Area Under The ROC Curve Values Research Articles

Validating the CHAMPS Score: A Novel and Reliable Prognostic Score of Non-Variceal Upper Gastrointestinal Bleeding.

The Charlson Comorbidity Index ≥2, in-Hospital onset, Albumin <2.5 g/dL, altered Mental status, Eastern Cooperative Oncology Group Performance status ≥2, Steroid use (CHAMPS) score is a novel and promising prognostic tool. We present an initial external validation of the CHAMPS score for predicting mortality in acute nonvariceal upper gastrointestinal bleeding (NVUGIB) across multiple clinical outcomes. A prospective cohort study was conducted on adult patients with NVUGIB admitted to the Department of Gastroenterology between November 2022 and June 2023. The CHAMPS score performance in predicting in-hospital outcomes was evaluated by employing area under the receiver operating characteristic (AUROC) curves, followed by a comparative analysis with five pre-existing scores. A total of 140 patients were included in the study. The CHAMPS score showed its highest performance in predicting mortality rates (AUROC = 0.89), significantly outperforming the Glasgow-Blatchford Bleeding Score (GBS) as well as the Albumin level <3.0 mg/dL, International normalized ratio >1.5, altered Mental status, Systolic blood pressure ≤90 mmHg, and age >65 years (AIMS65) score (AUROC = 0.72 and 0.71, respectively; all p < 0.05). Subgroup analysis for bleeding-related and non-bleeding-related mortality further confirmed the robust predictive capability of the CHAMPS score (AUROC = 0.88 and 0.87, respectively). The CHAMPS score failed to predict rebleeding and intervention reliably, exhibiting AUROC values of 0.43 and 0.55, respectively. The optimal CHAMPS score cutoff value for predicting mortality was 3 points, achieving 100% sensitivity and 71.2% specificity. In the low-risk category defined by both CHAMPS and GBS scores, mortality and rebleeding rates were 0%. However, within the CHAMPS score-based low-risk group, 58.8% required intervention, contrasting with a 0% intervention rate for the GBS score-based low-risk group (GBS score ≤1). The CHAMPS score consistently demonstrated a robust predictive performance for mortality (AUROC > 0.8), facilitating the identification of high-risk patients requiring aggressive treatment and low-risk patients in need of localized treatment or safe discharge after successful bleeding control.

Predictive accuracy of wideband absorbance in children with large vestibular aqueduct syndrome: A single-center retrospective study

ObjectivesThis study aimed to assess the clinical significance of Wideband Absorbance (WBA) in children with Large Vestibular Aqueduct Syndrome (LVAS), which could potentially serve as diagnostic and predictive markers for LVAS in children. DesignThis was a single-center retrospective case-control study. Audiological measurements and Wideband Acoustic Immittance (WAI) were performed. Propensity score matching (PSM) was considered to treat group imbalance. The Receiver Operating Characteristic (ROC) curves and area under the ROC curve (AUC) were used to evaluate the sensitivity and specificity of WBA. Study sampleParticipants included 42 children with LVAS and 163 normal children aged 6 months −11 years recruited from clinical audiology settings between 2019 and 2021. ResultsThe WBA at Tympanometric Peak Pressure (WBATPP) and Ambient Pressure (WBAA) in the LVAS group were significantly lower than those of the control group at 1259–2000 Hz but higher at 4000–6349 Hz (p < 0.05, power >0.8). The WBAA (1587 Hz) AUC value was 0.805, identifying a score ≤0.565 as indicative of a LVAS risk. ConclusionsWBA holds promise in distinguishing LVAS from the normal condition and warrants further exploration as a tool to examine the influence of inner ear pressure on acoustic energy transmission in the middle ear.

Modeling landslide susceptibility using alternating decision tree and support vector

Globally, but especially in the Chinese Loess Plateau, landslides are considered to be one of the most severe and significant geological hazards. The purpose of this study is to design two ensemble machine learning methods, which are denoted as ADTree-Dagging and SVM-Dagging, for modeling landslide susceptibility in Lanzhou City (China). For this aim, the slope units extracted by the curvature watersheds method are used to construct landslide susceptibility modeling, and ten landslide conditioning factors are included in the landslide susceptibility evaluation (altitude, slope angle, slope aspect, cutting depth, surface roughness, relief amplitude, gully density, rainfall, distance to roads, and lithology). The conditioning factors selection and spatial correlation analysis were implemented by using the correlation attribute evaluation method and the frequency ratio model. The comprehensive performance of the models was tested using the receiver operating characteristic (ROC), area under the ROC curve (AUC), the root mean square error (RMSE), and several other performance metrics. For the training dataset, the results show that the SVM-Dagging model acquire the largest AUC value (0.953), lowest RMSE (0.3125), highest positive predictive value (96.0%), highest negative predictive value (91.2%), highest sensitivity (91.6%), highest specificity (95.8%), highest accuracy (93.6%), and highest Kappa (0.873). Similar results are observed in the validation dataset. Results demonstrated that the Dagging technique has improved significantly the prediction ability of SVM and ADTree models. The Dagging method can combine different models by leveraging the strengths of each model to create methods with higher flexibility than traditional machine learning methods. Therefore, in this study, the proposed new models can be applied for land-use planning and management of landslide susceptibility in the study area and in other areas containing similar geological conditions.

Multimodal explainable artificial intelligence identifies patients with non-ischaemic cardiomyopathy at risk of lethal ventricular arrhythmias

The efficacy of an implantable cardioverter-defibrillator (ICD) in patients with a non-ischaemic cardiomyopathy for primary prevention of sudden cardiac death is increasingly debated. We developed a multimodal deep learning model for arrhythmic risk prediction that integrated late gadolinium enhanced (LGE) cardiac magnetic resonance imaging (MRI), electrocardiography (ECG) and clinical data. Short-axis LGE-MRI scans and 12-lead ECGs were retrospectively collected from a cohort of 289 patients prior to ICD implantation, across two tertiary hospitals. A residual variational autoencoder was developed to extract physiological features from LGE-MRI and ECG, and used as inputs for a machine learning model (DEEP RISK) to predict malignant ventricular arrhythmia onset. In the validation cohort, the multimodal DEEP RISK model predicted malignant ventricular arrhythmias with an area under the receiver operating characteristic curve (AUROC) of 0.84 (95% confidence interval (CI) 0.71–0.96), a sensitivity of 0.98 (95% CI 0.75–1.00) and a specificity of 0.73 (95% CI 0.58–0.97). The models trained on individual modalities exhibited lower AUROC values compared to DEEP RISK [MRI branch: 0.80 (95% CI 0.65–0.94), ECG branch: 0.54 (95% CI 0.26–0.82), Clinical branch: 0.64 (95% CI 0.39–0.87)]. These results suggest that a multimodal model achieves high prognostic accuracy in predicting ventricular arrhythmias in a cohort of patients with non-ischaemic systolic heart failure, using data collected prior to ICD implantation.

Evaluation of alanine aminotransferase/aspartate aminotransferase ratio and high-density lipoprotein for predicting neonatal adverse outcomes associated with intrahepatic cholestasis of pregnancy.

To determine the association between lipid metabolism and intrahepatic cholestasis of pregnancy (ICP), and explore the value of maternal alanine aminotransferase/aspartate aminotransferase (ALT/AST) and high-density lipoprotein (HDL) in predicting adverse neonatal outcomes in women with ICP. A total of 147 pregnant women with ICP admitted to The Fourth Hospital of Shijiazhuang and 120 normal pregnant women in the same period were selected in this study. The Mann-Whitney U test and Chi-square tests were used to compare the differences in clinical data. Multivariate logistic regression was used to analyze the relationship between ALT/AST and the occurrence of adverse pregnancy outcomes in patients with ICP. The combined predictive value of ALT/AST and HDL was determined by receiver operating characteristic (ROC) curve analysis. Among 147 women with ICP, 122 women had total bile acid (TBA) levels of 10-39.9 µmol/L, and 25 had TBA ≥ 40 µmol/L. There was significantly lower gestational age in patients with severe ICP than in those with mild and control groups (all p<0.05), and the weight of newborns in the maternal ICP group was significantly lower than in the control group (p<0.05). Increasing TBA levels was associated with higher AST, ALT, ALT/AST, and lower HDL level (all p<0.05). Meanwhile, higher levels of ALT/AST was positively associated with neonatal hyperbilirubinemia [adjusted odds ratio (AOR) = 4.019, 95% CI [1.757-9.194, p=0.001] and cardiac injury [AOR = 3.500, 95% CI [1.535-7.987], p=0.003]. HDL was a significant protective factor for neonatal hyperbilirubinemia and cardiac injury [AOR = 0.315, 95% CI [0.126-0.788], p=0.014; AOR = 0.134 (0.039-0.461), p=0.001]. The area under the ROC curve (AUC) for prediction of neonatal hyperbilirubinemia by ALT/AST combined with HDL was 0.668 [95% CI [56.3-77.3%], p = 0.002], and the sensitivity and specificity were 47.1% and 84.0%, respectively. To predict neonatal cardiac injury, the AUC value was 0.668 [95% CI [56.4-77.1%], p = 0.002], with sensitivity and specificity were 41.2% and 87.1%, respectively. The levels of higher ALT/AST and lower HDL were significantly associated with the risk of ICP-related adverse neonatal outcomes. Moreover, ALT/AST combined with HDL has moderate clinical value in predicting the adverse outcomes of neonatal hyperbilirubinemia and cardiac injury.

Prediction of the potential distribution of Chimonobambusautilis (Poaceae, Bambusoideae) in China, based on the MaxEnt model.

Chimonobambusautilis is a unique edible bamboo species valued for its economic and nutritional benefits. However, its existence in natural habitats is at risk due to environmental shifts and human interventions. This research utilised the maximum entropy model (MaxEnt) to predict potential habitats for Ch.utilis in China, identifying key environmental factors influencing its distribution and analysing changes in suitable habitats under future climate conditions. The results show that the results of the MaxEnt model have high prediction accuracy, with an AUC (Area Under the receiver operating characteristic Curve) value of 0.997. Precipitation in the driest month (Bio14), altitude (Alt) and isothermality (Bio03) emerged as the primary environmental factors influencing the Ch.utilis distribution. Currently, the suitable habitats area for Ch.utilis is 10.55 × 104 km2. Projections for the 2050s and 2090s indicate potential changes in suitable habitats ranging from -3.79% to 10.52%. In general, the most suitable habitat area will decrease and shrink towards higher latitude areas in the future. This study provides a scientific basis for the introduction, cultivation and conservation of Ch.utilis.

Comparison of gene set scoring methods for reproducible evaluation of tuberculosis gene signatures

BackgroundBlood-based transcriptional gene signatures for tuberculosis (TB) have been developed with potential use to diagnose disease. However, an unresolved issue is whether gene set enrichment analysis of the signature transcripts alone is sufficient for prediction and differentiation or whether it is necessary to use the original model created when the signature was derived. Intra-method comparison is complicated by the unavailability of original training data and missing details about the original trained model. To facilitate the utilization of these signatures in TB research, comparisons between gene set scoring methods cross-data validation of original model implementations are needed.MethodsWe compared the performance of 19 TB gene signatures across 24 transcriptomic datasets using both rrebuilt original models and gene set scoring methods. Existing gene set scoring methods, including ssGSEA, GSVA, PLAGE, Singscore, and Zscore, were used as alternative approaches to obtain the profile scores. The area under the ROC curve (AUC) value was computed to measure performance. Correlation analysis and Wilcoxon paired tests were used to compare the performance of enrichment methods with the original models.ResultsFor many signatures, the predictions from gene set scoring methods were highly correlated and statistically equivalent to the results given by the original models. In some cases, PLAGE outperformed the original models when considering signatures’ weighted mean AUC values and the AUC results within individual studies.ConclusionGene set enrichment scoring of existing gene sets can distinguish patients with active TB disease from other clinical conditions with equivalent or improved accuracy compared to the original methods and models. These data justify using gene set scoring methods of published TB gene signatures for predicting TB risk and treatment outcomes, especially when original models are difficult to apply or implement.

Comparison of lung lesion assessment using free-breathing dynamic contrast-enhanced 1.5-T MRI with a golden-angle radial stack-of-stars VIBE sequence and CT.

Few studies have investigated the feasibility of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) using a free-breathing golden-angle radial stack-of-stars volume-interpolated breath-hold examination (FB radial VIBE) sequence in the lung. To investigate whether DCE-MRI using the FB radial VIBE sequence can assess morphological and kinetic parameters in patients with pulmonary lesions, with computed tomography (CT) as the reference. In total, 43 patients (30 men; mean age = 64 years) with one lesion each were prospectively enrolled. Morphological and kinetic features on MRI were calculated. The diagnostic performance of morphological MR features was evaluated using a receiver operating characteristic (ROC) curve. Kinetic features were compared among subgroups based on histopathological subtype, lesion size, and lymph node metastasis. The maximum diameter was not significantly different between CT and MRI (3.66 ± 1.62 cm vs. 3.64 ± 1.72 cm; P = 0.663). Spiculation, lobulation, cavitation or bubble-like areas of low attenuation, and lymph node enlargement had an area under the ROC curve (AUC) >0.9, while pleural indentation yielded an AUC of 0.788. The lung cancer group had significantly lower Ktrans, Ve, and initial AUC values than the other cause inflammation group (0.203, 0.158, and 0.589 vs. 0.597, 0.385, and 1.626; P < 0.05) but significantly higher values than the tuberculosis group (P < 0.05). Morphology features derived from FB radial VIBE have high correlations with CT, and kinetic analyses show significant differences between benign and malignant lesions. DCE-MRI with FB radial VIBE could serve as a complementary quantification tool to CT for radiation-free assessments of lung lesions.

Differentiating Benign From Malignant Ovarian Masses With Solid Components: Diagnostic Performance of CEUS Combined With IOTA Simple Rules and O-RADS

ObjectiveThis study aimed to apply the International Ovarian Tumor Analysis (IOTA) Simple Rules (SR), the Ovarian-Adnexal Reporting and Data System (O-RADS) and contrast-enhanced ultrasound (CEUS) in an identical cohort of Chinese patients and to analyze their performance in discrimination of ovarian masses with solid components. MethodsThis was a two-center retrospective study that included a total of 94 ovarian lesions in 86 women enrolled from January 2018 to February 2023. The lesions were classified by using the IOTA terminology and CEUS was performed for the lesions exhibiting solid components on ultrasonography, IOTA SR and O-RADS were applied, and CEUS images were analyzed retrospectively. We assessed the time to wash-in, time to peak intensity (PI), PI compared to myometrium, and time to wash-out, and observed statistically significant differences between benign and malignant lesions in the first three parameters. CEUS characteristics were employed to determine CEUS scores for benign (score 0) and malignant (score 3) lesions. Subsequently, the lesions were reassessed based on the IOTA SR and O-RADS classifications and CEUS scores. The sensitivity, specificity, and area under the receiver-operating-characteristics curve (AUC) of the different models were also determined. ResultsAmong the 94 ovarian lesions, 46 (48.9%) were benign and 48 (51.1%) were malignant. It was found that in the 60 lesions to which the SR could be applied, the sensitivity, specificity, and AUC was 0.900, 0.667, and 0.783, respectively. The sensitivity, specificity, and AUC of O-RADS was observed to be 1.000, 0.283 and 0.641, respectively. When SR and O-RADS were combined with CEUS, their sensitivity, specificity, and AUC values were increased to 0.917, 0.891, 0.904, and 0.958, 0.783, 0.871, respectively. ConclusionIOTA SR and O-RADS exhibited relatively low specificity in differentiating malignant from benign ovarian lesions with the solid components, and their diagnostic performance can be significantly improved when combined with CEUS.

AI Detection of Glottic Neoplasm Using Voice Signals, Demographics, and Structured Medical Records.

This study investigated whether artificial intelligence (AI) models combining voice signals, demographics, and structured medical records can detect glottic neoplasm from benign voice disorders. We used a primary dataset containing 2-3 s of vowel "ah", demographics, and 26 items of structured medical records (e.g., symptoms, comorbidity, smoking and alcohol consumption, vocal demand) from 60 patients with pathology-proved glottic neoplasm (i.e., squamous cell carcinoma, carcinoma in situ, and dysplasia) and 1940 patients with benign voice disorders. The validation dataset comprised data from 23 patients with glottic neoplasm and 1331 patients with benign disorders. The AI model combined convolutional neural networks, gated recurrent units, and attention layers. We used 10-fold cross-validation (training-validation-testing: 8-1-1) and preserved the percentage between neoplasm and benign disorders in each fold. Results from the AI model using voice signals reached an area under the ROC curve (AUC) value of 0.631, and additional demographics increased this to 0.807. The highest AUC of 0.878 was achieved when combining voice, demographics, and medical records (sensitivity: 0.783, specificity: 0.816, accuracy: 0.815). External validation yielded an AUC value of 0.785 (voice plus demographics; sensitivity: 0.739, specificity: 0.745, accuracy: 0.745). Subanalysis showed that AI had higher sensitivity but lower specificity than human assessment (p < 0.01). The accuracy of AI detection with additional medical records was comparable with human assessment (82% vs. 83%, p = 0.78). Voice signal alone was insufficient for AI differentiation between glottic neoplasm and benign voice disorders, but additional demographics and medical records notably improved AI performance and approximated the prediction accuracy of humans. NA Laryngoscope, 134:4585-4592, 2024.

MRI-Based Prediction of Clinical Improvement after Ventricular Shunt Placement for Normal Pressure Hydrocephalus: Development and Evaluation of an Integrated Multisequence Machine Learning Algorithm.

Symptoms of normal pressure hydrocephalus (NPH) are sometimes refractory to shunt placement, with limited ability to predict improvement for individual patients. We evaluated an MRI-based artificial intelligence method to predict postshunt NPH symptom improvement. Patients with NPH who underwent MRI before shunt placement at a single center (2014-2021) were identified. Twelve-month postshunt improvement in mRS, incontinence, gait, and cognition were retrospectively abstracted from clinical documentation. 3D deep residual neural networks were built on skull-stripped T2-weighted and FLAIR images. Predictions based on both sequences were fused by additional network layers. Patients from 2014-2019 were used for parameter optimization, while those from 2020-2021 were used for testing. Models were validated on an external validation data set from a second institution (n = 33). Of 249 patients, n = 201 and n = 185 were included in the T2-based and FLAIR-based models according to imaging availability. The combination of T2-weighted and FLAIR sequences offered the best performance in mRS and gait improvement predictions relative to models trained on imaging acquired by using only 1 sequence, with area under the receiver operating characteristic (AUROC) values of 0.7395 [0.5765-0.9024] for mRS and 0.8816 [0.8030-0.9602] for gait. For urinary incontinence and cognition, combined model performances on predicting outcomes were similar to FLAIR-only performance, with AUROC values of 0.7874 [0.6845-0.8903] and 0.7230 [0.5600-0.8859]. Application of a combined algorithm by using both T2-weighted and FLAIR sequences offered the best image-based prediction of postshunt symptom improvement, particularly for gait and overall function in terms of mRS.

Analysis of risk factors and construction of nomogram model for local lymph node metastasis in salivary gland mucoepidermoid carcinoma

Objective: To analyze the risk factors affecting regional lymph node metastasis in salivary gland mucoepidermoid carcinoma (MEC) and to establish a nomogram model for individually predicting lymph node metastasis in salivary gland MEC. Methods: The clinical data of 2 152 patients with salivary gland MEC from 1975 to 2020 were collected from the Surveillance, Epidemiology, and End Results (SEER) database of the National Cancer Institute. The collected data were divided into training cohort (1 506 cases) and validation cohort (646 cases) according to the ratio of 7∶3. Single-factor regression and multi-factor logistic regression were used to screen factors related to local lymph node metastasis in salivary gland MEC, with constructing of a nomogram. Calibration curve, receiver operating characteristic (ROC) curve, area under the ROC curve (AUC) and decision curve analysis were used to evaluate model performance in the validation cohort and the total cohort. Statistical tests were performed using SPSS (26.0) and R (4.3.0) software. Results: Multivariate logistic regression results showed that M stage [OR(95%CI):12.360(3.295-46.365), P=0.014], pathological grade Ⅱ、Ⅲ、Ⅳ[OR(95%CI): 1.956(1.329-2.879), 9.654(6.309-14.772), 9.298(6.072-14.238), P<0.001], T staging T2, T3, T4[OR(95%CI): 1.706(0.932-3.124), 3.021(1.790-5.096), 3.311(1.925-5.695), P<0.001], and gender [OR(95%CI):0.759(0.593-0.972), P=0.029] were independent factors affecting local lymph node metastasis in salivary gland MEC. Through verification in the validation cohort and the total cohort, the AUC values were greater than 0.8, and the calibration curve was close to the perfect reference line, proving that the constructed nomogram model had good specificity and sensitivity for predicting local lymph node metastasis in salivary gland MEC. Conclusion: M stage, pathological grade, T stage, and gender are risk factors for predicting regional lymph node metastasis and the established-nomogram has good predictive performance for local lymph node metastasis in salivary gland MEC.

External validation of three diabetes prediction scores in a Spanish cohort: does adding high risk for depression improve the validation of the FINDRISC score (FINDRISC-MOOD)?

ObjectivesTo evaluate the external validity of the FINDRISC, DESIR and ADA risk scores for the prediction of diabetes in a Spanish population aged >45 years and to test the possible...

Penggunaan Algoritma Gaussian Naïve Bayes &amp; Decision Tree Untuk Klasifikasi Tingkat Kemenangan Pada Game Mobile Legends

The development of technology and the internet has increased the popularity of online games, such as Mobile Legends. However, in competitions, players often experience defeat due to various factors, including player skills, team strategies, and the right hero selection. The right hero selection is very important to increase the chances of winning. Therefore, the Mobile Legends Professional League (MPL) has become a focus for competitive teams around the world. This study aimed to determine the classification of victory in MPL matches based on draft pick. Gaussian Naïve Bayes and Decision Tree were used as classification algorithm models in this study. The process in this study included cleaning data, data transformation (labeling), handling imbalanced data, scaling, splitting, and hyperparameter. The evaluation stage used confusion matrix, correlation data, and AU-ROC curve. The results of this study showed that the Decision Tree method had better performance than Gaussian Naïve Bayes in classifying data using the confusion matrix. The AUC (area under the receiver operating characteristic curve) analysis showed that the decision tree had better performance than Gaussian naive Bayes in predicting positive and negative data. This is indicated by the higher AUC value for the Decision Tree, which is 0.67 compared to Gaussian Naïve Bayes which is 0.48. Classification models with higher AUC values can more accurately distinguish between positive and negative data. In this study, the Decision Tree had a higher AUC value than Gaussian Naïve Bayes so the Decision Tree could more accurately classify victory and defeat data.

Impact Exploration of Spatiotemporal Feature Derivation and Selection on Machine Learning-Based Predictive Models for Post-Embolization Cerebral Aneurysm Recanalization.

To enhance the performance of machine learning (ML) models for the post-embolization recanalization of cerebral aneurysms, we evaluated the impact of hemodynamic feature derivation and selection method on six ML algorithms. We utilized computational fluid dynamics (CFD) to simulate hemodynamics in 66 cerebral aneurysms from 65 patients, including 57 stable and nine recanalized aneurysms. We derived a total of 107 features for each aneurysm, encompassing four clinical features, 12 morphological features, and 91 hemodynamic features. To investigate the influence of feature derivation and selection methods on the ML models, we employed two derivation methods, simplified and fully derived, in combination with four selection methods: all features, statistically significant analysis, stepwise multivariate logistic regression analysis (stepwise-LR), and recursive feature elimination (RFE). Model performance was assessed using the area under the receiver operating characteristic curve (AUROC) and precision-recall curve (AUPRC) on both the training and testing datasets. The AUROC values on the testing dataset exhibited a wide-ranging spectrum, spanning from 0.373 to 0.863. Fully derived features and the RFE selection method demonstrated superior performance in intra-model comparisons. The multi-layer perceptron (MLP) model, trained with RFE-selected fully derived features, achieved the best performance on the testing dataset, with an AUROC value of 0.863 (95% CI: 0.684- 1.000). Our study demonstrated the importance of feature derivation and selection in determining the performance of ML models. This enabled the development of accurate decision-making models without the need to invade the patient.

Machine learning models on a web application to predict short-term postoperative outcomes following anterior cervical discectomy and fusion

BackgroundThe frequency of anterior cervical discectomy and fusion (ACDF) has increased up to 400% since 2011, underscoring the need to preoperatively anticipate adverse postoperative outcomes given the procedure’s expanding use. Our study aims to accomplish two goals: firstly, to develop a suite of explainable machine learning (ML) models capable of predicting adverse postoperative outcomes following ACDF surgery, and secondly, to embed these models in a user-friendly web application, demonstrating their potential utility.MethodsWe utilized data from the National Surgical Quality Improvement Program database to identify patients who underwent ACDF surgery. The outcomes of interest were four short-term postoperative adverse outcomes: prolonged length of stay (LOS), non-home discharges, 30-day readmissions, and major complications. We utilized five ML algorithms - TabPFN, TabNET, XGBoost, LightGBM, and Random Forest - coupled with the Optuna optimization library for hyperparameter tuning. To bolster the interpretability of our models, we employed SHapley Additive exPlanations (SHAP) for evaluating predictor variables’ relative importance and used partial dependence plots to illustrate the impact of individual variables on the predictions generated by our top-performing models. We visualized model performance using receiver operating characteristic (ROC) curves and precision-recall curves (PRC). Quantitative metrics calculated were the area under the ROC curve (AUROC), balanced accuracy, weighted area under the PRC (AUPRC), weighted precision, and weighted recall. Models with the highest AUROC values were selected for inclusion in a web application.ResultsThe analysis included 57,760 patients for prolonged LOS [11.1% with prolonged LOS], 57,780 for non-home discharges [3.3% non-home discharges], 57,790 for 30-day readmissions [2.9% readmitted], and 57,800 for major complications [1.4% with major complications]. The top-performing models, which were the ones built with the Random Forest algorithm, yielded mean AUROCs of 0.776, 0.846, 0.775, and 0.747 for predicting prolonged LOS, non-home discharges, readmissions, and complications, respectively.ConclusionsOur study employs advanced ML methodologies to enhance the prediction of adverse postoperative outcomes following ACDF. We designed an accessible web application to integrate these models into clinical practice. Our findings affirm that ML tools serve as vital supplements in risk stratification, facilitating the prediction of diverse outcomes and enhancing patient counseling for ACDF.

Application of Machine Learning Algorithms to Predict Osteoporotic Fractures in Women.

Predicting the risk of osteoporotic fractures is vital for prevention. Traditional methods such as the Fracture Risk Assessment Tool (FRAX) model use clinical factors. This study examined the predictive power of the FRAX score and machine-learning algorithms trained on FRAX parameters. We analyzed the data of 2,147 female participants from the Ansan cohort study. The FRAX parameters employed in this study included age, sex (female), height and weight, current smoking status, excessive alcohol consumption (>3 units/d of alcohol), and diagnosis of rheumatoid arthritis. Osteoporotic fracture was defined as one or more fractures of the hip, spine, or wrist during a 10-year observation period. Machine-learning algorithms, such as gradient boosting, random forest, decision tree, and logistic regression, were employed to predict osteoporotic fractures with a 70:30 training-to-test set ratio. We evaluated the area under the receiver operating characteristic curve (AUROC) scores to assess and compare the performance of these algorithms with the FRAX score. Of the 2,147 participants, 3.5% experienced osteoporotic fractures. Those with fractures were older, shorter in height, and had a higher prevalence of rheumatoid arthritis, as well as higher FRAX scores. The AUROC for the FRAX was 0.617. The machine-learning algorithms showed AUROC values of 0.662, 0.652, 0.648, and 0.637 for gradient boosting, logistic regression, decision tree, and random forest, respectively. This study highlighted the immense potential of machine-learning algorithms to improve osteoporotic fracture risk prediction in women when complete FRAX parameter information is unavailable.

A CT based radiomics analysis to predict the CN0 status of thyroid papillary carcinoma: a two- center study

ObjectivesTo develop and validate radiomics model based on computed tomography (CT) for preoperative prediction of CN0 status in patients with papillary thyroid carcinoma (PTC).MethodsA total of 548 pathologically confirmed LNs (243 non-metastatic and 305 metastatic) two distinct hospitals were retrospectively assessed. A total of 396 radiomics features were extracted from arterial-phase CT images, where the strongest features containing the most predictive potential were further selected using the least absolute shrinkage and selection operator (LASSO) regression method. Delong test was used to compare the AUC values of training set, test sets and cN0 group.ResultsThe Rad-score showed good discriminating performance with Area Under the ROC Curve (AUC) of 0.917(95% CI, 0.884 to 0.950), 0.892 (95% CI, 0.833 to 0.950) and 0.921 (95% CI, 868 to 0.973) in the training, internal validation cohort and external validation cohort, respectively. The test group of CN0 with a AUC of 0.892 (95% CI, 0.805 to 0.979). The accuracy was 85.4% (sensitivity = 81.3%; specificity = 88.9%) in the training cohort, 82.9% (sensitivity = 79.0%; specificity = 88.7%) in the internal validation cohort, 85.4% (sensitivity = 89.7%; specificity = 83.8%) in the external validation cohort, 86.7% (sensitivity = 83.8%; specificity = 91.3%) in the CN0 test group.The calibration curve demonstrated a significant Rad-score (P-value in H-L test > 0.05). The decision curve analysis indicated that the rad-score was clinically useful.ConclusionsRadiomics has shown great diagnostic potential to preoperatively predict the status of cN0 in PTC.

Predictive value of the BDH2-MN2 nomogram model for prognosis at 3 months after receiving intravenous thrombolysis in patients with acute ischemic stroke.

The present study focused on developing a nomogram model to predict the 3-month survival of patients with acute ischemic stroke (AIS) receiving intravenous thrombolysis with tissue plasminogen activator (tPA). A total of 709 patients were enrolled in the present study, including 496 patients in the training set and 213 patients in the validation set. All data were statistically analyzed using R software. We applied LASSO regression analysis to construct nomograms by screening statistically significant predictors from all variables.The model discrimination was evaluated based on the area under the receiver operating characteristic curve (AUC-ROC). LASSO regression analysis was conducted for all variables, which revealed BNP, DNT, HCY, HDL, MHR, NHR and post-thrombolysis NIHSS as independent predictors of adverse outcomes at 3 months after intravenous thrombolysis. Accordingly, these seven factors were incorporated in the nominated BDH2-MN2 nomogram. The resulting AUC-ROC values determined for the training and validation sets were 0.937 (95% CI: 0.822-0.954) and 0.898 (95% CI: 0.748-0.921), respectively. A robust BDH2-MN2 (BNP, DNT, HCY, HDL, MHR, NHR and post-thrombolysis NIHSS) nomogram model was successfully developed and validated. The developed nomogram enables prediction of adverse outcomes of individual AIS patients receiving intravenous thrombolysis with alteplase for 3 months.

Use of MRI Radiomics Models in Evaluating the Low HER2 Expression in Breast Cancer.

To investigate the magnetic resonance imaging (MRI) radiomics models in evaluating the human epidermal growth factor receptor 2(HER2) expression in breast cancer. The MRI data of 161 patients with invasive ductal carcinoma (non-special type) of breast cancer were retrospectively collected, and the MRI radiomics models were established based on the MRI imaging features of the fat suppression T2 weighted image (T2WI) sequence, dynamic contrast-enhanced (DCE)-T1WIsequence and joint sequences. The T-test and the least absolute shrinkage and selection operator (LASSO) algorithm were used for feature dimensionality reduction and screening, respectively, and the random forest (RF) algorithm was used to construct the classification model. The model established by the LASSO-RF algorithm was used in the ROC curve analysis. In predicting the low expression state of HER2 in breast cancer, the radiomics models of the fat suppression T2WI sequence, DCE-T1WI sequence, and the combination of the two sequences showed better predictive efficiency. In the receiver operating characteristic (ROC) curve analysis for the verification set of low, negative, and positive HER2 expression, the area under the ROC curve (AUC) value was 0.81, 0.72, and 0.62 for the DCE-T1WI sequence model, 0.79, 0.65 and 0.77 for the T2WI sequence model, and 0.84, 0.73 and 0.66 for the joint sequence model, respectively. The joint sequence model had the highest AUC value. The MRI radiomics models can be used to effectively predict the HER2 expression in breast cancer and provide a non-invasive and early assistant method for clinicians to formulate individualized and accurate treatment plans