722) ( Fig. 3 A ) —were selected as candidate algorithms. In the LASSO model, 12 core features were identified at the optimal lambda value ( Fig. 3 B ) ; in the GBM model, the top 10 features by importance were defined as key features ( Fig. 3 C ) ; in the XGB model, the top 10 features by importance were designated as key features ( Fig. 3 D ) . The intersection of these features yielded 5 core indicators, including MC, BWHD, BMI, RDH, and ILI ( Fig. 3 E ) . Using these core features, models based on five machine learning algorithms were constructed in the training set. After ten-fold cross-validation, the XGB model achieved the highest AUC value (0.809) ( Fig. 4 A ) , and it also exhibited a high AUC value (0.718) in the validation set ( Fig. 4 B ) . Detailed performance metrics, including accuracy, sensitivity, specificity, kappa coefficient, and F1-score, are summarized in Table 2 . Decision curve analysis indicated that the XGB model provided superior net clinical benefit across a broad range of threshold probabilities in both cohorts ( Figs. 4 C and 4 D ) . SHAP analysis demonstrated that all five core indicators contributed positively to adverse outcome risk ( Fig. 4 E ) . A nomogram was subsequently constructed to visualize the relative contribution of each predictor ( Fig. 4 F ) . Table 2 10-Fold Cross Validation Yielded the Average Scores of the Individual Models. Classifier AUC Sensitivity Specificity Accuracy Kappa F1 GBM 0.76(± 0.02) 0.17(± 0.07) 0.99(± 0.01) 0.80(± 0.01) 0.21(± 0.09) 0.30(± 0.06) KNN 0.70(± 0.06) 0.06(± 0.06) 0.95(± 0.02) 0.75(± 0.02) 0.01(± 0.09) 0.15(± 0.07) NNET 0.60(± 0.10) 0.06(± 0.11) 0.98(± 0.05) 0.77(± 0.02) 0.04(± 0.07) 0.25(± 0.09) SVM 0.59(± 0.07) 0.05(± 0.07) 0.98(± 0.03) 0.78(± 0.01) 0.05(± 0.07) 0.21(± 0.01) XGB 0.77(± 0.03) 0.23(± 0.09) 0.98(± 0.01) 0.82(± 0.01) 0.28(± 0.11) 0.39(± 0.04) LASSO 0.76(± 0.04) 0.07(± 0.03) 0.99(± 0.01) 0.79(± 0.00) 0.09(± 0.03) 0.13(± 0.03) Note: GBM: Gradient Boosting Machine, KNN: K-Nearest Neighbors, NNET: Neural Network, SVM: Support Vector Machine, XGB: Extreme Gradient Boosting, LASSO: Least Absolute Shrinkage and Selection Operator, AUC: Area Under the ROC Curve Identification of Risk Thresholds To enhance clinical applicability, risk thresholds for continuous variables were determined using the optimal XGB model and the Youden index. BWHD ≥ 1.715 mm was identified as a high-risk threshold, associated with a 1.7-fold increase in adverse outcomes in the training cohort and a 1.6-fold increase in the validation cohort (Figs. 5 A–C). RDH ≥ 37.2% was associated with a 0.8-fold and 1.1-fold increase in adverse outcomes in the training and validation cohorts, respectively (Figs. 5 D–F). BMI ≥ 23.346 kg/m² was also identified as a high-risk threshold, increasing adverse outcome incidence by 0.8-fold in the training cohort and 0.6-fold in the validation cohort (Figs. 5 G–I). All differences were statistically significant. Collectively, these findings establish a quantitative, threshold-based risk stratification framework based on preoperative clinical and MRI indicators, enabling the translation of predictive features into clinically actionable decision thresholds with consistent performance across both cohorts. Discussion Although percutaneous endoscopic interlaminar discectomy (PEID) is widely regarded as an effective minimally invasive treatment for lumbar disc herniation (LDH), postoperative adverse outcomes—including recurrence, complications, and delayed recovery—are still reported in 3.6–12.4% of patients.[ 15 , 16 ] Accurate prediction of such outcomes therefore remains a critical clinical challenge. In this study, we developed and validated a machine learning–based predictive model for adverse outcomes following PEID using a cohort of 414 patients. Several important findings emerged. First, a robust prediction model was established based on five core indicators and demonstrated favorable discrimination and clinical utility. Second, Modic changes (MC), basal width of the herniated disc (BWHD), body mass index (BMI), ratio of disc herniation (RDH), and interspinous ligament injury (ILI) were identified as key contributors to adverse postoperative outcomes. Third, clinically actionable risk thresholds for BWHD, RDH, and BMI were quantitatively defined for the first time. Among the identified predictors, MC emerged as a prominent imaging factor associated with adverse outcomes after PEID, showing a strong positive contribution in both SHAP analyses and nomogram visualization. MC reflect pathological alterations of the vertebral endplates and adjacent bone marrow and are considered markers of segmental degeneration and biomechanical imbalance.[ 17 ] From a pathological perspective, MC represent chronic reactive changes characterized by inflammatory infiltration, fatty degeneration, fibrosis, and impaired micro-injury repair, which may hinder optimal stress redistribution after decompression. Moreover, MC are often accompanied by disc height loss, reduced segmental stability, and disturbances in the local immune microenvironment, potentially compromising postoperative recovery and symptom resolution. These suggest that patients with preoperative MC may benefit from more cautious postoperative management, including delayed weight bearing, tailored rehabilitation programs, and extended follow-up to facilitate early detection of persistent pain or recurrent symptoms. Interspinous ligament (ISL) injury was another key predictor of adverse outcomes, underscoring the importance of posterior ligamentous complex (PLC) integrity in postoperative recovery. As a critical stabilizing structure, injury to the interspinous ligament disrupts normal biomechanical balance, increases intervertebral micromotion, and alters load transmission patterns.[ 18 ] These changes may provoke persistent mechanical irritation, local inflammation, and compensatory paravertebral muscle tension, thereby contributing to prolonged pain and delayed functional recovery. While previous studies have linked ILI to postoperative pain and instability, its role in comprehensive adverse outcomes—including recurrence and delayed recovery—has been insufficiently quantified.[ 19 ] Notably, ILI can be reliably assessed on preoperative MRI and is relatively unaffected by postoperative behavioral factors, making it a stable and clinically valuable risk marker. In the context of PEID, where soft tissue disruption is minimal, pre-existing ILI may unmask occult instability and compromise postoperative outcomes. Enhanced postoperative posture control, delayed loading, and prolonged rehabilitation may therefore be warranted in these patients. This study also confirmed BWHD and RDH as important quantitative imaging indicators influencing adverse outcomes after PEID. Unlike traditional descriptors such as herniation type or volume, basal morphology parameters more accurately reflect the structural relationship between annulus fibrosus rupture and herniated disc material.[ 20 ] A wider herniated base likely represents a larger annular defect and impaired segmental stability, which may limit postoperative stress redistribution and delay recovery. Consistent with previous biomechanical hypotheses that wide-based herniations are more difficult to stabilize,[ 21 ] our results demonstrated that increased BWHD and RDH were associated not only with recurrence but also with suboptimal pain relief and delayed functional improvement. Importantly, RDH normalizes BWHD to vertebral body size, enhancing interindividual comparability and generalizability. By applying quantitative measurements and machine learning–based modeling, we identified BWHD ≥ 1.715 mm and RDH ≥ 37.2% as high-risk thresholds, with patients exceeding these values exhibiting markedly higher risks of adverse outcomes. These findings provide objective, preoperatively accessible indicators that may be readily incorporated into clinical risk stratification. BMI was another significant clinical predictor identified in this study. Although elevated BMI has long been associated with poorer outcomes in spinal surgery, its impact on minimally invasive procedures such as PEID has often been underestimated.[ 22 , 23 ] Our model revealed that adverse effects of BMI emerged at a lower threshold (23.346 kg/m²) than traditionally reported in open or fusion surgeries, where thresholds typically exceed 25–30 kg/m².[ 24 ] This finding suggests that PEID, which relies on limited posterior working space and endoscopic channel stability, may be more sensitive to biomechanical loading. Even mild overweight status may increase axial load and flexion torque, placing continuous stress on postoperative tissues and hindering recovery. Thus, the identified BMI threshold represents not only a statistically optimal cutoff but also a biomechanically meaningful “sensitive turning point” for minimally invasive spinal surgery, highlighting the importance of early weight management and individualized rehabilitation strategies. Collectively, this study advances current understanding of adverse outcomes after PEID by integrating multidimensional clinical and imaging variables into an interpretable machine learning framework. Unlike previous studies that primarily focused on qualitative risk factors, this work quantitatively defined actionable risk thresholds and translated predictive features into a clinically usable nomogram. By combining SHAP interpretation with decision curve analysis, the model bridges the gap between predictive accuracy and clinical applicability, offering a practical tool for individualized risk assessment and postoperative management. Several limitations should be acknowledged. First, this was a single-center retrospective study, and multicenter prospective validation is required to confirm model generalizability. Second, although SHAP improved interpretability, unmeasured confounding factors may still influence model performance. Finally, variables such as psychological status, occupational workload, and postoperative rehabilitation compliance were not included and may further refine predictive accuracy in future studies. In conclusion, this study demonstrates that MC, BWHD and RDH, BMI, and ILI are key determinants of adverse outcomes following PEID. By defining clinically meaningful risk thresholds and constructing an interpretable prediction model, this work provides a data-driven foundation for personalized risk stratification and optimized postoperative management. Further large-scale validation may facilitate the transition of PEID outcome prediction from empirical judgment to precision-based decision-making. Conclusion This study identified five key predictors of adverse outcome risk after PEID: Modic Changes (MC), Basal Width Of The Herniated Disc (BWHD), Body Mass Index (BMI), Ratio Of Disc Herniation(RDH), and Interspinous Ligament Injury (ILI). The XGBoost model exhibited stable performance in both the training set (AUC = 0.809) and validation set (AUC = 0.718), and combined with SHAP values and a Nomogram, it enabled interpretable individualized risk prediction. Meanwhile, this study, for the first time, integrated these core indicators with specific thresholds, achieving an innovative transition from “feature identification” to “quantification of risk thresholds”. Declarations Data availability statement/Availability of data and material All data and materials are available. The corresponding author can provide the datasets used and/or analyzed during the current study upon reasonable request. Funding statement This study was supported by the National Natural Science Foundation of China (82360430), science and technology Bureau project of Zunyi city (HZ2023-195), Guizhou Provincial Basic Research Program (Natural Science) (Grant no. Qian Jiao Ke He Jichu-ZK[2023]-yiban 584), The Youth Science and Technology Talent Support Program of Guizhou Provincial Association for Science and Technology (GASTYESS202514). Conflict of interest disclosure The authors declare no conflicts of interest for this work. Ethics approval statement The norms and standards on which this study is based are the Declaration of Helsinki. The ethics of this study have been approved by the Biomedical Research Ethics Committee of the Affiliated Hospital of Zunyi Medical University, Ethics Review Approval Number: KLL-2025-597 Consent for publication/ Patient consent statement The publication of this article has obtained the informed consent of the research subjects and participants. Permission to reproduce material from other sources All data and materials are available. The corresponding authors can provide the datasets used and/or analyzed during the current study upon reasonable request. Authors' contributions LH, JA and HQ: conceptualization; XD, CP and XW: data curation; XD and CP: formal analysis; JA and HQ: funding acquisition; LH, JA and HQ: investigation; XD and CP: methodology; LH, JA and HQ: project administration; JA and HQ: resources; XD and CP: software; LH, JA and HQ: supervision; XD and CP: validation; XD and CP: visualization; XD, CP, XW, LH, JA and HQ: writing – original draft; XD, CP, XW, LH, JA and HQ: writing – review and editing. All authors read and approved the final manuscript. Acknowledgments: All authors thank the hospital staff for their guidance, support, and collaboration. References Zhang AS, Xu A, Ansari K et al (2023) Lumbar Disc Herniation: Diagnosis and Management. Am J Med 136:645–651. https://doi.org/10.1016/j.amjmed.2023.03.024 Modic MT, Ross JS (2007) Lumbar degenerative disk disease. Radiology 245:43–61. https://doi.org/10.1148/radiol.2451051706 Liu S, Zhang X, Xiong Y, He H (2025) Minimally invasive surgery for lumbar disc herniation: a meta-analysis of efficacy and safety. Int J Surg Lond Engl 111:5623–5636. https://doi.org/10.1097/JS9.0000000000002434 Xu J, Li Y, Wang B et al (2020) Minimum 2-Year Efficacy of Percutaneous Endoscopic Lumbar Discectomy versus Microendoscopic Discectomy: A Meta-Analysis. World Neurosurg 138:19–26. https://doi.org/10.1016/j.wneu.2020.02.096 Kang TW, Park SY, Oh H et al (2021) Risk of reoperation and infection after percutaneous endoscopic lumbar discectomy and open lumbar discectomy: a nationwide population-based study. Bone Jt J 103–B:1392–1399. https://doi.org/10.1302/0301-620X.103B8.BJJ-2020-2541.R2 Jiang X, Gu L, Xu G et al (2023) Nomogram for predicting the unfavourable outcomes of percutaneous endoscopic transforaminal discectomy for lumbar disc herniation: a retrospective study. Front Surg 10:1188517. https://doi.org/10.3389/fsurg.2023.1188517 Huang H, Yang M, Fu Z et al (2024) Predicting and analysing of the unfavourable outcomes of early applicated percutaneous endoscopic interlaminar discectomy for lumbar disc herniation: development and validation based nomogram. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 33:906–914. https://doi.org/10.1007/s00586-024-08141-9 Topol EJ (2019) High-performance medicine: the convergence of human and artificial intelligence. Nat Med 25:44–56. https://doi.org/10.1038/s41591-018-0300-7 Tang M, Wang S, Wang Y et al (2024) Nomogram Development and Validation for Predicting Postoperative Recurrent Lumbar Disc Herniation Based on Paraspinal Muscle Parameters. J Pain Res 17:2121–2131. https://doi.org/10.2147/JPR.S459846 Huang H, Yang M, Fu Z et al (2024) Predicting and analysing of the unfavourable outcomes of early applicated percutaneous endoscopic interlaminar discectomy for lumbar disc herniation: development and validation based nomogram. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 33:906–914. https://doi.org/10.1007/s00586-024-08141-9 Tang M, Wang S, Wang Y et al (2024) Development and Validation of a Nomogram Predicting Postoperative Recurrent Lumbar Disc Herniation Based on Activity Factors. Risk Manag Healthc Policy 17:689–699. https://doi.org/10.2147/RMHP.S453819 Shen Z, Zhong Z-M, Wu Q et al (2019) Predictors for Poor Outcomes After Percutaneous Endoscopic Lumbar Discectomy: A Retrospective Study of 241 Patients. World Neurosurg 126:e422–e431. https://doi.org/10.1016/j.wneu.2019.02.068 Aljawadi A, Sethi G, Islam A et al (2020) Sciatica Presentations and Predictors of Poor Outcomes Following Surgical Decompression of Herniated Lumbar Discs: A Review Article. Cureus 12:e11605. https://doi.org/10.7759/cureus.11605 Shi H, Zhu L, Jiang Z-L, Wu X-T (2021) Radiological risk factors for recurrent lumbar disc herniation after percutaneous transforaminal endoscopic discectomy: a retrospective matched case-control study. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 30:886–892. https://doi.org/10.1007/s00586-020-06674-3 Yin S, Du H, Yang W et al (2018) Prevalence of Recurrent Herniation Following Percutaneous Endoscopic Lumbar Discectomy: A Meta-Analysis. Pain Physician 21:337–350 Kim HS, You JD, Ju CI (2019) Predictive Scoring and Risk Factors of Early Recurrence after Percutaneous Endoscopic Lumbar Discectomy. BioMed Res Int 2019:6492675. https://doi.org/10.1155/2019/6492675 Huang H, Yang M, Fu Z et al (2024) Predicting and analysing of the unfavourable outcomes of early applicated percutaneous endoscopic interlaminar discectomy for lumbar disc herniation: development and validation based nomogram. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 33:906–914. https://doi.org/10.1007/s00586-024-08141-9 Mo AZ, Lockey S, Mo F (2025) The Posterior Ligamentous Complex: Anatomic and Biomechanical Considerations in Injury Classification and Management. J Am Acad Orthop Surg 33:e1332–e1342. https://doi.org/10.5435/JAAOS-D-22-00908 Jiang J, Chen M, Huang D-A et al (2024) High intensity in interspinous ligaments: a diagnostic sign of lumbar instability and back pain for degenerative lumbar spondylolisthesis. BMC Musculoskelet Disord 25:949. https://doi.org/10.1186/s12891-024-08081-x Gülensoy B, Güzel E, Kuzu Kumcu M et al (2023) Recurrence of lumbar disk herniation after microdiscectomy: a two-center retrospective analysis of 1214 cases and identification of risk factors. Turk J Med Sci 53:1254–1261. https://doi.org/10.55730/1300-0144.5691 Li Z-P, Liu L-L, Liu H et al (2024) Radiologic Analysis of Causes of Early Recurrence After Percutaneous Endoscopic Transforaminal Discectomy. Glob Spine J 14:113–121. https://doi.org/10.1177/21925682221096061 Chen XT, Shahrestani S, Ballatori AM et al (2021) The Influence of Body Mass Index in Obese and Morbidly Obese Patients on Complications and 30- and 90-day Readmissions Following Lumbar Spine Fusion. Spine 46:965–972. https://doi.org/10.1097/BRS.0000000000003933 Patel N, Bagan B, Vadera S et al (2007) Obesity and spine surgery: relation to perioperative complications. J Neurosurg Spine 6:291–297. https://doi.org/10.3171/spi.2007.6.4.1 Katsevman GA, Daffner SD, Brandmeir NJ et al (2020) Complexities of spine surgery in obese patient populations: a narrative review. Spine J Off J North Am Spine Soc 20:501–511. https://doi.org/10.1016/j.spinee.2019.12.011 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 29 Jan, 2026 Editor assigned by journal 29 Jan, 2026 Submission checks completed at journal 28 Jan, 2026 First submitted to journal 24 Jan, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8684995","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":582738503,"identity":"50eab12a-24f1-4071-94fe-5be0df47bbb5","order_by":0,"name":"Xuan Deng","email":"","orcid":"","institution":"Affiliated Hospital of Zunyi Medical College","correspondingAuthor":false,"prefix":"","firstName":"Xuan","middleName":"","lastName":"Deng","suffix":""},{"id":582738504,"identity":"ab5f2b5a-4a7c-47e0-8e4e-7db92bcaf17f","order_by":1,"name":"Chengrui Peng","email":"","orcid":"","institution":"Affiliated Hospital of Zunyi Medical College","correspondingAuthor":false,"prefix":"","firstName":"Chengrui","middleName":"","lastName":"Peng","suffix":""},{"id":582738505,"identity":"e4e82761-7060-4316-b2e3-397d7a8304c7","order_by":2,"name":"Xiuqian Wang","email":"","orcid":"","institution":"Affiliated Hospital of Zunyi Medical College","correspondingAuthor":false,"prefix":"","firstName":"Xiuqian","middleName":"","lastName":"Wang","suffix":""},{"id":582738506,"identity":"3f13cf51-c4d3-40e1-9101-da57be280a45","order_by":3,"name":"Li He","email":"","orcid":"","institution":"Affiliated Hospital of Zunyi Medical College","correspondingAuthor":false,"prefix":"","firstName":"Li","middleName":"","lastName":"He","suffix":""},{"id":582738507,"identity":"e931bd21-ca9d-445f-9ea0-413267d3e2b0","order_by":4,"name":"Hu Qian","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3UlEQVRIiWNgGAWjYNCCAgYDMP3BwMaOSC0GEC2MMwrSkknTwszz4RBjAyHFuu29Bx/zGNgZ87MffyZtY3CAmYH98NEN+LSYnTmXbMxjkGwm2ZNjJp1jcIePgSct7QZeLTeAKnkMmG0MbvCwAbU8Y2aQ4DEjpMX8N49BvY39DfZn0hYGhxkbiNBixsxjcNjMQILBTJqBKC1nzhhLzjE4bixxJsfYsscgLZmNoF+O9xh+eFNRbdjffvzhjR9/bOz42Q8fw6sFBJh4kHlshJSDAOMPYlSNglEwCkbByAUAbLBD0j2p6AgAAAAASUVORK5CYII=","orcid":"","institution":"Affiliated Hospital of Zunyi Medical College","correspondingAuthor":true,"prefix":"","firstName":"Hu","middleName":"","lastName":"Qian","suffix":""},{"id":582738508,"identity":"d6136d13-a5b8-474f-8d1e-29efca4d8392","order_by":5,"name":"Jun Ao","email":"","orcid":"","institution":"Affiliated Hospital of Zunyi Medical College","correspondingAuthor":false,"prefix":"","firstName":"Jun","middleName":"","lastName":"Ao","suffix":""}],"badges":[],"createdAt":"2026-01-24 08:24:52","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8684995/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8684995/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102940103,"identity":"d18bca78-31d4-4c44-af85-bea1728588c1","added_by":"auto","created_at":"2026-02-18 17:03:39","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":768087,"visible":true,"origin":"","legend":"\u003cp\u003eVisual representation of RDH(A/B).\u003c/p\u003e","description":"","filename":"Fig.1.png","url":"https://assets-eu.researchsquare.com/files/rs-8684995/v1/68d62e625d455956dc68323e.png"},{"id":102940108,"identity":"3f9c7a7a-a864-4c49-8292-281640672175","added_by":"auto","created_at":"2026-02-18 17:03:40","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":812001,"visible":true,"origin":"","legend":"\u003cp\u003eFlow Chart of This Study.\u003c/p\u003e","description":"","filename":"Fig.2.png","url":"https://assets-eu.researchsquare.com/files/rs-8684995/v1/a4fe6ba1c7c1e05f16176aa0.png"},{"id":102964200,"identity":"5e06f10f-ae1b-4de3-95d8-3c94ab29c255","added_by":"auto","created_at":"2026-02-19 04:21:43","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":146600,"visible":true,"origin":"","legend":"\u003cp\u003eScreening of Core Indicators. (A) ROC curves and area under the curve (AUC) of the five machine learning models. (B) Lambda value distribution map of LASSO under different feature numbers. (C) Indicator importance ranking plot of the GBM model. (D) Indicator importance ranking plot of the XGB model. (E) Intersection plot of indicators screened by the three machine learning algorithms.\u003c/p\u003e","description":"","filename":"Fig.3.png","url":"https://assets-eu.researchsquare.com/files/rs-8684995/v1/d345b1b8be08734058fcadd7.png"},{"id":102964324,"identity":"94f7748b-efd3-491d-a890-c24d35f6b8f7","added_by":"auto","created_at":"2026-02-19 04:22:04","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":135100,"visible":true,"origin":"","legend":"\u003cp\u003eConstruction and Validation of the Core Indicator Prediction Model. (A) ROC curve and AUC of the XGB model in the training set. (B) ROC curve and AUC of the XGB model in the validation set. (C) Prediction decision curve of the XGB model in the training set. (D) Prediction decision curve of the XGB model in the validation set. (E) SHAP bee swarm plot of feature importance for the XGB model in the training set. (F) Nomogram of the prediction model.\u003c/p\u003e","description":"","filename":"Fig.4.png","url":"https://assets-eu.researchsquare.com/files/rs-8684995/v1/7627f56df9120ad487417626.png"},{"id":102940105,"identity":"306b3b60-d06a-4c32-b44d-73de44b76619","added_by":"auto","created_at":"2026-02-18 17:03:39","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":157359,"visible":true,"origin":"","legend":"\u003cp\u003eScreening and Validation of Risk Thresholds for Continuous Variables. (A) Distribution of BWHD and confirmation of the risk threshold. (B) Risk comparison before and after the threshold of BWHD in the training set. (C) Risk comparison before and after the threshold of BWHD in the validation set. (D) Distribution of BMI and confirmation of the risk threshold. (E) Risk comparison before and after the BMI threshold in the training set. (F) Risk comparison before and after the BMI threshold in the validation set. (G) Distribution of RDH and confirmation of the risk threshold. (H) Risk comparison before and after the threshold of RDH in the training set. (I) Risk comparison before and after the threshold of RDH in the validation set.\u003c/p\u003e","description":"","filename":"Fig.5.png","url":"https://assets-eu.researchsquare.com/files/rs-8684995/v1/858eedf9cd0aad62365213bd.png"},{"id":102965480,"identity":"97232e1b-e87f-48dc-ac59-f2dca7e5568f","added_by":"auto","created_at":"2026-02-19 04:31:39","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3106152,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8684995/v1/12b9931b-aa9a-4e99-8cdf-f93fc25afdd4.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Risk factors and risk thresholds for adverse outcomes after PEID in patients with lumbar disc herniation: a retrospective cohort study","fulltext":[{"header":"Introduction","content":"\u003cp\u003eLumbar disc herniation (LDH) is a prevalent degenerative spinal disorder and a leading cause of low back pain and sciatica, substantially impairing patients\u0026rsquo; quality of life, work capacity, and social functioning.[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e] With advances in minimally invasive spinal surgery, percutaneous endoscopic interlaminar discectomy (PEID) has become an important treatment option for LDH owing to its minimal tissue trauma, rapid recovery, and shortened hospital stay.[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e] Nevertheless, despite its overall favorable efficacy, a subset of patients still experience adverse outcomes, including persistent symptoms, recurrence, or reoperation, which continue to pose clinical challenges.[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]\u003c/p\u003e \u003cp\u003eRecent studies have explored potential factors associated with adverse outcomes after PEID, focusing on clinical characteristics and imaging parameters such as baseline patient status, intervertebral disc degeneration, spinal stability, and intraoperative factors.[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e] However, most existing studies have primarily identified associations rather than defining actionable risk thresholds, limiting their utility in guiding precise perioperative decision-making. Moreover, traditional regression-based approaches are often insufficient for modeling the complex nonlinear relationships inherent in multidimensional clinical data. As a result, robust and clinically applicable predictive tools for adverse outcomes after PEID remain scarce.\u003c/p\u003e \u003cp\u003eWith the rapid development of artificial intelligence and medical big data, machine learning has demonstrated considerable advantages in clinical prediction by efficiently extracting key features, modeling nonlinear relationships, and optimizing predictive performance.[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] Algorithms such as support vector machine (SVM) are well suited for high-dimensional nonlinear classification, while ensemble methods including extreme gradient boosting (XGB) and gradient boosting machine (GBM) have further enhanced predictive accuracy in clinically oriented models. Increasing evidence indicates that machine learning outperforms conventional statistical methods in orthopedics, imaging diagnostics, and prediction of reoperation or recurrence.[\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e] However, despite these advances, machine learning\u0026ndash;based studies specifically addressing adverse outcomes after PEID remain limited, and no widely accepted predictive models have been established for this clinical scenario.\u003c/p\u003e \u003cp\u003eTherefore, the present study enrolled patients with LDH who underwent single-level PEID between October 2018 and June 2024. By integrating multiple machine learning algorithms for feature selection and model construction, this study aimed to: (1) identify key clinical and imaging indicators associated with adverse postoperative outcomes after PEID; (2) compare the predictive performance of different machine learning models to determine the optimal approach; and (3) for the first time, define clinically meaningful risk thresholds for key continuous variables. This approach extends beyond traditional correlational analyses and provides a quantitative, clinically applicable framework to support personalized risk stratification and postoperative management in patients undergoing PEID.\u003c/p\u003e"},{"header":"Material and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePatients population\u003c/h2\u003e \u003cp\u003eA retrospective cohort of patients with LDH who underwent PEID at our institution between October 2018 and June 2024 was enrolled in this study. All patients had imaging-confirmed LDH and received surgery after failure of conservative treatment. Inclusion criteria were: (1) a definitive diagnosis of single-segment LDH based on clinical and imaging findings; (2) treatment with PEID; (3) availability of high-quality preoperative MRI suitable for all measurements; and (4) a minimum follow-up of 18 months with complete clinical data. Exclusion criteria comprised: (1) multilevel LDH requiring simultaneous surgical intervention; (2) severe spinal stenosis or spondylolisthesis requiring open surgery; (3) follow-up \u0026lt;\u0026thinsp;18 months; (4) incomplete clinical or imaging data; and (5) non-degenerative lumbar conditions, including malignancy or spinal infection. Adverse outcomes were defined as clinical scenarios including, but not limited to, inadequate relief of symptoms following PEID, pain recurrence, postoperative functional limitations, surgical complications, and the need for revision surgery due to suboptimal treatment efficacy.[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eData Collection and Processing\u003c/h3\u003e\n\u003cp\u003eData collected from the hospital database primarily comprised clinical data and imaging information. Clinical variables included age, sex, smoking history, BMI, admission blood glucose level, disease onset duration, length of hospital stay, operative time, intraoperative blood loss, presence of hypertension, presence of diabetes mellitus, history of trauma, and smoking status. All imaging parameters were measured at the surgical segment, with T2-weighted \u003cb\u003eimages\u003c/b\u003e used for MRI assessments. Key imaging metrics included intervertebral disc height, vertebral body height, disc height index (DHI), BWHD, sROM, MC type, ILI, type of disc herniation, and Pfirrmann classification.[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e] The RDH was defined as the ratio of BWHD to vertebral body width at the midline \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u003cb\u003e)\u003c/b\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eMachine Learning–Based Feature Selection and Model Development\u003c/h3\u003e\n\u003cp\u003eAll enrolled patients were randomly divided into a training set (n\u0026thinsp;=\u0026thinsp;290, including 65 adverse outcomes) and a validation set (n\u0026thinsp;=\u0026thinsp;124, including 28 adverse outcomes) at a 7:3 ratio. In the training set, six machine learning algorithms (SVM, XGB, GBM, LASSO, KNN, and NNET) were applied for feature selection and model construction, with performance assessed using AUC, accuracy, sensitivity, specificity, kappa, and F1-score. The top three models based on AUC were selected, and the overlapping features were identified as core indicators. Models were then rebuilt using these core indicators with ten-fold cross-validation, and the algorithm with the highest mean AUC was chosen as the optimal model. Decision curve analysis evaluated clinical utility, while SHAP values and a nomogram were used to interpret feature contributions. Finally, the optimal model was validated in the validation set using the same evaluation metrics.\u003c/p\u003e\n\u003ch3\u003eIdentification and Validation of Risk Thresholds\u003c/h3\u003e\n\u003cp\u003eRisk thresholds for continuous variables were determined in the training cohort using receiver operating characteristic (ROC) curve analysis. Optimal cutoff values were identified based on the Youden index (J\u0026thinsp;=\u0026thinsp;sensitivity\u0026thinsp;+\u0026thinsp;specificity\u0026thinsp;\u0026minus;\u0026thinsp;1). The selected thresholds were subsequently validated in the validation cohort to assess their robustness and discriminatory ability.\u003c/p\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eStatistical analyses were performed using SPSS Statistics (version 22.0; IBM Corp., Chicago, IL, USA) and R software (version 4.3.2). Continuous variables were compared using the independent t-test or Mann\u0026ndash;Whitney U test, as appropriate, while categorical variables were analyzed using the χ\u0026sup2; test. Machine learning analyses were conducted in R using the \u0026ldquo;glmnet,\u0026rdquo; \u0026ldquo;rms,\u0026rdquo; \u0026ldquo;rmda,\u0026rdquo; and \u0026ldquo;pROC\u0026rdquo; packages. All tests were two-sided, and a P value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eBaseline Patient Data\u003c/h2\u003e \u003cp\u003eThe flow chart of this retrospective study is presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. A total of 414 patients with LDH were included, of whom 93 (22.5%) experienced adverse outcomes and 321 (77.5%) did not. No significant differences were observed between the two groups regarding baseline characteristics such as age or smoking history (P\u0026thinsp;\u0026gt;\u0026thinsp;0.05). In contrast, patients with adverse outcomes exhibited significantly higher BWHD, RDH, and BMI values, as well as a higher prevalence of MC and ILI (all P\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Detailed baseline characteristics are presented in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eResults of difference analysis\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eControls (N\u0026thinsp;=\u0026thinsp;321)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCases (N\u0026thinsp;=\u0026thinsp;93)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTotal (N\u0026thinsp;=\u0026thinsp;414)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSex\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.112\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003emale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e204\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e272\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003efemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e117\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e142\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAge(year)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e45.80\u0026thinsp;\u0026plusmn;\u0026thinsp;13.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e47.17\u0026thinsp;\u0026plusmn;\u0026thinsp;12.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e46.11\u0026thinsp;\u0026plusmn;\u0026thinsp;13.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.40\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eBlood Glucose(mmol/L)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.07\u0026thinsp;\u0026plusmn;\u0026thinsp;1.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.22\u0026thinsp;\u0026plusmn;\u0026thinsp;1.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5.10\u0026thinsp;\u0026plusmn;\u0026thinsp;1.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.18\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eBMI(kg/m\u0026sup2;)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e24.31\u0026thinsp;\u0026plusmn;\u0026thinsp;3.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e24.74\u0026thinsp;\u0026plusmn;\u0026thinsp;3.49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e24.41\u0026thinsp;\u0026plusmn;\u0026thinsp;3.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.27\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHypertension\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.15\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e280\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e355\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDiabetes Mellitus\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.08\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e300\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e381\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSmoking\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.44\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e259\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e330\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLHS\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7.44\u0026thinsp;\u0026plusmn;\u0026thinsp;4.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7.95\u0026thinsp;\u0026plusmn;\u0026thinsp;4.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7.56\u0026thinsp;\u0026plusmn;\u0026thinsp;4.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDisease Duration(day)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e26.83\u0026thinsp;\u0026plusmn;\u0026thinsp;45.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e29.98\u0026thinsp;\u0026plusmn;\u0026thinsp;39.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e27.54\u0026thinsp;\u0026plusmn;\u0026thinsp;44.34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.06\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eBlood Loss(ml)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18.04\u0026thinsp;\u0026plusmn;\u0026thinsp;11.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e19.35\u0026thinsp;\u0026plusmn;\u0026thinsp;8.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e18.34\u0026thinsp;\u0026plusmn;\u0026thinsp;10.48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eOperation Time(min)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e101.95\u0026thinsp;\u0026plusmn;\u0026thinsp;36.79\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e103.17\u0026thinsp;\u0026plusmn;\u0026thinsp;40.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e102.22\u0026thinsp;\u0026plusmn;\u0026thinsp;37.50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.90\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePfirrmann Classification\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.75\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eⅠ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eⅡ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eⅢ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e113\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eⅣ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e184\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e232\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eⅤ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eIDH\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11.62\u0026thinsp;\u0026plusmn;\u0026thinsp;2.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11.69\u0026thinsp;\u0026plusmn;\u0026thinsp;2.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e11.64\u0026thinsp;\u0026plusmn;\u0026thinsp;2.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.80\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eVBH\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e29.97\u0026thinsp;\u0026plusmn;\u0026thinsp;3.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e29.88\u0026thinsp;\u0026plusmn;\u0026thinsp;3.37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e29.95\u0026thinsp;\u0026plusmn;\u0026thinsp;3.327\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.83\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDHI\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.39\u0026thinsp;\u0026plusmn;\u0026thinsp;0.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.39\u0026thinsp;\u0026plusmn;\u0026thinsp;0.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.39\u0026thinsp;\u0026plusmn;\u0026thinsp;0.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.63\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eBWHD\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.63\u0026thinsp;\u0026plusmn;\u0026thinsp;0.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.06\u0026thinsp;\u0026plusmn;\u0026thinsp;0.79\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.72\u0026thinsp;\u0026plusmn;\u0026thinsp;0.70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eRDH\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.33\u0026thinsp;\u0026plusmn;\u0026thinsp;0.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.37\u0026thinsp;\u0026plusmn;\u0026thinsp;0.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.34\u0026thinsp;\u0026plusmn;\u0026thinsp;0.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSS(\u0026deg;)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e33.11\u0026thinsp;\u0026plusmn;\u0026thinsp;10.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e34.40\u0026thinsp;\u0026plusmn;\u0026thinsp;10.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e33.40\u0026thinsp;\u0026plusmn;\u0026thinsp;10.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.28\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eALL\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.50\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL2/3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL3/4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e120\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL4/5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e47\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL5/S1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eMC\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo Modic changes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e291\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e364\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrade I\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrade II\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrade III\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLeft-right Distribution\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeft side\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e153\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e203\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRight side\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e121\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e152\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCentral\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFar lateral\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLee Classification\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.83\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEntrance zone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e277\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e358\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMid zone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eExit zone\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLateral Cobb Angle\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e24.24\u0026thinsp;\u0026plusmn;\u0026thinsp;12.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e24.75\u0026thinsp;\u0026plusmn;\u0026thinsp;12.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e24.36\u0026thinsp;\u0026plusmn;\u0026thinsp;12.49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.73\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLumbar Spine ROM on X-ray\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e22.82\u0026thinsp;\u0026plusmn;\u0026thinsp;11.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e21.44\u0026thinsp;\u0026plusmn;\u0026thinsp;12.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e22.51\u0026thinsp;\u0026plusmn;\u0026thinsp;12.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.33\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLumbar Extension Cobb Angle on X-ray\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e32.15\u0026thinsp;\u0026plusmn;\u0026thinsp;12.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e32.98\u0026thinsp;\u0026plusmn;\u0026thinsp;13.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e32.34\u0026thinsp;\u0026plusmn;\u0026thinsp;12.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.58\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLumbar Flexion Cobb Angle on X-ray\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9.39\u0026thinsp;\u0026plusmn;\u0026thinsp;13.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11.53\u0026thinsp;\u0026plusmn;\u0026thinsp;14.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e9.87\u0026thinsp;\u0026plusmn;\u0026thinsp;13.51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.18\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eResponsible Segment ROM\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.87\u0026thinsp;\u0026plusmn;\u0026thinsp;5.22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6.279\u0026thinsp;\u0026plusmn;\u0026thinsp;3.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5.96\u0026thinsp;\u0026plusmn;\u0026thinsp;4.95s\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.19\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eMMFI\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e17.10\u0026thinsp;\u0026plusmn;\u0026thinsp;10.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15.80\u0026thinsp;\u0026plusmn;\u0026thinsp;7.80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e16.81\u0026thinsp;\u0026plusmn;\u0026thinsp;10.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.76\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAAS\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.59\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e259\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e331\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eZygapophyseal Joint Hypertrophy\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.69\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e293\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e376\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e28\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eCalcification\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.59\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e259\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e331\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eILI\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e294\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e362\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"5\"\u003e\u003csup\u003e*\u003c/sup\u003e\u003cb\u003eBMI\u003c/b\u003e:Body Mass Index \u003cb\u003eLHS\u003c/b\u003e:Length of Hospital Stay \u003cb\u003eFHS\u003c/b\u003e:First Herniated Segment \u003cb\u003eIDH\u003c/b\u003e:Intervertebral Disc Height \u003cb\u003eVBH\u003c/b\u003e:Vertebral Body Height \u003cb\u003eDHI\u003c/b\u003e:Disc Height Index \u003cb\u003eBWHD\u003c/b\u003e:Basal Width of Herniated Disc \u003cb\u003eRDH\u003c/b\u003e:Ratio Of Disc Herniation \u003cb\u003eSS\u003c/b\u003e:Sacral Slope \u003cb\u003eALL\u003c/b\u003e:Apex of Lumbar Lordosis \u003cb\u003eMC\u003c/b\u003e:Modic Changes \u003cb\u003eROM\u003c/b\u003e:Range of Motion \u003cb\u003eAAS\u003c/b\u003e:Abdominal Aortic Sclerosis \u003cb\u003eILI\u003c/b\u003e:Interspinous Ligament Injury \u003cb\u003eMMFI\u003c/b\u003e:Multifidus Muscle Fat Infiltration Rate \u003cb\u003eFPT\u003c/b\u003e:first protruding type\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003ePerformance of Adverse Outcome Prediction Models in the Training and Validation Cohorts\u003c/h3\u003e\n\u003cp\u003eSix machine learning algorithms were used for feature selection, and the three algorithms with the highest AUC values\u0026mdash;LASSO (AUC\u0026thinsp;=\u0026thinsp;0.681), XGB (AUC\u0026thinsp;=\u0026thinsp;0.701), and GBM (AUC\u0026thinsp;=\u0026thinsp;0.722) \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA\u003cb\u003e)\u003c/b\u003e\u0026mdash;were selected as candidate algorithms. In the LASSO model, 12 core features were identified at the optimal lambda value \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB\u003cb\u003e)\u003c/b\u003e; in the GBM model, the top 10 features by importance were defined as key features \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC\u003cb\u003e)\u003c/b\u003e; in the XGB model, the top 10 features by importance were designated as key features \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD\u003cb\u003e)\u003c/b\u003e. The intersection of these features yielded 5 core indicators, including MC, BWHD, BMI, RDH, and ILI \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eE\u003cb\u003e)\u003c/b\u003e. Using these core features, models based on five machine learning algorithms were constructed in the training set. After ten-fold cross-validation, the XGB model achieved the highest AUC value (0.809) \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA\u003cb\u003e)\u003c/b\u003e, and it also exhibited a high AUC value (0.718) in the validation set \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB\u003cb\u003e)\u003c/b\u003e. Detailed performance metrics, including accuracy, sensitivity, specificity, kappa coefficient, and F1-score, are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. Decision curve analysis indicated that the XGB model provided superior net clinical benefit across a broad range of threshold probabilities in both cohorts \u003cb\u003e(\u003c/b\u003eFigs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD\u003cb\u003e)\u003c/b\u003e. SHAP analysis demonstrated that all five core indicators contributed positively to adverse outcome risk \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE\u003cb\u003e)\u003c/b\u003e. A nomogram was subsequently constructed to visualize the relative contribution of each predictor \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eF\u003cb\u003e)\u003c/b\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e10-Fold Cross Validation Yielded the Average Scores of the Individual Models.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClassifier\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSensitivity\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSpecificity\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eAccuracy\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eKappa\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eF1\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eGBM\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.76(\u0026plusmn;\u0026thinsp;0.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.17(\u0026plusmn;\u0026thinsp;0.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e \u003cp\u003e0.99(\u0026plusmn;\u0026thinsp;0.01)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e \u003cp\u003e0.80(\u0026plusmn;\u0026thinsp;0.01)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c6\"\u003e \u003cp\u003e0.21(\u0026plusmn;\u0026thinsp;0.09)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c7\"\u003e \u003cp\u003e0.30(\u0026plusmn;\u0026thinsp;0.06)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eKNN\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.70(\u0026plusmn;\u0026thinsp;0.06)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.06(\u0026plusmn;\u0026thinsp;0.06)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e \u003cp\u003e0.95(\u0026plusmn;\u0026thinsp;0.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e \u003cp\u003e0.75(\u0026plusmn;\u0026thinsp;0.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c6\"\u003e \u003cp\u003e0.01(\u0026plusmn;\u0026thinsp;0.09)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c7\"\u003e \u003cp\u003e0.15(\u0026plusmn;\u0026thinsp;0.07)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eNNET\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.60(\u0026plusmn;\u0026thinsp;0.10)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.06(\u0026plusmn;\u0026thinsp;0.11)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e \u003cp\u003e0.98(\u0026plusmn;\u0026thinsp;0.05)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e \u003cp\u003e0.77(\u0026plusmn;\u0026thinsp;0.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c6\"\u003e \u003cp\u003e0.04(\u0026plusmn;\u0026thinsp;0.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c7\"\u003e \u003cp\u003e0.25(\u0026plusmn;\u0026thinsp;0.09)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSVM\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.59(\u0026plusmn;\u0026thinsp;0.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.05(\u0026plusmn;\u0026thinsp;0.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e \u003cp\u003e0.98(\u0026plusmn;\u0026thinsp;0.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e \u003cp\u003e0.78(\u0026plusmn;\u0026thinsp;0.01)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c6\"\u003e \u003cp\u003e0.05(\u0026plusmn;\u0026thinsp;0.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c7\"\u003e \u003cp\u003e0.21(\u0026plusmn;\u0026thinsp;0.01)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eXGB\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.77(\u0026plusmn;\u0026thinsp;0.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.23(\u0026plusmn;\u0026thinsp;0.09)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e \u003cp\u003e0.98(\u0026plusmn;\u0026thinsp;0.01)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e \u003cp\u003e0.82(\u0026plusmn;\u0026thinsp;0.01)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c6\"\u003e \u003cp\u003e0.28(\u0026plusmn;\u0026thinsp;0.11)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c7\"\u003e \u003cp\u003e0.39(\u0026plusmn;\u0026thinsp;0.04)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLASSO\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.76(\u0026plusmn;\u0026thinsp;0.04)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.07(\u0026plusmn;\u0026thinsp;0.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c4\"\u003e \u003cp\u003e0.99(\u0026plusmn;\u0026thinsp;0.01)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c5\"\u003e \u003cp\u003e0.79(\u0026plusmn;\u0026thinsp;0.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c6\"\u003e \u003cp\u003e0.09(\u0026plusmn;\u0026thinsp;0.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c7\"\u003e \u003cp\u003e0.13(\u0026plusmn;\u0026thinsp;0.03)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"7\"\u003eNote: GBM: Gradient Boosting Machine, KNN: K-Nearest Neighbors, NNET: Neural Network, SVM: Support Vector Machine, XGB: Extreme Gradient Boosting, LASSO: Least Absolute Shrinkage and Selection Operator, AUC: Area Under the ROC Curve\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eIdentification of Risk Thresholds\u003c/h2\u003e \u003cp\u003eTo enhance clinical applicability, risk thresholds for continuous variables were determined using the optimal XGB model and the Youden index. BWHD\u0026thinsp;\u0026ge;\u0026thinsp;1.715 mm was identified as a high-risk threshold, associated with a 1.7-fold increase in adverse outcomes in the training cohort and a 1.6-fold increase in the validation cohort (Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA\u0026ndash;C). RDH\u0026thinsp;\u0026ge;\u0026thinsp;37.2% was associated with a 0.8-fold and 1.1-fold increase in adverse outcomes in the training and validation cohorts, respectively (Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD\u0026ndash;F). BMI\u0026thinsp;\u0026ge;\u0026thinsp;23.346 kg/m\u0026sup2; was also identified as a high-risk threshold, increasing adverse outcome incidence by 0.8-fold in the training cohort and 0.6-fold in the validation cohort (Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eG\u0026ndash;I). All differences were statistically significant.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eCollectively, these findings establish a quantitative, threshold-based risk stratification framework based on preoperative clinical and MRI indicators, enabling the translation of predictive features into clinically actionable decision thresholds with consistent performance across both cohorts.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eAlthough percutaneous endoscopic interlaminar discectomy (PEID) is widely regarded as an effective minimally invasive treatment for lumbar disc herniation (LDH), postoperative adverse outcomes\u0026mdash;including recurrence, complications, and delayed recovery\u0026mdash;are still reported in 3.6\u0026ndash;12.4% of patients.[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] Accurate prediction of such outcomes therefore remains a critical clinical challenge. In this study, we developed and validated a machine learning\u0026ndash;based predictive model for adverse outcomes following PEID using a cohort of 414 patients. Several important findings emerged. First, a robust prediction model was established based on five core indicators and demonstrated favorable discrimination and clinical utility. Second, Modic changes (MC), basal width of the herniated disc (BWHD), body mass index (BMI), ratio of disc herniation (RDH), and interspinous ligament injury (ILI) were identified as key contributors to adverse postoperative outcomes. Third, clinically actionable risk thresholds for BWHD, RDH, and BMI were quantitatively defined for the first time.\u003c/p\u003e \u003cp\u003eAmong the identified predictors, MC emerged as a prominent imaging factor associated with adverse outcomes after PEID, showing a strong positive contribution in both SHAP analyses and nomogram visualization. MC reflect pathological alterations of the vertebral endplates and adjacent bone marrow and are considered markers of segmental degeneration and biomechanical imbalance.[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] From a pathological perspective, MC represent chronic reactive changes characterized by inflammatory infiltration, fatty degeneration, fibrosis, and impaired micro-injury repair, which may hinder optimal stress redistribution after decompression. Moreover, MC are often accompanied by disc height loss, reduced segmental stability, and disturbances in the local immune microenvironment, potentially compromising postoperative recovery and symptom resolution. These suggest that patients with preoperative MC may benefit from more cautious postoperative management, including delayed weight bearing, tailored rehabilitation programs, and extended follow-up to facilitate early detection of persistent pain or recurrent symptoms.\u003c/p\u003e \u003cp\u003eInterspinous ligament (ISL) injury was another key predictor of adverse outcomes, underscoring the importance of posterior ligamentous complex (PLC) integrity in postoperative recovery. As a critical stabilizing structure, injury to the interspinous ligament disrupts normal biomechanical balance, increases intervertebral micromotion, and alters load transmission patterns.[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e] These changes may provoke persistent mechanical irritation, local inflammation, and compensatory paravertebral muscle tension, thereby contributing to prolonged pain and delayed functional recovery. While previous studies have linked ILI to postoperative pain and instability, its role in comprehensive adverse outcomes\u0026mdash;including recurrence and delayed recovery\u0026mdash;has been insufficiently quantified.[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e] Notably, ILI can be reliably assessed on preoperative MRI and is relatively unaffected by postoperative behavioral factors, making it a stable and clinically valuable risk marker. In the context of PEID, where soft tissue disruption is minimal, pre-existing ILI may unmask occult instability and compromise postoperative outcomes. Enhanced postoperative posture control, delayed loading, and prolonged rehabilitation may therefore be warranted in these patients.\u003c/p\u003e \u003cp\u003eThis study also confirmed BWHD and RDH as important quantitative imaging indicators influencing adverse outcomes after PEID. Unlike traditional descriptors such as herniation type or volume, basal morphology parameters more accurately reflect the structural relationship between annulus fibrosus rupture and herniated disc material.[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e] A wider herniated base likely represents a larger annular defect and impaired segmental stability, which may limit postoperative stress redistribution and delay recovery. Consistent with previous biomechanical hypotheses that wide-based herniations are more difficult to stabilize,[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e] our results demonstrated that increased BWHD and RDH were associated not only with recurrence but also with suboptimal pain relief and delayed functional improvement. Importantly, RDH normalizes BWHD to vertebral body size, enhancing interindividual comparability and generalizability. By applying quantitative measurements and machine learning\u0026ndash;based modeling, we identified BWHD\u0026thinsp;\u0026ge;\u0026thinsp;1.715 mm and RDH\u0026thinsp;\u0026ge;\u0026thinsp;37.2% as high-risk thresholds, with patients exceeding these values exhibiting markedly higher risks of adverse outcomes. These findings provide objective, preoperatively accessible indicators that may be readily incorporated into clinical risk stratification.\u003c/p\u003e \u003cp\u003eBMI was another significant clinical predictor identified in this study. Although elevated BMI has long been associated with poorer outcomes in spinal surgery, its impact on minimally invasive procedures such as PEID has often been underestimated.[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e] Our model revealed that adverse effects of BMI emerged at a lower threshold (23.346 kg/m\u0026sup2;) than traditionally reported in open or fusion surgeries, where thresholds typically exceed 25\u0026ndash;30 kg/m\u0026sup2;.[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e] This finding suggests that PEID, which relies on limited posterior working space and endoscopic channel stability, may be more sensitive to biomechanical loading. Even mild overweight status may increase axial load and flexion torque, placing continuous stress on postoperative tissues and hindering recovery. Thus, the identified BMI threshold represents not only a statistically optimal cutoff but also a biomechanically meaningful \u0026ldquo;sensitive turning point\u0026rdquo; for minimally invasive spinal surgery, highlighting the importance of early weight management and individualized rehabilitation strategies.\u003c/p\u003e \u003cp\u003eCollectively, this study advances current understanding of adverse outcomes after PEID by integrating multidimensional clinical and imaging variables into an interpretable machine learning framework. Unlike previous studies that primarily focused on qualitative risk factors, this work quantitatively defined actionable risk thresholds and translated predictive features into a clinically usable nomogram. By combining SHAP interpretation with decision curve analysis, the model bridges the gap between predictive accuracy and clinical applicability, offering a practical tool for individualized risk assessment and postoperative management. Several limitations should be acknowledged. First, this was a single-center retrospective study, and multicenter prospective validation is required to confirm model generalizability. Second, although SHAP improved interpretability, unmeasured confounding factors may still influence model performance. Finally, variables such as psychological status, occupational workload, and postoperative rehabilitation compliance were not included and may further refine predictive accuracy in future studies.\u003c/p\u003e \u003cp\u003eIn conclusion, this study demonstrates that MC, BWHD and RDH, BMI, and ILI are key determinants of adverse outcomes following PEID. By defining clinically meaningful risk thresholds and constructing an interpretable prediction model, this work provides a data-driven foundation for personalized risk stratification and optimized postoperative management. Further large-scale validation may facilitate the transition of PEID outcome prediction from empirical judgment to precision-based decision-making.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study identified five key predictors of adverse outcome risk after PEID: Modic Changes (MC), Basal Width Of The Herniated Disc (BWHD), Body Mass Index (BMI), Ratio Of Disc Herniation(RDH), and Interspinous Ligament Injury (ILI). The XGBoost model exhibited stable performance in both the training set (AUC\u0026thinsp;=\u0026thinsp;0.809) and validation set (AUC\u0026thinsp;=\u0026thinsp;0.718), and combined with SHAP values and a Nomogram, it enabled interpretable individualized risk prediction. Meanwhile, this study, for the first time, integrated these core indicators with specific thresholds, achieving an innovative transition from \u0026ldquo;feature identification\u0026rdquo; to \u0026ldquo;quantification of risk thresholds\u0026rdquo;.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability statement/Availability of data and material\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data and materials are available. The corresponding author can provide the datasets used and/or analyzed during the current study upon reasonable request.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported by the National Natural Science Foundation of China (82360430), science and technology Bureau project of Zunyi city (HZ2023-195), Guizhou Provincial Basic Research Program (Natural Science) (Grant no. Qian Jiao Ke He Jichu-ZK[2023]-yiban 584), The Youth Science and Technology Talent Support Program of Guizhou Provincial Association for Science and Technology (GASTYESS202514).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of interest disclosure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no conflicts of interest for this work.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe norms and standards on which this study is based are the Declaration of Helsinki. The ethics of this study have been approved by the Biomedical Research Ethics Committee of the Affiliated Hospital of Zunyi Medical University, Ethics Review Approval Number: KLL-2025-597\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication/ Patient consent statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe publication of this article has obtained the informed consent of the research subjects and participants.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePermission to reproduce material from other sources\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data and materials are available. The corresponding authors can provide the datasets used and/or analyzed during the current study upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors' contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eLH, JA and HQ: conceptualization; XD, CP and XW: data curation; XD and CP: formal analysis; JA and HQ: funding acquisition; LH, JA and HQ: investigation; XD and CP: methodology; LH, JA and HQ: project administration; JA and HQ: resources; XD and CP: software; LH, JA and HQ: supervision; XD and CP: validation; XD and CP: visualization; XD, CP, XW, LH, JA and HQ: writing – original draft; XD, CP, XW, LH, JA and HQ: writing – review and editing. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors thank the hospital staff for their guidance, support, and collaboration.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eZhang AS, Xu A, Ansari K et al (2023) Lumbar Disc Herniation: Diagnosis and Management. Am J Med 136:645\u0026ndash;651. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.amjmed.2023.03.024\u003c/span\u003e\u003cspan address=\"10.1016/j.amjmed.2023.03.024\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eModic MT, Ross JS (2007) Lumbar degenerative disk disease. Radiology 245:43\u0026ndash;61. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1148/radiol.2451051706\u003c/span\u003e\u003cspan address=\"10.1148/radiol.2451051706\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu S, Zhang X, Xiong Y, He H (2025) Minimally invasive surgery for lumbar disc herniation: a meta-analysis of efficacy and safety. Int J Surg Lond Engl 111:5623\u0026ndash;5636. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1097/JS9.0000000000002434\u003c/span\u003e\u003cspan address=\"10.1097/JS9.0000000000002434\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu J, Li Y, Wang B et al (2020) Minimum 2-Year Efficacy of Percutaneous Endoscopic Lumbar Discectomy versus Microendoscopic Discectomy: A Meta-Analysis. World Neurosurg 138:19\u0026ndash;26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.wneu.2020.02.096\u003c/span\u003e\u003cspan address=\"10.1016/j.wneu.2020.02.096\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKang TW, Park SY, Oh H et al (2021) Risk of reoperation and infection after percutaneous endoscopic lumbar discectomy and open lumbar discectomy: a nationwide population-based study. Bone Jt J 103\u0026ndash;B:1392\u0026ndash;1399. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1302/0301-620X.103B8.BJJ-2020-2541.R2\u003c/span\u003e\u003cspan address=\"10.1302/0301-620X.103B8.BJJ-2020-2541.R2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang X, Gu L, Xu G et al (2023) Nomogram for predicting the unfavourable outcomes of percutaneous endoscopic transforaminal discectomy for lumbar disc herniation: a retrospective study. Front Surg 10:1188517. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fsurg.2023.1188517\u003c/span\u003e\u003cspan address=\"10.3389/fsurg.2023.1188517\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang H, Yang M, Fu Z et al (2024) Predicting and analysing of the unfavourable outcomes of early applicated percutaneous endoscopic interlaminar discectomy for lumbar disc herniation: development and validation based nomogram. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 33:906\u0026ndash;914. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00586-024-08141-9\u003c/span\u003e\u003cspan address=\"10.1007/s00586-024-08141-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTopol EJ (2019) High-performance medicine: the convergence of human and artificial intelligence. Nat Med 25:44\u0026ndash;56. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41591-018-0300-7\u003c/span\u003e\u003cspan address=\"10.1038/s41591-018-0300-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang M, Wang S, Wang Y et al (2024) Nomogram Development and Validation for Predicting Postoperative Recurrent Lumbar Disc Herniation Based on Paraspinal Muscle Parameters. J Pain Res 17:2121\u0026ndash;2131. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2147/JPR.S459846\u003c/span\u003e\u003cspan address=\"10.2147/JPR.S459846\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang H, Yang M, Fu Z et al (2024) Predicting and analysing of the unfavourable outcomes of early applicated percutaneous endoscopic interlaminar discectomy for lumbar disc herniation: development and validation based nomogram. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 33:906\u0026ndash;914. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00586-024-08141-9\u003c/span\u003e\u003cspan address=\"10.1007/s00586-024-08141-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang M, Wang S, Wang Y et al (2024) Development and Validation of a Nomogram Predicting Postoperative Recurrent Lumbar Disc Herniation Based on Activity Factors. Risk Manag Healthc Policy 17:689\u0026ndash;699. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2147/RMHP.S453819\u003c/span\u003e\u003cspan address=\"10.2147/RMHP.S453819\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShen Z, Zhong Z-M, Wu Q et al (2019) Predictors for Poor Outcomes After Percutaneous Endoscopic Lumbar Discectomy: A Retrospective Study of 241 Patients. World Neurosurg 126:e422\u0026ndash;e431. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.wneu.2019.02.068\u003c/span\u003e\u003cspan address=\"10.1016/j.wneu.2019.02.068\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAljawadi A, Sethi G, Islam A et al (2020) Sciatica Presentations and Predictors of Poor Outcomes Following Surgical Decompression of Herniated Lumbar Discs: A Review Article. Cureus 12:e11605. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7759/cureus.11605\u003c/span\u003e\u003cspan address=\"10.7759/cureus.11605\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShi H, Zhu L, Jiang Z-L, Wu X-T (2021) Radiological risk factors for recurrent lumbar disc herniation after percutaneous transforaminal endoscopic discectomy: a retrospective matched case-control study. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 30:886\u0026ndash;892. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00586-020-06674-3\u003c/span\u003e\u003cspan address=\"10.1007/s00586-020-06674-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYin S, Du H, Yang W et al (2018) Prevalence of Recurrent Herniation Following Percutaneous Endoscopic Lumbar Discectomy: A Meta-Analysis. Pain Physician 21:337\u0026ndash;350\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim HS, You JD, Ju CI (2019) Predictive Scoring and Risk Factors of Early Recurrence after Percutaneous Endoscopic Lumbar Discectomy. BioMed Res Int 2019:6492675. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1155/2019/6492675\u003c/span\u003e\u003cspan address=\"10.1155/2019/6492675\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang H, Yang M, Fu Z et al (2024) Predicting and analysing of the unfavourable outcomes of early applicated percutaneous endoscopic interlaminar discectomy for lumbar disc herniation: development and validation based nomogram. Eur Spine J Off Publ Eur Spine Soc Eur Spinal Deform Soc Eur Sect Cerv Spine Res Soc 33:906\u0026ndash;914. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00586-024-08141-9\u003c/span\u003e\u003cspan address=\"10.1007/s00586-024-08141-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMo AZ, Lockey S, Mo F (2025) The Posterior Ligamentous Complex: Anatomic and Biomechanical Considerations in Injury Classification and Management. J Am Acad Orthop Surg 33:e1332\u0026ndash;e1342. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5435/JAAOS-D-22-00908\u003c/span\u003e\u003cspan address=\"10.5435/JAAOS-D-22-00908\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang J, Chen M, Huang D-A et al (2024) High intensity in interspinous ligaments: a diagnostic sign of lumbar instability and back pain for degenerative lumbar spondylolisthesis. BMC Musculoskelet Disord 25:949. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s12891-024-08081-x\u003c/span\u003e\u003cspan address=\"10.1186/s12891-024-08081-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eG\u0026uuml;lensoy B, G\u0026uuml;zel E, Kuzu Kumcu M et al (2023) Recurrence of lumbar disk herniation after microdiscectomy: a two-center retrospective analysis of 1214 cases and identification of risk factors. Turk J Med Sci 53:1254\u0026ndash;1261. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.55730/1300-0144.5691\u003c/span\u003e\u003cspan address=\"10.55730/1300-0144.5691\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi Z-P, Liu L-L, Liu H et al (2024) Radiologic Analysis of Causes of Early Recurrence After Percutaneous Endoscopic Transforaminal Discectomy. Glob Spine J 14:113\u0026ndash;121. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/21925682221096061\u003c/span\u003e\u003cspan address=\"10.1177/21925682221096061\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen XT, Shahrestani S, Ballatori AM et al (2021) The Influence of Body Mass Index in Obese and Morbidly Obese Patients on Complications and 30- and 90-day Readmissions Following Lumbar Spine Fusion. Spine 46:965\u0026ndash;972. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1097/BRS.0000000000003933\u003c/span\u003e\u003cspan address=\"10.1097/BRS.0000000000003933\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatel N, Bagan B, Vadera S et al (2007) Obesity and spine surgery: relation to perioperative complications. J Neurosurg Spine 6:291\u0026ndash;297. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3171/spi.2007.6.4.1\u003c/span\u003e\u003cspan address=\"10.3171/spi.2007.6.4.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKatsevman GA, Daffner SD, Brandmeir NJ et al (2020) Complexities of spine surgery in obese patient populations: a narrative review. Spine J Off J North Am Spine Soc 20:501\u0026ndash;511. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.spinee.2019.12.011\u003c/span\u003e\u003cspan address=\"10.1016/j.spinee.2019.12.011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"european-spine-journal","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"esjo","sideBox":"Learn more about [European Spine Journal](http://link.springer.com/journal/586)","snPcode":"586","submissionUrl":"https://submission.springernature.com/new-submission/586/3","title":"European Spine Journal","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"PEID, adverse outcomes, machine learning, predictive model, risk threshold","lastPublishedDoi":"10.21203/rs.3.rs-8684995/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8684995/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003ePercutaneous endoscopic interlaminar discectomy (PEID) is a common surgical technique for lumbar disc herniation (LDH), but the risk factors for adverse outcomes remain controversial. This study aims to develop and validate a predictive model based on machine learning algorithms to identify key clinical indicators influencing adverse outcomes after PEID.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eThis retrospective study included 414 LDH patients who underwent single-level PEID between October 2018 and June 2024. Data were divided into training (n\u0026thinsp;=\u0026thinsp;290) and validation (n\u0026thinsp;=\u0026thinsp;124) sets. Six machine learning algorithms were used for feature selection, identifying core indicators. Models were constructed based on these indicators and evaluated for predictive performance.\u003c/p\u003e\u003ch2\u003eResult\u003c/h2\u003e \u003cp\u003eFive core indicators were identified: Modic Changes (MC), Basal Width Of The Herniated Disc (BWHD), Body Mass Index (BMI), Ratio Of Disc Herniation (RDH), and Interspinous Ligament Injury (ILI). The XGB model performed best, with an AUC of 0.809 in the training set and 0.718 in the validation set. Risk thresholds for BWHD, BMI, and RDH were 1.715mm, 23.346kg/m\u0026sup2;, and 37.2%, respectively.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eMC, ILI, BWHD, BMI, and RDH are risk factors for postoperative adverse outcomes in PEID patients. The model provides useful clinical guidance, with validated risk thresholds for key indicators.\u003c/p\u003e","manuscriptTitle":"Risk factors and risk thresholds for adverse outcomes after PEID in patients with lumbar disc herniation: a retrospective cohort study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-18 17:03:34","doi":"10.21203/rs.3.rs-8684995/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-01-30T01:14:42+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-01-30T01:09:20+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-01-29T02:50:20+00:00","index":"","fulltext":""},{"type":"submitted","content":"European Spine Journal","date":"2026-01-24T08:02:12+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"european-spine-journal","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"esjo","sideBox":"Learn more about [European Spine Journal](http://link.springer.com/journal/586)","snPcode":"586","submissionUrl":"https://submission.springernature.com/new-submission/586/3","title":"European Spine Journal","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"4221c9ab-7aa9-4c07-a45e-9081315a1010","owner":[],"postedDate":"February 18th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-28T14:10:27+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-18 17:03:34","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8684995","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8684995","identity":"rs-8684995","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}