AUROC Curve Research Articles

Fasting intact insulin by mass spectrometry is associated with metabolic dysfunction-associated steatotic liver disease in youth.

Fasting intact insulin concentrations can predict metabolic dysfunction-associated steatotic liver disease (MASLD) in adults without diabetes; however, research in youth is limited. We sought to determine whether fasting intact insulin, measured by liquid chromatography-tandem mass spectrometry, is associated with MASLD in children. This retrospective cross-sectional analysis used data and samples from children who participated in studies across 3 universities between 2014 and 2022. Key measurements included fasting intact insulin, ALT, and hepatic steatosis assessed by MRI techniques. MASLD was defined as hepatic steatosis ≥5% by MRI with at least 1 cardiometabolic risk factor. The optimal cutoff points to identify MASLD were determined by maximizing the Youden index, and the AUROC curves were compared using the DeLong test. The analysis included 184 children (28% male; 14.9 ± 2.6y; 57% Hispanic race/ethnicity; body mass index 32.5 ± 8.1kg/m2; 64% with MASLD, 43% with polycystic ovary syndrome, and 5% with other liver diseases). Fasting intact insulin and ALT levels were significantly higher in children with MASLD (p < 0.05). Fasting intact insulin was strongly associated with MASLD with an AUROC of 0.83 (0.77-0.90), sensitivity of 71%, and specificity of 85%. When combined with ALT (intact insulin × ALT [μU/mL × U/L]), the AUROC was 0.88 (0.83-0.94), with a sensitivity of 89% and specificity of 81%. The improvement in AUROC over intact insulin alone was not statistically significant (p = 0.089) but was statistically significant from ALT (p = 0.022). Optimal cutoff points for intact insulin and intact insulin × ALT were 20μU/mL and 522μU/mL × U/L, respectively. In pediatric patients, measurements of fasting intact insulin alone and combined with ALT provide a noninvasive strategy for identifying the presence of MASLD.

The diagnostic value of peripheral blood lymphocyte testing in children with infectious mononucleosis

ObjectiveTo investigate the diagnostic value of peripheral blood lymphocyte testing in children with infectious mononucleosis (IM).MethodsA total of 135 children with IM as the IM group and 100 healthy volunteers as the healthy group were included in this retrospective study. Peripheral blood lymphocyte subsets marked as CD3+, CD4+, CD8+, CD16 + CD56+, and CD19 + in the peripheral blood were quantified using flow cytometry. Statistical analysis was performed using the chi-square test, Kruskal-Wallis test, AUROC curve, and Kappa consistency test to assess the diagnostic value of these markers in IM.ResultsThe AUROC curve for CD8 + cells and for CD4+/CD8 + ratios both achieved a value of 1 with the sensitivity and specificity of 100% (P<0.001). The Kappa coefficients were 1 for CD8+, CD4+/CD8 + ratios and the combined EBV analysis, indicating a 100% consistency with the clinical diagnosis. Significant differences were also observed in the CD3+, CD4+, CD16 + CD56+, and CD19 + lymphocyte subsets between the IM group and the healthy group (P<0.05).ConclusionThe evaluation of CD8 + and CD4+/CD8 + ratios in peripheral blood lymphocytes represents a significant advancement in the diagnosis of IM. Peripheral blood lymphocyte testing offers a reliable, sensitive, and specific diagnostic tool to enhance the clinical management of children with IM.

ACEF score and lactate: lifeline predictors in endocarditis valve procedures: insights from a single-center study.

This study aimed to assess the prognostic value of routinely collected laboratory parameters, specifically lactate, troponin-T, and CK-MB, in predicting mortality in patients with surgically treated endocarditis. Additionally, the study evaluated the effectiveness of two mortality scores, EuroSCORE II and ACEF II Score, in this clinical context. We retrospectively analyzed data from 130 patients diagnosed with endocarditis who underwent surgery at a single tertiary center over nine years. The study utilized preoperative mortality scores and laboratory parameters collected within the first 24h post-surgery. Statistical analyses included AUROC curves, Kaplan-Meier survival analyses, and correlation analyses to determine predictive values and associations with patient outcomes. Among the 130 patients, 28 (21.5%) died within one year post-surgery. Elevated postoperative lactate levels were significantly associated with increased short- and long-term mortality, with AUROC values indicating strong predictive capability. The ACEF II Score also demonstrated significant predictive value for mortality at various time points, outperforming EuroSCORE II. Higher lactate levels and ACEF II Scores correlated with increased hazard ratios for mortality. Kaplan-Meier analyses revealed significant survival differences based on lactate and ACEF II Score thresholds. Postoperative lactate levels and the ACEF II Score are robust predictors of mortality in patients undergoing cardiac surgery for endocarditis. Integrating these parameters into clinical practice can enhance risk stratification and guide therapeutic decisions, improving patient outcomes through personalized care. Further studies are needed to validate these findings across diverse populations and explore additional biomarkers for refined predictive accuracy.

Thickness Speed Progression Index (TSPI): Machine Learning Approach for Keratoconus Detection: Thickness-based machine learning model for KC detection

To develop and validate a pachymetry-based machine learning index for differentiating keratoconus, keratoconus suspect, and normal corneas. Development and validation of a machine learning diagnostic algorithm. This retrospective study included 349 eyes of 349 patients with normal, frank keratoconus (KC), and keratoconus suspect (KCS) corneas. KCS corneas included topographically/tomographically normal (TNF) and borderline fellow eyes (TBF) of patients with asymmetric keratoconus. Six parameters were derived from the Corneal Thickness Progression map on the Galilei Dual Scheimpflug-Placido system and fed into a machine-learning algorithm to create the Thickness Speed Progression Index (TSPI). The model was trained with 5-fold cross-validation using a random search over 7 different machine learning algorithms, and the best model and hyperparameters were selected. 133 normal eyes, 141 KC eyes, and 75 KCS eyes, subdivided into 34 TNF and 41 TBF eyes, were included. In experiment 1 (normal and KC), the best model (Random Forest) achieved an accuracy of 100% and AUROC of 1.00 for both normal and KC groups. In experiment 2 (normal, KCS, and KC), the model achieved an overall accuracy of 91%, and AUROC curves of 0.93, 0.83, and 0.99 in detecting normal, KCS, and KC corneas respectively. In experiment 3 (normal, TNF, TBF, and KC), the model achieved an accuracy of 87% with AUROC curves of 0.91, 0.60, 0.77, and 0.94 for normal, TNF, TBF, and KC corneas, respectively. Using data solely based on pachymetry, machine learning algorithms such as the TSPI are able to discriminate normal corneas from keratoconus and keratoconus suspects corneas with reasonable accuracy.

Improving the prediction of major adverse cardiovascular events in patients with hypertrophic cardiomyopathy by adding plasma SVEP1 to conventional clinical model including NT-proBNP

Abstract Background Hypertrophic cardiomyopathy (HCM) is the most common genetic heart disease with a prevalence between 1 in 200 to 500 [1]. HCM can lead to major adverse cardiovascular events (MACE) such as heart failure (HF) hospitalization and cardiac death [2,3]. A protein named "sushi, von Willebrand factor type A, EGF and pentraxin domain containing 1" (SVEP1) is a large extracellular matrix protein whose activity can be measured in plasma [4,5]. SVEP1 promotes inflammation and atherosclerosis [4,5]. SVEP1 is associated with risk of HF hospitalization and cardiac death in patients with HF with reduced ejection fraction [6]. However, no prior study has assessed the prognostic value of SVEP1 in patients with HCM. Purpose To test the hypothesis that the addition of SVEP1 improves the predictive performance of the conventional model based on clinical risk factors plus NT-proBNP for future MACE in patients with HCM. Methods We performed a multicenter prospective cohort study of patients with HCM [7]. The outcome was MACE, defined as a composite of HF hospitalization or cardiac death [6]. After dividing the cohort into four groups using the median values of SVEP1 and NT-proBNP, we compared the risk of the outcome event using the Cox proportional hazards model adjusting for 15 clinical risk factors known to be associated with MACE in HCM (Figure 1) [8-11]. We also developed the Lasso-regularized Cox proportional hazards model to predict the time to first MACE by adding SVEP1 to the 15 clinical risk factors with or without NT-proBNP in the training cohort (randomly selected 75% of the cohort). We then compared the predictive performance in the test cohort (25% of the cohort) using the C-statistic of each model. Results During a median follow-up of 1.9 (25 percentile-75 percentile, 0.7-4.3) years, the outcome event occurred in 63 (10%) of 610 patients. The four groups stratified by the median of SVEP1 and NT-proBNP levels had different risks of the outcome event (log-rank p&amp;lt;0.001: Figure 1). Even in the groups with lower-than-median NT-proBNP levels, the high-SVEP1 group had higher risks of MACE compared with the low-SVEP1 group (adjusted hazard ratio 4.52, p=0.042: Figure 1). In the prediction of time to first MACE, the addition of SVEP1 improved the C-statistic of the clinical model (p&amp;lt;0.001) and that of the clinical plus NT-proBNP model (p=0.01: Figure 2). The clinical plus SVEP1 model also outperformed the clinical plus NT-proBNP model (p=0.005: Figure 2). Conclusions In the present multicenter prospective cohort study of 610 patients with HCM, SVEP1 improved MACE risk stratification when added to the conventional model consisting of clinical risk factors plus NT-proBNP. SVEP1 also exhibited a superior predictive ability than NT-proBNP to predict MACE in patients with HCM. These data collectively indicate that SVEP1 provides additional risk stratification and may be a better biomarker than NT-proBNP for the prognostication of patients with HCM.Figure 1.Kaplan-Meier curves for MACEFigure 2.Time-dependent AUROC curves

Enhancing early detection of autistic spectrum disorder in children using machine learning approaches

Diagnosing Autism Spectrum Disorder (ASD) presents a multifaceted challenge, demanding accurate and efficient screening methods. Applying machine learning techniques offers a promising avenue for enhancing diagnostic accuracy and efficiency. This research investigates the efficiency of machine learning in distinguishing individuals with ASD from those without, utilizing a comprehensive dataset comprising screening questions, demographic factors, and ASD related diagnostic classifications. We applied chi-square feature selection technique and also tested Random Forest, Logistic Regression, Gradient Boosting Classifier, and Extra Trees Classifier. Each model showed optimal performance and exhibit high precision, recall, and F1-score for both ASD-positive and ASD-negative instances. Additionally, AUROC curves further validated the models’ exceptional discriminatory abilities, with exceptional results. Our findings highlight the potential of machine learning algorithms for enhancing ASD diagnosis accuracy and efficiency in clinical settings. Further research and validation on larger datasets are required to understand the importance of machine learning methods in ASD diagnosis.

Validation of SIPI and comparison with objective nursing workload

ABSTRACT Background Accurately estimating nursing care workload upon hospital admission is crucial for resource allocation. Despite the introduction of tools like the ‘Information System of Nursing Performance’ (SIPI) in Italy, studies evaluating their effectiveness are lacking. Aim This study aimed to validate SIPI based on nursing care demands in an internal medicine ward. Design Single-centre, observational study of all patients admitted to an internal medicine department in an Italian hospital from 1 September to 31 December 2022. Methods Nursing activities were documented over the first 3 days of hospitalization using a standardized form, measured as performance/5 min/patient/day. Multivariate analysis assessed SIPI’s association with nursing workload. SIPI’s ability to predict excessive nurse workload was evaluated using the AUROC curve. Results The study included 306 patients. Mean SIPI value was 53.9 (20.6). Daily average care demands varied by SIPI category: 20.7 performances in very low, 26.8 in medium-low, 28.7 in medium-high, and 32.4 in very high (P = 0.001). SIPI was independently associated with increased care demands (OR: 1.024; 1.002–1.047, P = 0.033). The AUROC for SIPI predicting excessive workload was 0.614 (95% CI: 0.545–0.683). Conclusions SIPI is useful for identifying nursing workload and correlates with patient care demands.

Using machine learning to predict the risk of short-term and long-term death in acute kidney injury patients after commencing CRRT

BackgroundThe mortality rate and prognosis of short-term and long-term acute kidney injury (AKI) patients who undergo continuous renal replacement therapy (CRRT) are different. Setting up risk stratification tools for both short-term and long-term deaths is highly important for clinicians.MethodA total of 1535 AKI patients receiving CRRT were included in this study, with 1144 from the training set (the Dryad database) and 391 from the validation set (MIMIC IV database). A model for predicting mortality within 10 and 90 days was built using nine different machine learning (ML) algorithms. AUROC, F1-score, accuracy, sensitivity, specificity, precision, and calibration curves were used to assess the predictive performance of various ML models.ResultsA total of 420 (31.1%) deaths occurred within 10 days, and 1080 (68.8%) deaths occurred within 90 days. The random forest (RF) model performed best in both predicting 10-day (AUROC: 0.80, 95% CI: 0.74–0.84; accuracy: 0.72, 95% CI: 0.67–0.76; F1-score: 0.59) and 90-day mortality (AUROC: 0.78, 95% CI: 0.73–0.83; accuracy: 0.73, 95% CI: 0.69–0.78; F1-score: 0.80). The importance of the feature shows that SOFA scores are rated as the most important risk factor for both 10-day and 90-day mortality.ConclusionOur study, utilizing multiple machine learning models, estimates the risk of short-term and long-term mortality among AKI patients who commence CRRT. The results demonstrated that the prognostic factors for short-term and long-term mortality are different. The RF model has the best prediction performance and has valuable potential for clinical application.

Stroop test, Quickstroop, and the 1-min animal naming test for minimal hepatic encephalopathy diagnosis: A multicenter study in Mexico

Introduction and ObjectivesAlthough the Psychometric Hepatic Encephalopathy Score (PHES) remains the gold standard in diagnosing minimal hepatic encephalopathy (MHE), its complexity limits its application in clinical practice. While more convenient tests, such as the Stroop test, Quickstroop, and the 1-min animal naming test (ANT-1), have emerged, they haven't been validated in our setting. Our objective was to validate these tests in our population. Patients and MethodsThis multicenter, observational, descriptive, and cross-sectional study was conducted in three hospitals in northeastern Mexico. MHE was defined as a PHES <-4. We included patients with cirrhosis aged >15 years without a history of overt hepatic encephalopathy. Data regarding sex, age, education, Child-Pugh/MELD-Na scores, etiology of cirrhosis, diabetes, hypertension, obesity, ascites, and clinically significant portal hypertension was collected. Fisher's exact test, Mann-Whitney U test, and receiver operating characteristic (ROC) curves were used for statistical analysis. ResultsOf the 121 patients included, 35.5 % were diagnosed with MHE. The presence of MHE was significantly associated with education level, years of study, and scores in the Stroop test, Quickstroop, and ANT-1. The AUROC curves were 77.9 %, 74.6 %, and 72.7 % for the Stroop test, Quickstroop, and ANT-1, respectively. The resulting cut-off points were 218.398 (sensitivity: 74 %; specificity: 74 %), 40.535 (sensitivity: 77 %; specificity: 68 %), and <16 animals (sensitivity: 58 %; specificity: 79 %), respectively. ConclusionsThese tests are valid diagnostic tools for detecting MHE in our population. Their simpler use and applicability could increase the early diagnosis of MHE and prompt primary prophylaxis initiation for overt hepatic encephalopathy.

Accurate prediction of all-cause mortality in patients with metabolic dysfunction-associated steatotic liver disease using electronic health records

Introduction and ObjectivesDespite the huge clinical burden of MASLD, validated tools for early risk stratification are lacking, and heterogeneous disease expression and a highly variable rate of progression to clinical outcomes result in prognostic uncertainty. We aimed to investigate longitudinal electronic health record-based outcome prediction in MASLD using a state-of-the-art machine learning model. Patients and Methodsn = 940 patients with histologically-defined MASLD were used to develop a deep-learning model for all-cause mortality prediction. Patient timelines, spanning 12 years, were fully-annotated with demographic/clinical characteristics, ICD-9 and -10 codes, blood test results, prescribing data, and secondary care activity. A Transformer neural network (TNN) was trained to output concomitant probabilities of 12-, 24-, and 36-month all-cause mortality. In-sample performance was assessed using 5-fold cross-validation. Out-of-sample performance was assessed in an independent set of n = 528 MASLD patients. ResultsIn-sample model performance achieved AUROC curve 0.74–0.90 (95 % CI: 0.72–0.94), sensitivity 64 %-82 %, specificity 75 %–92 % and Positive Predictive Value (PPV) 94 %-98 %. Out-of-sample model validation had AUROC 0.70–0.86 (95 % CI: 0.67–0.90), sensitivity 69 %–70 %, specificity 96 %–97 % and PPV 75 %–77 %. Key predictive factors, identified using coefficients of determination, were age, presence of type 2 diabetes, and history of hospital admissions with length of stay >14 days. ConclusionsA TNN, applied to routinely-collected longitudinal electronic health records, achieved good performance in prediction of 12-, 24-, and 36-month all-cause mortality in patients with MASLD. Extrapolation of our technique to population-level data will enable scalable and accurate risk stratification to identify people most likely to benefit from anticipatory health care and personalized interventions.

Development of a nomogram for predicting perioperative blood transfusions in major hepatobiliary and colorectal surgeries: a retrospective study

Background: Major hepatobiliary and colorectal surgeries are associated with a risk of blood transfusions. However, risk assessment tools for predicting blood transfusions have not been studied extensively among patients undergoing these types of surgeries. We evaluated the risk factors for perioperative blood transfusions in patients who underwent major hepatobiliary and colorectal surgeries and subsequently created a nomogram. Methods: Medical records of patients who underwent elective major hepatobiliary and colorectal surgeries in a single university hospital in Malaysia from 2015 to 2020 were retrospectively reviewed. A nomogram to predict transfusions risk was developed, and its discriminatory ability was tested using the area under the receiver operating characteristic (ROC) curve. Results: Data from 293 patients (61.1% male) with an average age of 59.7 years (± SD 14.51) were analysed. The prevalence of anaemia was 61.1%. A total of 127 patients (43.3%) received at least 1 unit of packed red cells transfusions. On multivariable analysis, gender (odds ratio [OR 1.646), preoperative haemoglobin of 8.0 g/dl or less (OR 0.777), Charlson Comorbidity Index score (OR 1.14) and procedure type (versus colonic surgery, major hepatectomy, OR 6.094; other pancreatomy, OR 1.487; Whipple’s procedure, OR 9.667; and anterior resection, OR 3.569) were associated with a significantly higher risk of transfusions. All 4 of these factors were included in the nomogram. The nomogram’s discrimination and calibration results showed an AUROC curve of 0.754. Conclusion: The nomogram effectively predicted the need for blood transfusions in major colorectal and hepatobiliary surgeries in our patients.

HES V2.0 outperforms GALAD for detection of HCC: A phase 3 biomarker study in the United States.

The original hepatocellular carcinoma early detection screening (HES) score, which combines alpha-fetoprotein (AFP) with age, alanine aminotransferase, and platelets, has better performance than AFP alone for early HCC detection. We have developed HES V2.0 by adding AFP-L3 and des-gamma-carboxy prothrombin to the score and compared its performance to GALAD and ASAP scores among patients with cirrhosis. We conducted a prospective-specimen collection, retrospective-blinded-evaluation phase 3 biomarker cohort study in patients with cirrhosis enrolled in imaging and AFP surveillance. True-positive rate (TPR)/sensitivity and false-positive rate for any or early HCC were calculated for GALAD, ASAP, and HES V2.0 scores within 6, 12, and 24 months of HCC diagnosis. We calculated the AUROC curve and estimated TPR based on an optimal threshold at a fixed false-positive rate of 10%. We analyzed 2331 patients, of whom 125 developed HCC (71% in the early stages). For any HCC, HES V2.0 had higher TPR than GALAD overall (+7.2%), at 6 months (+3.6%), at 12 months (+7.2%), and 24 months (+13.0%) before HCC diagnosis. HES V2.0 had higher TPR than ASAP for all time points (+5.9% to +12.0%). For early HCC, HES V2.0 had higher sensitivity/TPR than GALAD overall (+6.7%), at 12 months (+6.3%), and 24 months (+14.6%) but not at 6 months (+0.0%) and higher than ASAP for all time points (+13.4% to +18.0%). In a prospective cohort study, HES V2.0 had a significantly higher performance for identifying new HCC, including early stage, than GALAD or ASAP.

Feasibility study of ResNet-50 in the distinction of intraoral neural tumors using histopathological images.

Neural tumors are difficult to distinguish based solely on cellularity and often require immunohistochemical staining to aid in identifying the cell lineage. This article investigates the potential of a Convolutional Neural Network for the histopathological classification of the three most prevalent benign neural tumor types: neurofibroma, perineurioma, and schwannoma. A model was developed, trained, and evaluated for classification using the ResNet-50 architecture, with a database of 30 whole-slide images stained in hematoxylin and eosin (106, 782 patches were generated from and divided among the training, validation, and testing subsets, with strategies to avoid data leakage). The model achieved an accuracy of 70% (64% normalized), and showed satisfactory results for differentiating two of the three classes, reaching approximately 97% and 77% as true positives for neurofibroma and schwannoma classes, respectively, and only 7% for perineurioma class. The AUROC curves for neurofibroma and schwannoma classes was 0.83%, and 0.74% for perineurioma. However, the specificity rate for the perineurioma class was greater (83%) than in the other two classes (neurofibroma with 61%, and schwannoma with 60%). This investigation demonstrated significant potential for proficient performance with a limitation regarding the perineurioma class (the limited feature variability observed contributed to a lower performance).

Application of the Rox Index as a Predictor of Respiratory Failure in Pediatric Patients Receiving High-Flow Oxygen Therapy Support at the Roberto Gilbert Elizalde Hospital

Background: High-flow oxygen therapy is an ideal treatment for managing respiratory failure. It avoids orotracheal intubation; however, some patients are refractory to this treatment. The ROX Index aims to detect early failure in high-flow oxygen therapy; nevertheless, its use has been poorly studied in pediatrics. Objective: To evaluate the applicability of the ROX Index as a tool to determine the failure of high-flow oxygen therapy in pediatric patients admitted to the Roberto Gilbert Hospital. Methodology: Observational, prospective, longitudinal cohort study with a descriptive approach including 209 patients between 1 month and 17 years old who required high-flow oxygen therapy within the first 24 hours. ROX Index measurements were taken at 2, 4, 6, 12, and 24 hours with follow-up to determine the outcome. Result: A total of 209 patients meeting the inclusion criteria were obtained; 27 (12.9%) required endotracheal intubation (ETI), and 182 (87%) received high-flow oxygen therapy (HFNC). A higher frequency of younger infants was observed among patients who required ETI, with a median age of 12 months. A significant difference in the ROX index was found from two hours onwards with cutoff points on the AUROC curve above 0.5. Conclusion: It was determined that the ROX Index can be used as a complementary tool to strengthen decision-making in pediatrics.

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.

The Usefulness of Vitamin K-Dependent Proteins in the Diagnosis of Colorectal Carcinoma.

Colorectal cancer (CRC) is one of the most common neoplasms in developed countries, with increasing incidence and mortality, even in young people. A variety of serum markers have been associated with CRC (CEA, CA 19-9), but neither should be used as a screening tool for the diagnosis or evolution staging of CRC. The sensitivity and specificity of these markers are not as good as is required, so new ones need to be found. Matrix Gla protein and PIVKA II are involved in carcinogenesis, but few studies have evaluated their usefulness in predicting the presence and severity of CRC. Two hundred patients were divided into three groups: 80 patients were included in the control group; 80 with CRC and without hepatic metastasis were included in Group 1; 40 patients with CRC and hepatic metastasis were included in Group 2. Vitamin K-dependent proteins (VKDPs) levels in plasma were determined. Patients with CRC without methastasis (Group 1) and CRC patients with methastasis (Group 2) presented significantly higher values of CEA, CA 19-9, PIVKA II (310.05 ± 38.22 vs. 430.13 ± 122.13 vs. 20.23 ± 10.90), and ucMGP (14,300.00 ± 2387.02 vs. 13,410.52 ± 2243.16 vs. 1780.31 ± 864.70) compared to control group (Group 0). Interestingly, Group 1 presented the greatest PIVKA II values. Out of all the markers, significant differences between the histological subgroups were found only for ucMGP, but only in non-metastatic CRC. Studying the discrimination capacity between the patients with CRC vs. those without, no significant differences were found between the classical tumor markers and the VKDP AUROC curves (PIVKA II and ucMGP AUROCs = 1). For the metastatic stage, the sensitivity and specificity of the VKDPs were lower in comparison with those of CA 19-9 and CEA, respectively (PIVKA II AUROC = 0.789, ucMGP AUROC = 0.608). The serum levels of these VKDPs are significantly altered in patients with colorectal carcinoma; it is possible to find additional value of these in the early stages of the disease.

Use of non-invasive scales for detecting esophageal varices in paediatric patients with portal vein thrombosis

IntroductionPortal vein thrombosis (PVT) is the most frequent cause of portal hypertension in paediatric population. Baveno VI Consensus considers endoscopic variceal ligation (EVL) as the second therapeutic option after Meso-Rex bypass (surgical shunt). AimAnalyse the diagnostic profitability of non-invasive scales in order to predict the risk of esophageal varices (OV) in children with PVT. Materials and methodsDescriptive retrospective study where every upper gastrointestinal endoscopy (UGE) carried on patients <15 years old with non-cirrhotic PVT were included. There were divided according to the presence of OV and sex, cause, age, previous gastrointestinal bleeding or treatments, results of UGE and scales (Clinical Prediction Rule-CPR), Varices Prediction Rule-VPR), King’s Variceal Prediction Score-K-VaPS) and Platelet count / Spleen diameter Ratio-PSR). Qualitative variables were expressed as absolute frequency and percentage, and quantitative variables as median and intercuartilic range. U Mann–Whitney and Hanley–McNeil tests were used for comparisons. Results45 UGE were analysed. 80% (n = 36) presented OV: median of 3(2–3) and 33,3%(n = 12) required EVL. Statistical differences were demonstrated between both groups: CPR (142,39 (132,22–166,53) vs. 122,75 (115,24–133,15); p = 0,003), VPR (9,91 (9,36 – 11,75) vs. 5,6 (3,34–8,39) p = 0,001), K-VaPS (117,86 (99,66–126,58) vs. 99,64 (94,88–110,18) p = 0,019), PSR (2384,62 (1902,22–3201,63) vs. 1252,5 (579,6–2144,42) p = 0,05), with and area under the curve AUROC > 75%, without statistical differences between scales. ConclusionsIn paediatric patients with non-cirrotic PVT non-invasive scales can be used as a tool to predict the presence of OV and raise the indication of UGE.

Synthetic minority over-sampling technique-enhanced machine learning models for predicting recurrence of postoperative chronic subdural hematoma.

Chronic subdural hematoma (CSDH) is a neurological condition with high recurrence rates, primarily observed in the elderly population. Although several risk factors have been identified, predicting CSDH recurrence remains a challenge. Given the potential of machine learning (ML) to extract meaningful insights from complex data sets, our study aims to develop and validate ML models capable of accurately predicting postoperative CSDH recurrence. Data from 447 CSDH patients treated with consecutive burr-hole irrigations at Wenzhou Medical University's First Affiliated Hospital (December 2014-April 2019) were studied. 312 patients formed the development cohort, while 135 comprised the test cohort. The Least Absolute Shrinkage and Selection Operator (LASSO) method was employed to select crucial features associated with recurrence. Eight machine learning algorithms were used to construct prediction models for hematoma recurrence, using demographic, laboratory, and radiological features. The Border-line Synthetic Minority Over-sampling Technique (SMOTE) was applied to address data imbalance, and Shapley Additive Explanation (SHAP) analysis was utilized to improve model visualization and interpretability. Model performance was assessed using metrics such as AUROC, sensitivity, specificity, F1 score, calibration plots, and decision curve analysis (DCA). Our optimized ML models exhibited prediction accuracies ranging from 61.0% to 86.2% for hematoma recurrence in the validation set. Notably, the Random Forest (RF) model surpassed other algorithms, achieving an accuracy of 86.2%. SHAP analysis confirmed these results, highlighting key clinical predictors for CSDH recurrence risk, including age, alanine aminotransferase level, fibrinogen level, thrombin time, and maximum hematoma diameter. The RF model yielded an accuracy of 92.6% with an AUC value of 0.834 in the test dataset. Our findings underscore the efficacy of machine learning algorithms, notably the integration of the RF model with SMOTE, in forecasting the recurrence of postoperative chronic subdural hematoma. Leveraging the RF model, we devised an online calculator that may serve as a pivotal instrument in tailoring therapeutic strategies and implementing timely preventive interventions for high-risk patients.

Evaluation of models to predict prognosis in patients with advanced hepatocellular carcinoma treated with TACE combined with apatinib.

The HAP, Six-and-Twelve, Up to Seven, and ALBI scores have been substantiated as reliable prognostic markers in patients presenting with intermediate and advanced hepatocellular carcinoma (HCC) undergoing transarterial chemoembolization (TACE) treatment. Given this premise, our research aims to assess the predictive efficacy of these models in patients with intermediate and advanced HCC receiving a combination of TACE and Apatinib. Additionally, we have conducted a meticulous comparative analysis of these four scoring systems to discern their respective predictive capacities and efficacies in combined therapy. Performing a retrospective analysis on the clinical data from 200 patients with intermediate and advanced HCC, we studied those who received TACE combined with Apatinib at the First Affiliated Hospital of the University of Science and Technology of China between June 2018 and December 2022. To identify the factors affecting survival, the study performed univariate and multivariate Cox regression analyses, with calculations of four different scores: HAP, Six-and-Twelve, Up to Seven, and ALBI. Lastly, Harrell's C-index was employed to compare the prognostic abilities of these scores. Cox proportional hazards model results revealed that the ALBI score, presence of portal vein tumor thrombus (PVTT, )and tumor size are independent determinants of prognostic survival. The Kaplan-Meier analyses showed significant differences in survival rates among patients classified by the HAP, Six-and-Twelve, Up to Seven, and ALBI scoring methods. Of the evaluated systems, the HAP scoring demonstrated greater prognostic precision, with a Harrell's C-index of 0.742, surpassing the alternative models (P < 0.05). In addition, an analysis of the area under the AU-ROC curve confirms the remarkable superiority of the HAP score in predicting short-term survival outcomes. Our study confirms the predictive value of HAP, Six-and-Twelve, Up to Seven, and ALBI scores in intermediate to advanced Hepatocellular Carcinoma (HCC) patients receiving combined Transarterial Chemoembolization (TACE) and Apatinib therapy. Notably, the HAP model excels in predicting outcomes for this specific HCC subgroup.

Prediction of complete response after neoadjuvant chemotherapy for invasive breast cancers: The utility of shear wave elastography and superb microvascular imaging in pretreatment breast ultrasound

PurposeTo investigate whether multiparametric parameters of pretreatment breast ultrasound (US) and clinicopathologic factors are associated with pathologic complete response (pCR) after neoadjuvant chemotherapy (NAC) for breast cancer. MethodsBetween November 2018 and September 2022, 88 patients who underwent NAC and subsequent surgery were included in this study (median age, 55 years; interquartile range [IQR], 45, 59.3). Multiparametric breast US including grayscale, shear wave elastography (SWE) and superb microvascular imaging (SMI) of pathologically proven invasive breast cancers were retrospectively reviewed. Clinicopathological and multiparametric parameters of breast US, including size, SWEmax, SWEratio and vascular index on SMI (SMIVI) were compared between the groups. Univariate and multivariate logistic regression analyses were performed to determine factors predicting pCR after NAC. AUROC curve analysis was performed to determine the predictors’ optimal cut-off values and diagnostic performance. ResultsThe pCR group (n = 24) showed a significantly smaller tumor size, lower SWEmax, higher Ki-67 index, higher hormone receptor negativity and negative axillary lymph node metastasis compared to the non-pCR group (n = 64). Multivariate regression analysis showed that SWEmax (adjusted odds ratio[aOR] = 0.956, 95 % confidence interval [CI] = 0.919–0.994, P = 0.025) and Ki-67 index (aOR = 1.083, 95 % CI = 1.012–1.159, P = 0.021) were independently associated with pathologically complete response. The optimal cut-off values for predicting pCR were 27.5 % for Ki-67 with an AUC of 0.743 and 134.8 kPa for SWEmax with an AUC of 0.779. A combination model including clinical factors and SWEmax showed the best diagnostic performance with an AUC of 0.876. ConclusionA higher Ki-67 index and lower SWEmax measured on pretreatment breast US were independently associated with pCR in invasive breast cancer after NAC.