Validation Cohort Research Articles

Proteomics Analysis of Renal Cell Line Caki-2 with AFMID Overexpression and Potential Biomarker Discovery in Urine.

Aromatic caninurine formamase (AFMID) is an enzyme involved in the tryptophan pathway, metabolizing N-formylkynurenine to kynurenine. AFMID had been found significantly downregulated in clear cell renal cell carcinoma (ccRCC) in both tissue and urine samples. Although ccRCC is characterized by a typical Warburg-like phenotype, mitochondrial dysfunction, and elevated fat deposition, it is unknown whether AFMID plays a role in tumorigenesis and the development of ccRCC. In the present study, AFMID overexpression had inhibitory effects for ccRCC cells, decreasing the rate of cell proliferation. Quantitative proteomics showed that AFMID overexpression altered cellular signaling pathways involved in cell growth and cellular metabolism pathways, including lipid metabolism and inositol phosphate metabolism. Further urine proteomic analysis indicated that cellular function dysfunction with AFMID overexpression could be reflected in the urine. The activity of predicted upregulators DDX58, TREX1, TGFB1, SMARCA4, and TNF in ccRCC cells and urine showed opposing change trends. Potential urinary biomarkers were tentatively discovered and further validated using an independent cohort. The protein panel of APOC3, UMOD, and CILP achieved an AUC value of 0.862 for the training cohort and 0.883 for the validation cohort. The present study is of significance in terms of highlighting various aspects of pathway changes associated with AFMID enzymes, discovering potential specific biomarkers for potential patient diagnosis, and therapeutic targeting.

Noninvasive, microbiome-based diagnosis of inflammatory bowel disease.

Despite recent progress in our understanding of the association between the gut microbiome and inflammatory bowel disease (IBD), the role of microbiome biomarkers in IBD diagnosis remains underexplored. Here we developed a microbiome-based diagnostic test for IBD. By utilization of metagenomic data from 5,979 fecal samples with and without IBD from different geographies and ethnicities, we identified microbiota alterations in IBD and selected ten and nine bacterial species for construction of diagnostic models for ulcerative colitis and Crohn's disease, respectively. These diagnostic models achieved areas under the curve >0.90 for distinguishing IBD from controls in the discovery cohort, and maintained satisfactory performance in transethnic validation cohorts from eight populations. We further developed a multiplex droplet digital polymerase chain reaction test targeting selected IBD-associated bacterial species, and models based on this test showed numerically higher performance than fecal calprotectin in discriminating ulcerative colitis and Crohn's disease from controls. Here we discovered universal IBD-associated bacteria and show the potential applicability of a multibacteria biomarker panel as a noninvasive tool for IBD diagnosis.

Machine Learning-Based Prediction of Neurodegenerative Disease in Patients With Type 2 Diabetes by Derivation and Validation in 2 Independent Korean Cohorts: Model Development and Validation Study.

Several machine learning (ML) prediction models for neurodegenerative diseases (NDs) in type 2 diabetes mellitus (T2DM) have recently been developed. However, the predictive power of these models is limited by the lack of multiple risk factors. This study aimed to assess the validity and use of an ML model for predicting the 3-year incidence of ND in patients with T2DM. We used data from 2 independent cohorts-the discovery cohort (1 hospital; n=22,311) and the validation cohort (2 hospitals; n=2915)-to predict ND. The outcome of interest was the presence or absence of ND at 3 years. We selected different ML-based models with hyperparameter tuning in the discovery cohort and conducted an area under the receiver operating characteristic curve (AUROC) analysis in the validation cohort. The study dataset included 22,311 (discovery) and 2915 (validation) patients with T2DM recruited between 2008 and 2022. ND was observed in 133 (0.6%) and 15 patients (0.5%) in the discovery and validation cohorts, respectively. The AdaBoost model had a mean AUROC of 0.82 (95% CI 0.79-0.85) in the discovery dataset. When this result was applied to the validation dataset, the AdaBoost model exhibited the best performance among the models, with an AUROC of 0.83 (accuracy of 78.6%, sensitivity of 78.6%, specificity of 78.6%, and balanced accuracy of 78.6%). The most influential factors in the AdaBoost model were age and cardiovascular disease. This study shows the use and feasibility of ML for assessing the incidence of ND in patients with T2DM and suggests its potential for use in screening patients. Further international studies are required to validate these findings.

Deep learning predicted perceived age is a reliable approach for analysis of facial ageing: A proof of principle study.

Perceived age (PA) has been associated with mortality, genetic variants linked to ageing and several age-related morbidities. However, estimating PA in large datasets is laborious and costly to generate, limiting its practical applicability. To determine if estimating PA using deep learning-based algorithms results in the same associations with morbidities and genetic variants as human-estimated perceived age. Self-supervised learning (SSL) and deep feature transfer (DFT) deep learning (DL) approaches were trained and tested on human-estimated PAs and their corresponding frontal face images of middle-aged to elderly Dutch participants (n = 2679) from a population-based study in the Netherlands. We compared the DL-estimated PAs with morbidities previously associated with human-estimated PA as well as genetic variants in the gene MC1R; we additionally tested the PA associations with MC1R in a new validation cohort (n = 1158). The DL approaches predicted PA in this population with a mean absolute error of 2.84 years (DFT) and 2.39 years (SSL). In the training-test dataset, we found the same significant (p < 0.05) associations for DL PA with osteoporosis, ARHL, cognition, COPD and cataracts and MC1R, as with human PA. We also found a similar but less significant association for SSL and DFT PAs (0.69 and 0.71 years per allele, p = 0.008 and 0.011, respectively) with MC1R variants in the validation dataset as that found with human, SSL and DFT PAs in the training-test dataset (0.79, 0.78 and 0.71 years per allele respectively; all p < 0.0001). Deep learning methods can automatically estimate PA from facial images with enough accuracy to replicate known links between human-estimated perceived age and several age-related morbidities. Furthermore, DL predicted perceived age associated with MC1R gene variants in a validation cohort. Hence, such DL PA techniques may be used instead of human estimations in perceived age studies thereby reducing time and costs.

Predicting the severity of mycoplasma pneumoniae pneumonia in pediatric and adult patients: a multicenter study

The purpose of this study is to develop a nomogram model for early prediction of the severe mycoplasma pneumoniae pneumonia (SMPP) in Pediatric and Adult Patients. A retrospective analysis was conducted on patients with MPP, classifying them into SMPP and non-severe MPP (NSMPP) groups. A total of 550 patients (NSMPP 374 and SMPP 176) were enrolled in the study and allocated to training, validation cohorts. 278 patients (NSMPP 224 and SMPP 54) were retrospectively collected from two institutions and allocated to testing cohort. The risk factors for SMPP were identified using univariate analysis. For radiomic feature selection, Spearman’s correlation and the least absolute shrinkage and selection operator (LASSO) were utilized. Logistic regression was used to build different models, including clinical, imaging, radiomics, and integrated models (combining clinical, imaging, and radiomics features selected). The model’s discrimination was evaluated using a receiver operating characteristic curve, its calibration with a calibration curve, and the results were visualized using the Hosmer–Lemeshow goodness-of-fit test. Thirteen clinical features and fourteen imaging features were selected for constructing the clinical and imaging models. Simultaneously, a set of twenty-five radiomics features were utilized to build the radiomics model. The integrated model demonstrated good calibration and discrimination in the training cohorts (AUC, 0.922; 95% CI: 0.900, 0.942), validation cohorts (AUC, 0.879; 95% CI: 0.806, 0.920), and testing cohorts (AUC, 0.877; 95% CI: 0.836, 0.916). The discriminatory and predictive efficacy of the clinical model in testing cohorts increased further after clinical and radiological features were incorporated (AUC, 0.849 vs. 0.922, P = 0.002). The model demonstrated exemplary predictive efficacy for SMPP by leveraging a comprehensive set of inputs, encompassing clinical data, quantitative and qualitative radiological features, along with radiomics features. The integration of these three aspects in the predictive model further enhanced the performance of the clinical model, indicating the potential for extensive clinical applications.

Development and validation of a web-based nomogram for acute kidney injury in acute non-variceal upper gastrointestinal bleeding patients.

Acute kidney injury (AKI) is a common and serious complication in patients with acute non-variceal upper gastrointestinal bleeding (NVUGIB). Early prediction and intervention are crucial for improving patient outcomes. Data for patients presenting with acute NVUGIB in this retrospective study were sourced from the MIMC-IV database. Patients were randomly allocated into training and validation cohorts for further analysis. Independent predictors for AKI were identified using least absolute shrinkage and selection operator regression and multivariable logistic regression analyses in the training cohort. Based on the logistic regression results, a nomogram was developed to predict early AKI onset in acute NVUGIB patients, and implemented as a web-based calculator for clinical application. The nomogram's performance was evaluated through discrimination, using the C-index, calibration curves, and decision curve analysis (DCA) to assess its clinical value. The study involved 1082 acute NVUGIB patients, with 406 developing AKI. A multivariable logistic regression identified five key AKI predictors: CKD, use of human albumin, chronic liver disease, glucose, and blood urea nitrogen. The nomogram was constructed based on independent predictors. The nomogram exhibited robust accuracy, evidenced by a C-index of 0.73 in the training cohort and 0.72 in the validation cohort. Calibration curves demonstrated satisfactory concordance between predicted and observed AKI occurrences. DCA revealed that the nomogram offered considerable clinical benefit within a threshold probability range of 7% to 54%. Our nomogram is a valuable tool for predicting AKI risk in patients with acute NVUGIB, offering potential for early intervention and improved clinical outcomes.

A CT-based machine learning model for using clinical-radiomics to predict malignant cerebral edema after stroke: a two-center study.

This research aimed to create a machine learning model for clinical-radiomics that utilizes unenhanced computed tomography images to assess the likelihood of malignant cerebral edema (MCE) in individuals suffering from acute ischemic stroke (AIS). The research included 179 consecutive patients with AIS from two different hospitals. These patients were randomly assigned to training (n = 143) and validation (n = 36) sets with an 8:2 ratio. Using 3DSlicer software, the radiomics features of regions impacted by infarction were derived from unenhanced CT scans. The radiomics features linked to MCE were pinpointed through a consistency test, Student's t test and the least absolute shrinkage and selection operator (LASSO) method for selecting features. Clinical parameters associated with MCE were also identified. Subsequently, machine learning models were constructed based on clinical, radiomics, and clinical-radiomics. Ultimately, the efficacy of these models was evaluated by measuring the operating characteristics of the subjects through their area under the curve (AUCs). Logistic regression (LR) was found to be the most effective machine learning algorithm, for forecasting the MCE. In the training and validation cohorts, the AUCs of clinical model were 0.836 and 0.773, respectively, for differentiating MCE patients; the AUCs of radiomics model were 0.849 and 0.818, respectively; the AUCs of clinical and radiomics model were 0.912 and 0.916, respectively. This model can assist in predicting MCE after acute ischemic stroke and can provide guidance for clinical treatment and prognostic assessment.

Predicting cerebral edema in patients with spontaneous intracerebral hemorrhage using machine learning.

The early prediction of cerebral edema changes in patients with spontaneous intracerebral hemorrhage (SICH) may facilitate earlier interventions and result in improved outcomes. This study aimed to develop and validate machine learning models to predict cerebral edema changes within 72 h, using readily available clinical parameters, and to identify relevant influencing factors. An observational study was conducted between April 2021 and October 2023 at the Quzhou Affiliated Hospital of Wenzhou Medical University. After preprocessing the data, the study population was randomly divided into training and internal validation cohorts in a 7:3 ratio (training: N = 150; validation: N = 65). The most relevant variables were selected using Support Vector Machine Recursive Feature Elimination (SVM-RFE) and Least Absolute Shrinkage and Selection Operator (LASSO) algorithms. The predictive performance of random forest (RF), GDBT, linear regression (LR), and XGBoost models was evaluated using the area under the receiver operating characteristic curve (AUROC), precision-recall curve (AUPRC), accuracy, F1-score, precision, recall, sensitivity, and specificity. Feature importance was calculated, and the SHapley Additive exPlanations (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME) methods were employed to explain the top-performing model. A total of 84 (39.1%) patients developed cerebral edema changes. In the validation cohort, GDBT outperformed LR and RF, achieving an AUC of 0.654 (95% CI: 0.611-0.699) compared to LR of 0.578 (95% CI, 0.535-0.623, DeLong: p = 0.197) and RF of 0.624 (95% CI, 0.588-0.687, DeLong: p = 0.236). XGBoost also demonstrated similar performance with an AUC of 0.660 (95% CI, 0.611-0.711, DeLong: p = 0.963). However, in the training set, GDBT still outperformed XGBoost, with an AUC of 0.603 ± 0.100 compared to XGBoost of 0.575 ± 0.096. SHAP analysis revealed that serum sodium, HDL, subarachnoid hemorrhage volume, sex, and left basal ganglia hemorrhage volume were the top five most important features for predicting cerebral edema changes in the GDBT model. The GDBT model demonstrated the best performance in predicting 72-h changes in cerebral edema. It has the potential to assist clinicians in identifying high-risk patients and guiding clinical decision-making.

Prognostic implications of metabolism-related genes in acute myeloid leukemia.

Acute myeloid leukemia(AML) is a diverse malignancy with a prognosis that varies, being especially unfavorable in older patients and those with high-risk characteristics. Metabolic reprogramming has become a significant factor in AML development , presenting new opportunities for prognostic assessment and therapeutic intervention. Metabolism-related differentially expressed genes (mDEGs) were identified by integrating KEGG metabolic gene lists with AML gene expression data from GSE63270. Using TCGA data, we performed consensus clustering and survival analysis to investigate the prognostic significance of mDEGs. A metabolic risk model was constructed using LASSO Cox reg ression and enhanced by a nomogram incorporated clinical characteristics. The model was validated through receiver operating characteristic (ROC) curves and survival statistics. Gene network analysis was conducted to identify critical prognostic factors. The tumor immune microenvironment was evaluated using CIBERSORT and ESTIMATE algorithms, followed by correlation analysis between immune checkpoint gene expression and risk scores. Drug sensitivity predictions and in vitro assays were performed to explore the effects of mDEGs on cell proliferation and chemoresistance. An 11-gene metabolic prognostic model was established and validated. High-risk patients had worse overall survival in both training and validation cohorts (p < 0.05). The risk score was an independent prognostic factor. High-risk patients showed increased immune cell infiltration and potential response to checkpoint inhibitors but decreased drug sensitivity. The model correlated with sensitivity to drugs such as venetoclax. Carbonic anhydrase 13 (CA13) was identified as a key gene related to prognosis and doxorubicin resistance. Knocking down CA13 reduced proliferation and increased cell death with doxorubicin treatment. A novel metabolic gene signature was developed to stratify risk and predict prognosis in AML, serving as an independent prognostic factor. CA13 was identified as a potential therapeutic target. This study provides new insights into the prognostic and therapeutic implications of metabolic genes in AML.

A deep learning approach to case prioritisation of colorectal biopsies.

To create and validate a weakly supervised artificial intelligence (AI) model for detection of abnormal colorectal histology, including dysplasia and cancer, and prioritise biopsies according to clinical significance (severity of diagnosis). Triagnexia Colorectal, a weakly supervised deep learning model, was developed for the classification of colorectal samples from haematoxylin and eosin (H&E)-stained whole slide images. The model was trained on 24 983 digitised images and assessed by multiple pathologists in a simulated digital pathology environment. The AI application was implemented as part of a point and click graphical user interface to streamline decision-making. Pathologists assessed the accuracy of the AI tool, its value, ease of use and integration into the digital pathology workflow. Validation of the model was conducted on two cohorts: the first, on 100 single-slide cases, achieved micro-average model specificity of 0.984, micro-average model sensitivity of 0.949 and micro-average model F1 score of 0.949 across all classes. A secondary multi-institutional validation cohort, of 101 single-slide cases, achieved micro-average model specificity of 0.978, micro-average model sensitivity of 0.931 and micro-average model F1 score of 0.931 across all classes. Pathologists reflected their positive impressions on the overall accuracy of the AI in detecting colorectal pathology abnormalities. We have developed a high-performing colorectal biopsy AI triage model that can be integrated into a routine digital pathology workflow to assist pathologists in prioritising cases and identifying cases with dysplasia/cancer versus non-neoplastic biopsies.

Development and validation of a new tool to estimate early mortality in patients with advanced cancer treated with immunotherapy

BackgroundImmune checkpoint inhibitors (ICIs) are standard treatments for advanced solid cancers. Resistance to ICIs, both primary and secondary, poses challenges, with early mortality (EM) within 30–90 days indicating a lack of benefit. Prognostic factors for EM, including the lung immune prognostic index (LIPI), remain underexplored.MethodsWe performed a retrospective, observational study including patients affected by advanced solid tumors, treated with ICI as single agent or combined with other agents. Logistic regression models identified factors associated with EM and 90-day progression risks. A nomogram for predicting 90-day mortality was built and validated within an external cohort.ResultsIn total, 637 patients received ICIs (single agent or in combination with other drugs) for advanced solid tumors. Most patients were male (61.9%), with NSCLC as the prevalent tumor (61.8%). Within the cohort, 21.3% died within 90 days, 8.4% died within 30 days, and 34.5% experienced early progression. Factors independently associated with 90-day mortality included ECOG PS 2 and a high/intermediate LIPI score. For 30-day mortality, lung metastasis and a high/intermediate LIPI score were independent risk factors. Regarding early progression, high/intermediate LIPI score was independently associated. A predictive nomogram for 90-day mortality combining LIPI and ECOG PS achieved an AUC of 0.76 (95% CI 0.71–0.81). The discrimination ability of the nomogram was confirmed in the external validation cohort (n = 255) (AUC 0.72, 95% CI 0.64–0.80).ConclusionLIPI and ECOG PS independently were able to estimate 90-day mortality, with LIPI also demonstrating prognostic validity for 30-day mortality and early progression.

Accurate birth weight prediction from fetal biometry using the Gompertz model

ObjectivesMonitoring of fetal growth and estimation of birth weight is of clinical importance. During pregnancy, ultrasound fetal biometry values including femur length, head circumference, abdominal circumference, biparietal diameter are measured and used to place fetuses on “growth charts”. There is no simple growth-model-based, predictive formula in use for fetal biometry. Estimation of fetal weight at birth currently depends on ultrasound data taken a short time before birth. Study designOur cohort (“Seethapathy cohort”) consists of ultrasound biometry measurements and other data for 774 pregnant women in Chennai, India, 2015–2017. We use the Gompertz model, a standard model for constrained growth, with just three intuitive parameters, to model the growth of fetal biometry, and a machine learning (ML) model trained on these parameters to predict birth weight (BW). ResultsThe Gompertz model convincingly fits the growth of fetal biometry values. Two Gompertz parameters—t0 (inflection time) and c (rate of decrease of growth rate)—seem universal to all fetuses, while the third, A, is an overall scale specific to each fetus, capturing individual variation. On the Seethapathy cohort we can infer A for each fetus from ultrasound data available by the 24 or 35 weeks. Our ML model predicts birth weight with < 8 % error, outperforming published methods that have access to late-term ultrasound data. The same model gives an 8.4 % error in BW prediction on an independent validation cohort of 365 women. ConclusionsThe Gompertz model fits fetal biometry growth and enables birth weight estimation without need of late-term ultrasounds. Aside from its clinical predictive value, we suggest its use for future growth standards, with almost all variation described by a single scale parameter A.

Multicenter Development and Prospective Validation of eCARTv5: A Gradient Boosted Machine Learning Early Warning Score.

Early detection of clinical deterioration using machine learning early warning scores may improve outcomes. However, most implemented scores were developed using logistic regression, only underwent retrospective validation, and were not tested in important subgroups. Our objective was to develop and prospectively validate a gradient boosted machine model (eCARTv5) for identifying clinical deterioration on the wards. Multicenter retrospective and prospective observational study. Inpatient admissions to the medical-surgical wards at seven hospitals in three health systems for model development (2006-2022) and at 21 hospitals from three health systems for retrospective (2009-2023) and prospective (2023-2024) external validation. All adult patients hospitalized at each participating health system during the study years. None MEASUREMENTS AND MAIN RESULTS: Predictor variables (demographics, vital signs, documentation, and laboratory values) were used in a gradient boosted trees algorithm to predict intensive care unit transfer or death in the next 24 hours. The developed model (eCART) was compared to the Modified Early Warning Score (MEWS) and the National Early Warning Score (NEWS) using the area under the receiver operating characteristic curve (AUROC). The development cohort included 901,491 admissions, the retrospective validation cohort included 1,769,461 admissions, and the prospective validation cohort included 205,946 admissions. In retrospective validation, eCART had the highest AUROC (0.835; 95%CI 0.834, 0.835), followed by NEWS (0.766 (95%CI 0.766, 0.767)), and MEWS (0.704 (95%CI 0.703, 0.704)). eCART's performance remained high (AUROC ≥0.80) across a range of patient demographics, clinical conditions, and during prospective validation. We developed eCART, which performed better than the NEWS and MEWS retrospectively, prospectively, and across a range of subgroups. These results served as the foundation for Food and Drug Administration clearance for its use in identifying deterioration in hospitalized ward patients.

B-136 From Result to Response: The Development of Laboratory-Based Scoring Models to Predict COVID-19 Patient Outcomes

Abstract Background Patient characteristics relating to increased risk for COVID-19 severity are well-documented, including older age and pre-existing health concerns; however, methods to accurately predict patients’ acute reaction to SARS-CoV-2 infections remain lacking. Identification of specific laboratory tests informative of patient outcomes could effectively screen incoming patients and better inform physicians regarding optimal treatment plans. The aim of this study was to examine associations between frequently ordered laboratory markers and COVID-19 patient mortality to develop an objective laboratory-based scoring system that estimates patients’ risk of mortality. Methods Study participants included inpatients from 3 hospitals recruited into the GENCOV project based in Ontario, Canada. Participants (n=325) were at least 18 years of age, provided consent, and were hospitalized within 1 month of having a PCR confirmed COVID-19 infection between January 2020 and February 2022. Extensive clinical data including patient demographics, laboratory results, and treatment outcomes were extracted from patient’s electronic medical records (EMR). Results for 32 biochemical and hematological tests including complete blood cell counts, coagulation (activated partial thromboplastin time, D-dimer, fibrinogen, prothrombin time), general chemistry (albumin, blood gases, creatine kinase, electrolytes, glucose, triglycerides), inflammatory (lactate dehydrogenase, ferritin, C-reactive protein), liver (alanine aminotransferase, aspartate aminotransferase, total bilirubin), renal (creatinine, urea) and cardiac (troponin, NT-proBNP, BNP) markers were collected from each chart. Univariable logistic regression with nested likelihood ratio tests were used to determine significant associations between laboratory results and patient outcomes. Significant markers were incorporated into multivariable models and variable risk values were assigned. Total calculated risk scores for each participant were compared using univariable and multivariable risk values. Validation of each scoring method was performed on a 20% subset of inpatients. Receiver operating characteristic (ROC) curves evaluated model performance. Results Six laboratory markers were associated with COVID-19 patient mortality when controlling for age and sex. Univariable regression showed that elevated creatinine, elevated lactate, elevated white blood cell, low base excess, low bicarbonate, or low pH results upon admission were significantly associated with increased odds of mortality, in comparison to test results within the reference range for these same markers. Multivariable regression revealed fewer markers (only low bicarbonate and low base excess) showing significant associations with COVID-19 mortality. Area under the ROC curves (AUC) determined that risk scores derived from univariable values performed similarly to multivariable values in validation (0.800 vs 0.802) and development cohorts (0.821 vs 0.829). Total risk scores calculated from the univariable vs multivariable models suggest higher sensitivity (90% vs 85%) than specificity (58% vs 67%) in the validation cohort, whereas the development cohort had higher sensitivity (88%) than specificity (66%) with univariable scores, and higher specificity (81%) than sensitivity (74%) with multivariable scores. Conclusions Laboratory results associated with COVID-19 mortality following both univariable and multivariable analyses suggest hospitalized patients infected with SARS-CoV-2 may present with acidosis. Total risk scores derived from univariable and multivariable values performed similarly in predicting mortality; however, differences in marker significance between models indicate discrepancies with risk interpretation. Further comparison of the created risk score assessments with equivalent models in other populations is warranted.

A-311 Evaluating the Predictive Value of the sFlt-1/PlGF Ratio in Hypertensive Pregnancies for Preeclampsia

Abstract Background Preeclampsia (PE) is a hypertensive pregnancy disorder affecting around 5% of all pregnancies, and it continues to be a leading cause of maternal and neonatal morbidity and mortality in the United States. The incidence of hypertensive pregnancies is on the rise, attributed to the increasing prevalence of obesity and chronic hypertension. The development of a straightforward diagnostic test to accurately differentiate PE from gestational or chronic hypertension is crucial to avoid unnecessary hospitalizations. Despite the ongoing development of prediction models for PE, conventional clinical and laboratory measures have not been effective in predicting disease progression. Recent research has focused on the ratio of two serum biomarkers: soluble fms-like tyrosine kinase-1 (sFlt-1) and placental growth factor (PlGF), particularly in patients with a clinical suspicion of PE in acute settings. These studies have shown promise, with many indicating that the sFlt-1/PlGF ratio can effectively predict the presence or absence of PE. This study was conducted to assess the predictive value of the sFlt-1/PlGF ratio in pregnancies complicated by hypertension, aiming to contribute to the early detection and management of PE. Methods Pregnant individuals with hypertensive disorders at or above the age of 18 receiving prenatal care at NewYork-Presbysterian-Weill Cornell Hospital were enrolled in this study. We measured the levels of sFlt-1 and PlGF using fully automated Elecsys immunoassays (Roche Diagnostics). Subsequently, a prospective analysis of the serum sFlt-1/PlGF ratio was performed. A receiver operating characteristic (ROC) curve was used to calculate the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the ratio in predicting PE. To assess the prognostic accuracy of the sFlt-1/PlGF ratio, patients were randomly assigned to a development and a validation cohort at a ratio of 30:70. Results We collected a total of 137 serum samples of hypertensive pregnancies, with all ≥20 weeks of gestational age (GA). Of these, 47 (34.3%) developed PE within 2 weeks. The samples were divided into two cohorts: 98 were included in the development cohort, and 39 in the validation cohort. Within these groups, 34 (34.7%) of the development cohort and 13 (33.3%) of the validation cohort developed PE within 2 weeks, respectively. In the development cohort, the median sFlt-1/PlGF ratio was 85 (Interquartile range, [IQR], 44-164) among pregnant women who developed PE within 2 weeks. This was significantly higher compared to a median ratio of 8 (IQR, 3-35) observed in those who did not develop PE (p&amp;lt;0.001). The area under the ROC curve was 0.87, indicating good discriminative ability. An optimized cutoff ratio of 38 was determined, based on the max of combined sensitivity and specificity. When applied into the validation cohort, this cutoff yields a 90% (95% confidence interval [CI], 70-97) NPV and 58% (95%CI, 37-76) PPV with an 85% sensitivity (95%CI, 58-96) and a 69% specificity (95%CI, 50-83). Conclusions In women with a hypertensive disorder of pregnancy presenting GA≥20 weeks of gestation, the measurement of serum sFlt-1/PlGF ratio demonstrated clinical value in predicting the development of PE.

Machine learning derived model for the prediction of bleeding in dual antiplatelet therapy patients.

This study aimed to develop a predictive model for assessing bleeding risk in dual antiplatelet therapy (DAPT) patients. A total of 18,408 DAPT patients were included. Data on patients' demographics, clinical features, underlying diseases, past history, and laboratory examinations were collected from Affiliated Dongyang Hospital of Wenzhou Medical University. The patients were randomly divided into two groups in a proportion of 7:3, with the most used for model development and the remaining for internal validation. LASSO regression, multivariate logistic regression, and six machine learning models, including random forest (RF), k-nearest neighbor imputing (KNN), decision tree (DT), extreme gradient boosting (XGBoost), light gradient boosting machine (LGBM), and Support Vector Machine (SVM), were used to develop prediction models. Model prediction performance was evaluated using area under the curve (AUC), calibration curves, decision curve analysis (DCA), clinical impact curve (CIC), and net reduction curve (NRC). The XGBoost model demonstrated the highest AUC. The model features were comprised of seven clinical variables, including: HGB, PLT, previous bleeding, cerebral infarction, sex, Surgical history, and hypertension. A nomogram was developed based on seven variables. The AUC of the model was 0.861 (95% CI 0.847-0.875) in the development cohort and 0.877 (95% CI 0.856-0.898) in the validation cohort, indicating that the model had good differential performance. The results of calibration curve analysis showed that the calibration curve of this nomogram model was close to the ideal curve. The clinical decision curve also showed good clinical net benefit of the nomogram model. This study successfully developed a predictive model for estimating bleeding risk in DAPT patients. It has the potential to optimize treatment planning, improve patient outcomes, and enhance resource utilization.

A-161 Diagnosis of Lupus Anticoagulant Cases using Deep Neural Networks

Abstract Background Laboratory identification of lupus anticoagulant (LAC) is critical for diagnosing antiphospholipid antibody syndrome, a common cause of unprovoked thromboembolic events and pregnancy morbidity. Commonly used clot-based assays for LAC include the dilute Russell’s viper venom time (DRVVT) and activated partial thromboplastin time (APTT), but these are susceptible to anticoagulant interferences. Despite standardization efforts, LAC interpretation remains challenging due to pre- and post-analytical complexity and limited expertise. This study explored the use of deep neural networks (DNN) to facilitate LAC diagnosis. Methods Retrospective data for LAC profiling with 13 features performed at Mayo Clinic between 2021 and 2022 was collected. The profiles of 7,202 patients were randomly divided into training (n=4608), validation (n=1153), and test cohorts (n=1441). Diagnostic outputs including LAC positivity (determined by DRVVT or APTT) and the presence of warfarin (WAR) and heparin (HEP), were confirmed by expert assessment. Three DNN architectures were developed and evaluated (Figure 1): (1) four single output DNN models with a domain knowledge driven selection of input features, (2) four single output DNN models using the same 13 inputs across all models, and (3) a multi-output DNN model. Model performance was evaluated using the F1 score and the index of balanced accuracy (IBA) for each output label. Results Architecture 1 achieved high diagnostic accuracy with IBA values of 0.989-1.000 for all four outputs and F1 scores of 0.945-0.991. Architecture 2 exhibited F1 scores of 0.902-0.973 and IBA values of 0.935-0.992. Architecture 3 achieved F1 scores of 0.933-0.967 and IBA values of 0.941-0.983. Conclusions All three architectures accurately classify LAC and common anticoagulation effects. The comparable performance of architectures 1 and 2 suggests that manual feature extraction is unnecessary. Given its relative simplicity and versatility, architecture 3 is the preferred choice for implementation in a clinical laboratory to standardize LAC diagnosis.

Integrative Biomarker Panel for Improved Lung Cancer Diagnosis Using Plasma microRNAs and Sputum Bacterial DNA.

This study aimed to evaluate if integrating diverse molecular biomarkers in plasma and sputum could improve the diagnosis of lung cancer. The study analyzed miRNAs in plasma and bacterial DNA in sputum from 58 lung cancer patients and 62 cancer-free smokers using droplet digital PCR. The individual plasma miRNA and sputum bacterial biomarkers had sensitivities of 62-71% and specificities of 61-79% for diagnosing lung cancer. A panel of plasma miRNA or sputum bacterial biomarkers produced sensitivities of 79-85% and specificities of 74-82%. An integromic signature consisting of two miRNAs in plasma and three bacterial biomarkers in sputum had a higher sensitivity (87%) and specificity (89%) compared to individual biomarkers. The signature's diagnostic value was confirmed in a validation cohort of 56 lung cancer patients and 59 controls, independent of tumor stage, histological type, and demographic factors. Integrating diverse molecular biomarkers in plasma and sputum could improve the diagnosis of lung cancer.

Performance analysis of a deep-learning algorithm to detect the presence of inflammation in MRI of sacroiliac joints in patients with axial spondyloarthritis

ObjectivesTo assess the ability of a previously trained deep-learning algorithm to identify the presence of inflammation on MRI of sacroiliac joints (SIJ) in a large external validation set of patients...

A Novel Nomogram Developed Based on Preoperative Immune Inflammation-Related Indicators for the Prediction of Postoperative Delirium Risk in Elderly Hip Fracture Cases: A Single-Center Retrospective Cohort Study.

Postoperative delirium (POD) commonly occurs in elderly individuals following hip fracture surgery, with unclear pathophysiological mechanism. Inflammation is a known factor affecting the onset of delirium. The current work aimed to examine the associations of preoperative immune inflammation-related indicators with POD occurrence in elderly cases following hip fracture surgery. The current retrospective cohort study included 437 elderly cases administered hip fracture surgery from January 2018 to December 2023. The clinicodemographic data and laboratory findings of all cases were retrospectively analyzed. Immune inflammation-related indicators were assessed, eg, MLR, NLR and PLR, as well as SII and SIRI. The bootstrap method was employed to assign cases at 7:3 to the training (48 and 258 cases in the POD and no-POD groups, respectively) and internal validation (13 and 118 cases in the POD and no-POD groups, respectively) cohorts. Next, LASSO, univariable and multivariable logistic regression analyses were applied to determine risk factors in the training cohort, based on which a nomogram model was built. The obtained nomogram was examined for accuracy by calibration plot analysis. Finally, the nomogram's clinical value was assessed by decision curve analysis (DCA), followed by internal validation based on the training cohort. Of all 437 cases, 61 developed POD, indicating a POD incidence of 13.96%. LASSO regression and multivariable analyses revealed preoperative SIRI independently predicted POD in the training cohort. The developed nomogram had an area under the curve (AUC) of 0.991 (95% CI 0.983~0.998) in the training cohort versus 0.986 (95% CI 0.966~1.000) in the validation cohort. Calibration curve analysis revealed nomogram-predicted and actual probabilities were in line. DCA demonstrated the novel nomogram could confer net benefits for POD prediction in elderly cases administered hip fracture surgery. The immune inflammation-related indicators SIRI could predict POD in elderly cases following hip fracture surgery.