Trochlear dysplasia is the primary anatomic risk factor for patellofemoral instability (PFI), but current classification systems rely on qualitative observations and are limited in their reproducibility. The Dejour Version 3.0 (2025) classification was established on quantitative magnetic resonance imaging (MRI)-based measurements in adults, but its validity in the pediatric population has yet to be evaluated. To (1) assess the accuracy of the Dejour MRI-based classification of trochlear dysplasia in the diagnosis of PFI in children and adolescents and (2) derive pediatrics-specific thresholds of MRI-based measurements of dysplasia and additional risk factors to optimally predict PFI. Case-control study; Level of evidence, 3. A total of 144 knees (127 patients) with objective PFI were age- and sex-matched to 144 controls. Four raters had excellent agreement on 7 measures of patellofemoral morphology: cartilaginous sulcus angle, lateral trochlear inclination, patellar tilt, lateral patellofemoral angle (LPFA), tibial tubercle-trochlear groove distance, sagittal central bump size, and Caton-Deschamps Index. Dejour Version 3.0 was assessed for sensitivity, specificity, and diagnostic accuracy, as defined by the area under the curve (AUC) for the respective receiver operating characteristic curves, within this study's pediatric sample. Regression tree analysis with recursive partitioning was utilized to identify pediatrics-specific threshold values on MRI. Resulting combinations were assessed for their sensitivity, specificity, and diagnostic accuracy. The AUCs for the 2 options with the highest sensitivity were compared using a random forest (RF) model to evaluate optimal diagnostic accuracy. Application of the 4 previously established adult cutoff combinations resulted in low/moderate sensitivity and fair/good diagnostic accuracy (range of AUCs, 0.79-0.87) in the study's pediatric cohort. Regression tree analysis yielded 5 cutoff combinations, of which 2 achieved a sensitivity >90%. The first cutoff was a singular cartilaginous sulcus angle measurement ≥151° (sensitivity: 93% [95% CI, 87.6%-96.6%]; specificity: 87% [95% CI, 80.2%-91.9%]; AUC, 0.94); the second cutoff combination incorporated an LPFA cutoff <0.45° if the cartilaginous sulcus angle was <151° (sensitivity: 98% [95% CI, 94.0%-99.6%]; specificity: 85% [95% CI, 78.6%-90.7%]; AUC, 0.97). The AUC for the second cutoff combination was noninferior to the AUC (RF) by a prespecified ΔAUC of 0.03 (P = .33). Application of the Dejour classification of trochlear dysplasia utilizing adult-specific thresholds yielded only moderate accuracy in the diagnosis of PFI in pediatric patients. The authors present an MRI-based classification system utilizing objective measurements of trochlear and patellofemoral morphology, emphasizing a 2-measurement combination of sulcus angle and LPFA that yielded excellent diagnostic accuracy. This new classification system is presented as a practical tool to objectively diagnose pediatric PFI in the clinical setting.

Incorporation of grade into stage in oral cavity squamous cell carcinoma: A novel staging schema.

A body of research suggests that many participants in alcohol use treatment research begin making changes to their drinking behavior prior to beginning treatment. Although several studies have examined factors associated with such change, results have been mixed or yielded small effects. Both theoretical and methodological limitations have hindered efforts to understand such change, highlighting the need for novel analytic approaches. This study aimed to examine how traditional modeling methods (linear and logistic regression) and machine learning (recursive partitioning, random forests, neural networks, and support vector machines) may be used to predict pretreatment drinking changes. Using baseline psychological constructs and demographic variables of an existing dataset of 175 predominately white (93.7%) participants, we randomly split the data into a training dataset (80% of available data) and a testing dataset (20% of available data) and used the training dataset to run the initial models. Then, using the remaining test data, we used the trained models to make predictions on new data. We ran models predicting the percent change in drinking and heavy drinking days, and classification models on whether participants were pretreatment changers (50% reduction in drinking days; 70% reduction in percent heavy drinking days). Overall, the neural-network models tended to have the highest predictive accuracy, although the model areas under the curves ranged from poor to acceptable. Variable importance algorithms indicated that demographic factors (e.g., education and income) and psychological constructs (e.g., processes of change) were among the predictors that contributed most to improved model performance. The identification of demographic variables as important predictors highlights the importance of understanding demographic and societal-level factors as potential drivers of pretreatment drinking changes.

We describe clinical and biologic characteristics of neuroblastoma in older children, adolescents, and young adults (OCAYA); describe survival outcomes in the post-immunotherapy era; and identify if there is an age cut-off that best discriminates outcomes. Patients diagnosed with neuroblastoma at ≥547 days between 2003 and 2022 from the International Neuroblastoma Risk Group Data Commons were compared by age subgroups. Recursive partitioning, dividing younger versus older at all monthly cut-points between 18 months and 15 years, was undertaken using Cox regression models of event-free survival (EFS), overall survival (OS), and OS post-relapse (OSPR). Kaplan-Meier curves of clinical/biologic subgroups were compared with log-rank tests. 7,835 patients met inclusion criteria: 18 months to <5 years (n=5841), 5 to <10 years (n=1488), 10 to <15 years (n=357), and ≥15 years (n=149) at diagnosis. Younger patients were more likely to have MYCN amplification (18 months to 5 years: 31%; 5-10 years: 15%) than older (10-15 years: 8%; ≥15 years: 7%) (p<0.0001), metastatic disease (p<0.0001), and high mitosis-karyorrhexis index (MKI) (p<0.0001) and less likely to have diploid tumors (p<0.001). Repeatedly dichotomizing the cohort, younger patients had superior EFS and OS (p<0.05) for all cut-offs ≤40 months (hazard ratios: 1.1-1.3). Among high-risk OCAYA (International Neuroblastoma Staging System [INSS] Stage 4; n=5005 [64% of cohort]), those diagnosed 2010-2022 had superior EFS/OS versus 2003-2009 in each age group (p<0.0001). OSPR remained poor for all OCAYA (5-year OSPR 14%±0.7%). For patients ≥547 days old, any age cut-off ≤40 months discriminated younger (superior EFS/OS) versus older patients; no cut-off was optimal. OCAYA diagnosed 2010-2022 (post-immunotherapy era) had superior outcomes versus 2003-2009. Stratification by comprehensive molecular biomarkers will likely best inform novel therapeutic strategies for OCAYA.

WHO grade 4 astrocytomas are associated with poor prognosis, and their prognostic factors remain controversial. This study aimed to identify the prognostic factors and develop a management algorithm for these patients. This study retrospectively included 151 CNS5 adult grade 4 astrocytomas from two medical centers. The tumors were classified as histologic grade 2/3 astrocytomas with CDKN2A/B homozygous deletion (molecular grade 4 astrocytoma, MA4), histologic grade 4 astrocytomas with CDKN2A/B homozygous deletion (molecular and histologic grade 4 astrocytoma, MHA4), and histologic grade 4 astrocytomas without CDKN2A/B homozygous deletion (histologic grade 4 astrocytoma, HA4). Prognostic factors were identified and incorporated into recursive partitioning analysis (RPA) for survival risk stratification. Histologic grade 4 astrocytomas with CDKN2A/B homozygous deletion, postoperative tumor volume (TV), and chemoradiotherapy were associated with patient survival. RPA identified three groups with distinct prognoses (p = 0.001). Group 1 had a median overall survival (OS) of 77.8 months, consisting of MA4 and HA4 with postoperative TV on FLAIR ≤ 28.5 mL. Group 2 had a median OS of 32.2 months, including MA4 and HA4 with postoperative TV on FLAIR > 28.5 mL receiving chemoradiotherapy, or MHA4 with postoperative TV on FLAIR ≤ 28.5 mL. Group 3 had a median OS of 14.9 months, including MA4 and HA4 with postoperative TV on FLAIR > 28.5 mL without chemoradiotherapy, or MHA4 with postoperative TV on FLAIR > 28.5 mL receiving chemoradiotherapy. Histologic grade 4 astrocytomas with CDKN2A/B homozygous deletion confer the worst survival. Maximal or complete resection, as assessed on FLAIR images, is critical to improving outcomes.

Non-relapse mortality in allogeneic haemopoietic stem cell transplantation (alloHSCT) is driven by acute graft-versus-host disease (aGVHD) and infection. The MAGIC Algorithm Probability (MAP) composite biomarker score is based on serum ST2 and REG3a levels at day(D) 7 post-transplant and predicts aGVHD-related mortality. While changes in MAP score predict aGVHD treatment response, their post-transplant predictive value and association with infection-related NRM (iNRM), independent of GVHD, are unclear. We evaluated the association of D0-21 MAP change (MAPΔ) versus D7 MAP with 6-month NRM (6mNRM) and developed a risk stratification model using random forest analysis and recursive partitioning. We prospectively enrolled 101 adult alloHSCT recipients from 2022-2024. Serum was collected weekly from D0-21 to derive MAPΔ. Associations with 6-month NRM were assessed using generalized linear and random forest modelling. 6mNRM was 14.8%, with 80% (12/15) of deaths attributed to infection without prior aGVHD. Both MAPΔ ≥0.055 and D7 MAP ≥0.16 were significantly associated with 6mNRM (p<0.0001), with MAPΔ showing superior predictive accuracy (Receiver Operator Characteristic AUC 0.934 vs. 0.779). In multivariable analysis, MAPΔ (OR 45; 95% CI 10.6-318; p=0.01), but not D7 MAP, was an independent predictor of 6mNRM. A classification tree incorporating MAPΔ, D7 MAP, and age-adjusted comorbidity index stratified patients into low, intermediate, and high-risk groups. MAPΔ was a powerful, independent predictor of 6mNRM, which was driven by infection rather than aGVHD in this cohort. We propose a tool based on MAPΔ to assess iNRM, which may guide interventions and clinical trial design. Further validation in larger cohorts is warranted.

ObjectiveIschemic stroke (IS) with hyperuricemia (HUA) correlates with poor outcomes, yet the shared pathophysiological traits remain unclear. This study examined metabolic parameters in HUA-IS comorbidity and developed an optimal interpretable Clinlabomics model for risk assessment.MethodsA total of 2,164 IS patients and 2,459 healthy controls (HCs) were retrospectively enrolled. Participants were divided into four groups: HUA-IS (comorbidity, n=1,082), non-HUA IS (n=1,082), HUA HCs (n=1,314), non-HUA HCs (n=1,145); the latter three were defined as the non-comorbidity group. After 1:1 propensity score matching (PSM), 1,031 cases were matched in each group. Ten metabolic parameters were analyzed: serum uric acid at admission (SUA_admission), SUA on the third day of hospitalization (SUA_3d), triglyceride-glucose index (TyG), triglyceride (TG), high-density lipoprotein cholesterol (HDL−C), atherogenic index of plasma (AIP), atherogenic coefficient (AC), lipoprotein combine index (LCI), Castelli’s risk index I (CRI-I), and Castelli’s risk index II (CRI-II). Univariate/multivariate logistic regression, quartile-based logistic regression, and restricted cubic spline (RCS) analysis were used to explore parameters - comorbidity associations. Post-PSM data were split 7:3 into training/testing sets, least absolute shrinkage and selection operator (LASSO) regression selected features, and 11 machine learning algorithms developed Clinlabomics models. Additionally, the optimal model was validated in the testing set and an independent validation set.ResultsAfter PSM, multivariate logistic regression identified AIP as the strongest risk factor (OR = 2.74, 95%CI: 1.80-4.19). The Q4 of TyG, TG, AIP, and LCI elevated comorbidity risk (P < 0.05). Besides, RCS showed nonlinear association of LCI with comorbidity (P < 0.05). The Recursive Partitioning and Regression Trees (rpart)-based Clinlabomics model exhibited favorable performance with F1-score, accuracy (ACC), and area under the curve (AUC) of 0.960, 0.960, and 0.986. At optimal hyperparameter (cp=0.0017), the model achieved AUCs of 0.987 (95%CI: 0.982-0.993), 0.955 (95%CI: 0.939-0.972), and 0.957 (95%CI: 0.915-0.999) in the training, testing, and validation datasets, respectively, correctly identifying 87.7% non-comorbidity and 98.0% comorbidity patients in validation. SHapley Additive exPlanations (SHAP) analysis identified UA_admission, UA_3d, TyG, TG, AIP and LCI as key metabolic indicators.ConclusionTyG, TG, AIP, and LCI were critical metabolic parameters for HUA-IS comorbidity, which warrant heightened attention in future comorbidity research.

Abstract Regulations of chemical exposures often focus on individual substances, neglecting the amplified toxicity that can arise from multiple concurrent exposures. We propose a novel methodology to identify critical thresholds in multivariate exposure spaces and estimate the effects of policy interventions that limit exposures within these thresholds. Our approach employs a recursive partitioning algorithm integrated with targeted maximum likelihood estimation (TMLE) to discover regions in the exposure space where the expected outcome is minimized or maximized. To address potential overfitting bias from using the same data for threshold discovery and effect estimation, we utilize cross-validated TMLE (CV-TMLE), which ensures asymptotic unbiasedness and efficiency. Simulation studies demonstrate convergence to the optimal exposure region and accurate estimation of intervention effects. We apply our method to synthetic mixture data, successfully identifying true interactions, and to NHANES data, discovering harmful metal exposures affecting telomere length. Our approach provides a flexible and interpretable framework for policy-makers to assess the impact of exposure regulations, and we offer an open-source implementation in the CVtreeMLE R package.

There is a paucity of research focused on risk and protective factors for depression in rural Latine adolescents. The present study first identified variables commonly described in conceptual models of depression etiology and maintenance in Latine adolescents and rural populations, including demographic (i.e., age, sex), cultural (i.e., acculturation), familial (i.e., family conflict, familism), and contextual factors (i.e., socioeconomic strain, parental education level, discrimination-related stress). A machine learning approach was then used to understand the relative contributions of these variables to depression in rural Latine adolescents. Participants (n = 670; Mage = 15.74; 53% female) were Latine adolescents in grades 9-12 recruited from a high school in a low-income rural area, who completed a battery of self-report measures. A data-driven recursive partitioning method was used to examine the joint contribution of these variables to depression severity. Using a conditional inference framework, adolescents with low depression scores were characterized by high familism and low discrimination-related stress, whereas adolescents with high depression scores endorsed low familism. Female adolescents had higher depression severity than their male counterparts. These findings are consistent with both conceptual models of depression in Latine youth and previous empirical studies, particularly those showing that familism and discrimination-related stress play a significant role as protective and risk factors, respectively. The identification of crucial variables using data-driven approaches could help improve screening for and treatment of depression in rural Latine youth who experience significant mental health inequities. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

Pediatric septic arthritis requires immediate recognition, as delayed diagnosis can cause severe joint damage and long-term dysfunction. In the absence of guideline-based cut-off values for laboratory markers, surgical decisions are often made on a case-by-case basis. Due to the scarcity of evidence specific to the pediatric population, treatment strategies are often based on adult data, highlighting the need for targeted research in this population. To address this gap, we developed a diagnostic algorithm that incorporated reliable predictive factors. Of 443 joint aspirations performed in our clinic (2014-2024), 132 (29.8%) were for suspected septic arthritis. After applying exclusion criteria, 80 cases were included. Clinical (fever, pain with movement, comorbidities), laboratory parameters at the time of joint aspiration (serum CRP, synovial white blood cell count [syWBC], serum white blood cell count [seWBC], synovial neutrophil perventage [syN%] and radiological data (radiographs, CT and MRI if available) were collected. Septic arthritis was defined by detection of pathogens in joint aspirate via culture or PCR. A pathogen was identified in 25% (20/80) of cases, with Kingella kingae (30%) being the most frequently detected organism, followed by Staphylococcus aureus (25%). Regression analysis revealed CRP (p < 0.01), syWBC (p = 0.04), but not syN% (p = 0.51) as predictors. ROC analysis yielded optimal cutoff values for CRP (69mg/L; AUC = 0.82; 95% CI 0.71-0.93) and syWBC (65,000 cells/µL; AUC = 0.79; 95% CI 0.66-0.92). A diagnostic algorithm using CRP > 69mg/L alone, or CRP < 69mg/L combined with syWBC > 110,000 cells/µL, predicted septic arthritis with a sensitivity of 85% (95% CI 0.62-0.97; p < 0.01) and a specificity of 90% (95% CI 0.79-0.96; p < 0.01). CRP was slightly more accurate than syWBC in predicting septic arthritis. When combined in an recursive partitioning model, these parameters demonstrated strong diagnostic performance. In cases where CRP measurements may be unreliable, an elevated syWBC count represents a CRP-independent alternative, although with reduced specificity. Level III: a retrospective case-control study.

Prognostic scores for patients with multiple brain metastases treated with repeated stereotactic radiosurgery - a secondary analysis of the CYBER-SPACE randomized phase 2 trial.

Do intact rotator cuff tendons and soft tissue tensioning affect stress shielding in reverse total shoulder arthroplasty? Findings from the Medacta shoulder system.

Refining of spectrophotometric direct peptide reactivity assay (Spectro-DPRA) for in vitro skin sensitization: A novel predictive model based on decision tree.

Abstract To develop genotypes, breeders test and select genotypes across sites and years representative of their target population of environments (TPE). This is not cost‐effective at early breeding stages when evaluating thousands of genotypes. We demonstrate the use of self‐organizing maps (SOMs), genomic estimated breeding values (GEBV), and recursive partitioning to identify environmental groups (EGs), environmental covariates differentiating among EGs, and hybrids stable across EGs. Genotypes were predicted based on 310 site‐specific estimation sets trained on as many sites in the Argentine maize TPE between 2018 and 2023. Analyses showed that the first cropping season in the Center and South was associated with drought and heat stress during grain fill; the second cropping season was likely to experience drought during the vegetative phase. The second season in the North showed higher temperatures and higher vapor pressure deficit (VPD). SOM grouped sites with similar genotypic ranking for grain yield in repeatable EGs irrespective of geographical zone, season, or year for lines and hybrids. Drought, heat, and high VPD at different phenological stages explained differences in yield among EGs. Lines and hybrids stable across EGs were identified using SOM. We infer that these hybrids would show good adaptation and stability across conditions used for model training. Using site‐specific GEBVs in an analogous way to phenotypic data allowed simulating the performance of lines and hybrids at early breeding stages beyond solely relying on phenotypic evaluations. Since SOM predominantly focuses on yield, it is critical to use SOM in combination with other methods and information available.

Evidence on the prognostic and staging effects of lymphovascular invasion (LVI) after neoadjuvant chemoradiotherapy for esophageal squamous cell carcinoma (ESCC) is limited. We aimed to determine the prognostic value of LVI and develop a modified post-neoadjuvant pathologic staging (ypStage) system integrating LVI and ypTNM stage to improve risk stratification. This multi-institutional study included patients with ESCC receiving neoadjuvant chemoradiotherapy and R0 resection. Recursive partitioning analysis (RPA) was conducted to derive prognostic groupings. A modified ypStage system was developed, validated, and compared with ypTNM stage. A total of 931 patients were divided into training (n = 565) and external validation (n = 366) cohorts. LVI was present in 115 patients (12.4%). LVI was an independent predictor of survival and disease recurrence, with hazard ratios of 1.70 for overall survival and 1.74 for recurrence-free survival. By integrating LVI status and ypTNM stage, nonmetastatic ESCC were classified into three stages with distinct prognoses. The proposed RPA stage provided superior hazard consistency, hazard discrimination, sample size balance, and outcome prediction over ypTNM stage. LVI was a strong prognostic factor, independent of the current ypTNM stage in ESCC. We developed an RPA-based ypStage system integrating LVI status and ypTNM stage that exhibited good prognostic performance.

In line with UNESCO’s Historic Urban Landscape approach, this study highlights the need for integrative tools that connect heritage conservation with broader urban development dynamics, balancing preservation and growth. While several machine-learning models have been applied to analyse the drivers of urban change, there remains a need for comparative analyses that assess their strengths, limitations, and potential for combined applications tailored to specific contexts. This study aims to compare the predictive accuracy of three land-use change models (Random Forest, Logistic Regression, and Recursive Partitioning Regression Trees) in estimating the probability of land-use transitions, as well as their interpretative capacity to identify the main factors driving these changes. Using data from the Bellavista neighborhood in Tomé, Chile, the models were assessed through prediction and performance metrics, probability maps, and an analysis of key driving factors. The results underscore the potential of integrating predictive (Random Forest) and interpretative (Logistic Regression and Recursive Partitioning Regression Trees) approaches to support heritage planning. Specifically, the research demonstrates how these models can be effectively combined by leveraging their respective strengths: employing Random Forest for spatial simulations, Logistic Regression for identifying associative factors, and Recursive Partitioning Regression Trees for generating intuitive decision rules. Overall, the study shows that land-use change models constitute valuable tools for managing urban transformation in heritage urban areas of intermediate cities.

Hepatocellular carcinoma (HCC) has a poor prognosis, particularly with spinal metastases. Current prognostic scores (e.g., Revised Tokuhashi, New England Spinal Metastasis Score) lack integration of tumor microenvironment (TME)-based molecular subtypes, limiting their utility in precision medicine. This study evaluated the prognostic value of these subtypes and whether they enhance established scoring systems.In a single-center retrospective cohort of 117 HCC patients undergoing surgery for spinal metastases (2009-2024), patients were stratified into three TME subtypes: immune-inflamed (n = 39), immune-excluded (n = 53), and immune-desert (n = 25). Overall survival (OS) was analyzed using Kaplan-Meier and Cox regression. The discriminative ability of four prognostic scores was assessed with time-dependent ROC curves. Recursive partitioning analysis (RPA) integrated molecular subtypes with clinical scores to develop novel decision trees.Median OS for the cohort was 13.1 months. TME subtype was a powerful independent prognostic factor, with immune-inflamed, immune-excluded, and immune-desert subtypes showing median OS of 17.2, 12.1, and 8.8 months, respectively (P < 0.001). Multivariable analysis confirmed this association (e.g., immune-desert aHR = 9.52, P < 0.001). The Revised Tokuhashi score showed the highest baseline discriminative ability for 1-year survival (AUROC = 0.726). Integrating TME subtype and postoperative systemic therapy significantly improved predictive accuracy across all models (AUROCs > 0.92). RPA generated clinically actionable decision trees, defining three distinct prognostic groups. TME-based molecular subtypes are critical independent survival determinants in HCC with spinal metastases. Their integration with clinical scores using RPA produces highly accurate predictive models and practical decision aids, advocating for a biology-augmented approach to personalize patient management.

Development of a new model for prediction of relevant treatment related adverse events in older individuals with gastrointestinal cancers.

Efficacy of metastatic lesion radiotherapy in de novo metastatic nasopharyngeal carcinoma patients receiving local regional radiotherapy and chemo-immunotherapy: a multicenter retrospective study.

Maternal circumstances and intergenerational transmission of health: A model-based recursive partitioning approach from Machine Learning.