Area Under ROC Curve Research Articles

Predictive value of whole brain perfusion on admission for shunt-dependent hydrocephalus fllowing aneurysmal subarachnoid hemorrhage

Objective: To explore the value of predicting shunt-dependent hydrocephalus (SDHC) in patients with aneurysmal subarachnoid hemorrhage (aSAH) based on whole brain CT perfusion(CTP) and clinical data within 24 hours at admission. Methods: The clinical and imaging data of aSAH patients who received interventional embolization in our hospital were retrospectively collected from March 2018 to August 2022. All patients underwent one-stop whole brain CT examination within 24 hours after symptom onset, and the qualitative and quantitative CTP parameters were obtained after post-processing. Follow-up was conducted once every 2 months by consulting electronic medical records or by telephone for 6 months. According to whether SDHC occurred or not, the patients were divided into SDHC group and non-SDHC group. The differences between the two groups were compared. Logistic regression model was used to analyze and determine the predictive factors of SDHC, and the SDHC predictive model was established. The effectiveness of the predictive model was evaluated by drawing the receiver operating characteristic (ROC) curve of the subjects. Results: A total of 414 patients were included, including 132 males and 282 females, aged (59±11) years. 17.6%(73/414) patients had SDHC. There were significant differences in the occurrence of acute hydrocephalus, the World Neurosurgical League Scale (WFNS), the Hunt-Hess scale, the modified Fisher score (mFS), and the qualitative and quantitative parameters of CTP between the two groups (both P<0.001). Multivariate logistic regression analysis showed that acute hydrocephalus (OR=8.621, 95%CI: 4.237-17.542),old age (OR=1.107, 95%CI: 1.068-1.148), high mFS and high Hunt-Hess classification (OR=3.740, 95%CI: 1.352-10.342) were the risk factors of SDHC in aSAH patients, and high mean cerebral blood flow (mCBF) (OR=0.931, 95%CI: 0.885-0.980) was a protective factor of SDHC.The area under ROC curve (AUC) of the prediction model constructed by these five variables was 0.923(95%CI: 0.89-0.95), with 84.5% sensitivity and 87.7% specificity. Conclusion: The mCBF and acute hydrocephalus, age, mFS and Hunt-Hess classification within 24 hours at admission can be used to predict SDHC for aSAH patients.

Development and validation a nomogram prediction model for early diagnosis of bloodstream infections in the intensive care unit.

This study aims to develop and validate a nomogram for predicting the risk of bloodstream infections (BSI) in critically ill patients based on their admission status to the Intensive Care Unit (ICU). Patients' data were extracted from the Medical Information Mart for Intensive Care-IV (MIMIC-IV) database (training set), the Beijing Friendship Hospital (BFH) database (validation set) and the eICU Collaborative Research Database (eICU-CRD) (validation set). Univariate logistic regression analyses were used to analyze the influencing factors, and lasso regression was used to select the predictive factors. Model performance was assessed using area under receiver operating characteristic curve (AUROC) and Presented as a Nomogram. Various aspects of the established predictive nomogram were evaluated, including discrimination, calibration, and clinical utility. The model dataset consisted of 14930 patients (1444 BSI patients) from the MIMIC-IV database, divided into the training and internal validation datasets in a 7:3 ratio. The eICU dataset included 2100 patients (100 with BSI) as the eICU validation dataset, and the BFH dataset included 419 patients (21 with BSI) as the BFH validation dataset. The nomogram was constructed based on Glasgow Coma Scale (GCS), sepsis related organ failure assessment (SOFA) score, temperature, heart rate, respiratory rate, white blood cell (WBC), red width of distribution (RDW), renal replacement therapy and presence of liver disease on their admission status to the ICU. The AUROCs were 0.83 (CI 95%:0.81-0.84) in the training dataset, 0.88 (CI 95%:0.88-0.96) in the BFH validation dataset, and 0.75 (95%CI 0.70-0.79) in the eICU validation dataset. The clinical effect curve and decision curve showed that most areas of the decision curve of this model were greater than 0, indicating that this model has a certain clinical effectiveness. The nomogram developed in this study provides a valuable tool for clinicians and nurses to assess individual risk, enabling them to identify patients at a high risk of bloodstream infections in the ICU.

Amide proton transfer-weighted imaging and stretch-exponential model DWI based 18F-FDG PET/MRI for differentiation of benign and malignant solitary pulmonary lesions

ObjectivesTo differentiate benign and malignant solitary pulmonary lesions (SPLs) by amide proton transfer-weighted imaging (APTWI), mono-exponential model DWI (MEM-DWI), stretched exponential model DWI (SEM-DWI), and 18F-FDG PET-derived parameters.MethodsA total of 120 SPLs patients underwent chest 18F-FDG PET/MRI were enrolled, including 84 in the training set (28 benign and 56 malignant) and 36 in the test set (13 benign and 23 malignant). MTRasym(3.5 ppm), ADC, DDC, α, SUVmax, MTV, and TLG were compared. The area under receiver-operator characteristic curve (AUC) was used to assess diagnostic efficacy. The Logistic regression analysis was used to identify independent predictors and establish prediction model.ResultsSUVmax, MTV, TLG, α, and MTRasym(3.5 ppm) values were significantly lower and ADC, DDC values were significantly higher in benign SPLs than malignant SPLs (all P < 0.01). SUVmax, ADC, and MTRasym(3.5 ppm) were independent predictors. Within the training set, the prediction model based on these independent predictors demonstrated optimal diagnostic efficacy (AUC, 0.976; sensitivity, 94.64%; specificity, 92.86%), surpassing any single parameter with statistical significance. Similarly, within the test set, the prediction model exhibited optimal diagnostic efficacy. The calibration curves and DCA revealed that the prediction model not only had good consistency but was also able to provide a significant benefit to the related patients, both in the training and test sets.ConclusionThe SUVmax, ADC, and MTRasym(3.5 ppm) were independent predictors for differentiation of benign and malignant SPLs, and the prediction model based on them had an optimal diagnostic efficacy.

Artificial intelligence based prediction model of in-hospital mortality among females with acute coronary syndrome: for the Jerusalem Platelets Thrombosis and Intervention in Cardiology (JUPITER-12) Study Group.

Despite ongoing efforts to minimize sex bias in diagnosis and treatment of acute coronary syndrome (ACS), data still shows outcomes differences between sexes including higher risk of all-cause mortality rate among females. Hence, the aim of the current study was to examine sex differences in ACS in-hospital mortality, and to implement artificial intelligence (AI) models for prediction of in-hospital mortality among females with ACS. All ACS patients admitted to a tertiary care center intensive cardiac care unit (ICCU) between July 2019 and July 2023 were prospectively enrolled. The primary outcome was in-hospital mortality. Three prediction algorithms, including gradient boosting classifier (GBC) random forest classifier (RFC), and logistic regression (LR) were used to develop and validate prediction models for in-hospital mortality among females with ACS, using only available features at presentation. A total of 2,346 ACS patients with a median age of 64 (IQR: 56-74) were included. Of them, 453 (19.3%) were female. Female patients had higher prevalence of NSTEMI (49.2% vs. 39.8%, p < 0.001), less urgent PCI (<2 h) rates (40.2% vs. 50.6%, p < 0.001), and more complications during admission (17.7% vs. 12.3%, p = 0.01). In-hospital mortality occurred in 58 (2.5%) patients [21/453 (5%) females vs. 37/1,893 (2%) males, HR = 2.28, 95% CI: 1.33-3.91, p = 0.003]. GBC algorithm outscored the RFC and LR models, with area under receiver operating characteristic curve (AUROC) of 0.91 with proposed working point of 83.3% sensitivity and 82.4% specificity, and area under precision recall curve (AUPRC) of 0.92. Analysis of feature importance indicated that older age, STEMI, and inflammatory markers were the most important contributing variables. Mortality and complications rates among females with ACS are significantly higher than in males. Machine learning algorithms for prediction of ACS outcomes among females can be used to help mitigate sex bias.

Conditional generative adversarial network driven radiomic prediction of mutation status based on magnetic resonance imaging of breast cancer

BackgroundBreast Cancer (BC) is a highly heterogeneous and complex disease. Personalized treatment options require the integration of multi-omic data and consideration of phenotypic variability. Radiogenomics aims to merge medical images with genomic measurements but encounter challenges due to unpaired data consisting of imaging, genomic, or clinical outcome data. In this study, we propose the utilization of a well-trained conditional generative adversarial network (cGAN) to address the unpaired data issue in radiogenomic analysis of BC. The generated images will then be used to predict the mutations status of key driver genes and BC subtypes.MethodsWe integrated the paired MRI and multi-omic (mRNA gene expression, DNA methylation, and copy number variation) profiles of 61 BC patients from The Cancer Imaging Archive (TCIA) and The Cancer Genome Atlas (TCGA). To facilitate this integration, we employed a Bayesian Tensor Factorization approach to factorize the multi-omic data into 17 latent features. Subsequently, a cGAN model was trained based on the matched side-view patient MRIs and their corresponding latent features to predict MRIs for BC patients who lack MRIs. Model performance was evaluated by calculating the distance between real and generated images using the Fréchet Inception Distance (FID) metric. BC subtype and mutation status of driver genes were obtained from the cBioPortal platform, where 3 genes were selected based on the number of mutated patients. A convolutional neural network (CNN) was constructed and trained using the generated MRIs for mutation status prediction. Receiver operating characteristic area under curve (ROC-AUC) and precision-recall area under curve (PR-AUC) were used to evaluate the performance of the CNN models for mutation status prediction. Precision, recall and F1 score were used to evaluate the performance of the CNN model in subtype classification.ResultsThe FID of the images from the well-trained cGAN model based on the test set is 1.31. The CNN for TP53, PIK3CA, and CDH1 mutation prediction yielded ROC-AUC values 0.9508, 0.7515, and 0.8136 and PR-AUC are 0.9009, 0.7184, and 0.5007, respectively for the three genes. Multi-class subtype prediction achieved precision, recall and F1 scores of 0.8444, 0.8435 and 0.8336 respectively. The source code and related data implemented the algorithms can be found in the project GitHub at https://github.com/mattthuang/BC_RadiogenomicGAN.ConclusionOur study establishes cGAN as a viable tool for generating synthetic BC MRIs for mutation status prediction and subtype classification to better characterize the heterogeneity of BC in patients. The synthetic images also have the potential to significantly augment existing MRI data and circumvent issues surrounding data sharing and patient privacy for future BC machine learning studies.

Monocyte-to-Hdl-Cholesterol Ratio as a Prognostic Marker in Covid-19

Background: There is an urgent need for mortality predictors for COVID-19 so that clinicians can diagnose severe cases and triage them as soon as possible. Many studies have suggested using hematologic markers to predict mortality and severity of COVID-19 disease. This study investigates the use of monocyte-to-high density lipoprotein cholesterol ratio (MHR) as a predictive marker for COVID-19 severity and mortality. Methods: This retrospective cross-sectional study was performed on 81 PCR-confirmed COVID-19 patients between 25 March 2020 to 26 June 2020. Patients were classified into two presentation categories: the non-severe group (n=37) and the severe group (n=44). Patients in the severe group were also divided into two subgroups: severe survivors (n=14) and severe non-survivors (n=30). In the receiver operating characteristic (ROC) analysis, optimal cut-off values of the monocyte count, high-density lipoprotein cholesterol (HDL-C), and MHR were calculated for the differentiation of severe and non-severe COVID-19 patients, as well as survivors and non-survivors. Results: A total of 81 patients, 29 (35.8%) males and 52 (64.2) females, with a median age of 71 (IQR 63-81) years. Both HDL-C and MHR showed a reasonable ability to distinguish severe disease from non-severe disease, while MHR had a higher area under curve (AUC) than HDL-C (0.799, 95%CI 0.704-0.894, p 0.001 vs 0.734, 95% Cl 0.626-0.843, p 0.001). Only MHR could distinguish survivors from non-survivors with an ROC AUC of 0.735 (95%Cl 0.619-0.850). The optimal cut-off values of MHR for predicting severe disease were 0.0061 (sensitivity: 66% and specificity: 66%) and 0.0066 (sensitivity: 70% and specificity: 62%) for predicting mortality. The optimal cut-off value of MHR for predicting severe disease was 0.0061 (sensitivity: 66% and specificity: 66%), and it was 0.0066 for predicting mortality among patients with severe disease (sensitivity: 70% and specificity: 62%). Conclusion: Our results showed that MHR was observed to be able to distinguish severe COVID-19 patients from non-severe patients as well as survivors from non-survivors.

Unraveling atherosclerotic cardiovascular disease risk factors through conditional probability analysis with Bayesian networks: insights from the AZAR cohort study

This study aimed at modelling the underlying predictor of ASCVD through the Bayesian network (BN). Data for the AZAR Cohort Study, which evaluated 500 healthcare providers in Iran, was collected through examinations, and blood samples. Two BNs were used to explore a suitable causal model for analysing the underlying predictor of ASCVD; Bayesian search through an algorithmic approach and knowledge-based BNs. Results showed significant differences in ASCVD risk factors across background variables’ levels. The diagnostic indices showed better performance for the knowledge-based BN (Area under ROC curve (AUC) = 0.78, Accuracy = 76.6, Sensitivity = 62.5, Negative predictive value (NPV) = 96.0, Negative Likelihood Ratio (LR−) = 0.48) compared to Bayesian search (AUC = 0.76, Accuracy = 72.4, Sensitivity = 17.5, NPV = 93.2, LR− = 0.83). In addition, we decided on knowledge-based BN because of the interpretability of the relationships. Based on this BN, being male (conditional probability = 63.7), age over 45 (36.3), overweight (51.5), Mets (23.8), diabetes (8.3), smoking (10.6), hypertension (12.1), high T-C (28.5), high LDL-C (23.9), FBS (12.1), and TG (25.9) levels were associated with higher ASCVD risk. Low and normal HDL-C levels also had higher ASCVD risk (35.3 and 37.4), while high HDL-C levels had lower risk (27.3). In conclusion, BN demonstrated that ASCVD was significantly associated with certain risk factors including being older and overweight male, having a history of Mets, diabetes, hypertension, having high levels of T-C, LDL-C, FBS, and TG, but Low and normal HDL-C and being a smoker. The study may provide valuable insights for developing effective prevention strategies for ASCVD in Iran.

Artificial Intelligence to Differentiate Pediatric Pseudopapilledema and True Papilledema on Fundus Photographs

PurposeTo develop and test an artificial intelligence (AI) model to aid in differentiating pediatric pseudopapilledema from true papilledema on fundus photographs. DesignMulti-center retrospective study Subjects851 fundus photographs from 235 children (age <18 years) with pseudopapilledema and true papilledema MethodsFour pediatric neuro-ophthalmologists at four different institutions contributed fundus photographs of children with confirmed diagnoses of papilledema or pseudopapilledema. An AI model to classify fundus photographs as papilledema or pseudopapilledema was developed using a DenseNet backbone and a tri-branch convolutional neural network. We performed 10-fold cross-validation and separately analyzed an external test set. The AI model’s performance was compared to two masked human expert pediatric neuro-ophthalmologists, who performed the same classification task. Main Outcome MeasuresAccuracy, sensitivity, and specificity of the AI model compared to human experts. ResultsThe area under receiver operating curve (AUC) of the AI model was 0.77 for the cross-validation set and 0.81 for the external test set. The accuracy of the AI model was 70.0% for the cross-validation set and 73.9% for the external test set. The sensitivity of the AI model was 73.4% for the cross-validation set and 90.4% for the external test set. The AI model’s accuracy was significantly higher than human experts on the cross validation set (p<0.002), and the model’s sensitivity was significantly higher on the external test set (p<0.0002). The specificity of the AI model and human experts was similar (56.4-67.3%). Moreover, the AI model was significantly more sensitive at detecting mild papilledema than human experts, whereas AI and humans performed similarly on photographs of moderate-to-severe papilledema. On review of the external test set, only one child (with nearly resolved pseudotumor cerebri) had both eyes with papilledema incorrectly classified as pseudopapilledema. ConclusionsWhen classifying fundus photographs of pediatric papilledema and pseudopapilledema, our AI model achieved over 90% sensitivity at detecting papilledema, superior to human experts. Due to the high sensitivity and low false negative rate, AI may be useful to triage children with suspected papilledema requiring work-up to evaluate for serious underlying neurologic conditions.

Blood lactate after pre-hospital blood transfusion for major trauma by helicopter emergency medical services.

The appropriate use of blood components is essential for ethical use of a precious, donated product. The aim of this study was to report in-hospital red blood cell (RBC) transfusion after pre-hospital transfusion by helicopter emergency medical service paramedics. A secondary aim was to assess the potential for venous blood lactate to predict ongoing transfusion. All patients who received RBC in air ambulance were transported to a single adult major trauma centre, had venous blood lactate measured on arrival and did not die before ability to transfuse RBC were included. The association of venous blood lactate with ongoing RBC transfusion was assessed using multi-variable logistic regression analysis and reported using adjusted odds ratios (aOR). The discriminative ability of venous blood lactate was assessed using area under receiver operating characteristics curve (AUROC). From 1 January 2016 to 15 May 2019, there were 165 eligible patients, and 128 patients were included. In-hospital transfusion occurred in 97 (75.8%) of patients. Blood lactate was associated with ongoing RBC transfusion (aOR: 2.00; 95% confidence interval [CI]: 1.36-2.94). Blood lactate provided acceptable discriminative ability for ongoing transfusion (AUROC: 0.78; 95% CI: 0.70-0.86). After excluding patients with early deaths, a quarter of those who had prehospital RBC transfusion had no further transfusion in hospital. Venous blood lactate appears to provide value in identifying such patients. Lactate levels after pre-hospital transfusion could be used as a biomarker for transfusion requirement after trauma.

A MaxEnt Model of Citrus Black Fly Aleurocanthus woglumi Ashby (Hemiptera: Aleyrodidae) under Different Climate Change Scenarios.

The citrus blackfly (CBF), Aleurocanthus woglumi Ashby, is an exotic pest native to Southeast Asia that has spread rapidly to the world's main centers of citrus production, having been recently introduced to Brazil. In this study, a maximum entropy niche model (MaxEnt) was used to predict the potential worldwide distribution of CBF under current and future climate change scenarios for 2030 and 2050. These future scenarios came from the Coupled Model Intercomparison Project Phase 6 (CMIP6), SSP1-2.6, and SSP5-8.5. The MaxEnt model predicted the potential distribution of CBF with area under receiver operator curve (AUC) values of 0.953 and 0.930 in the initial and final models, respectively. The average temperature of the coldest quarter months, precipitation of the rainiest month, isothermality, and precipitation of the driest month were the strongest predictors of CBF distribution, with contributions of 36.7%, 14.7%, 13.2%, and 10.2%, respectively. The model based on the current time conditions predicted that suitable areas for the potential occurrence of CBF, including countries such as Brazil, China, the European Union, the USA, Egypt, Turkey, and Morocco, are located in tropical and subtropical regions. Models from SSP1-2.6 (2030 and 2050) and SSP5-8.5 (2030) predicted that suitable habitats for CBF are increasing dramatically worldwide under future climate change scenarios, particularly in areas located in the southern US, southern Europe, North Africa, South China, and part of Australia. On the other hand, the SSP5-8.5 model of 2050 indicated a great retraction of the areas suitable for CBF located in the tropical region, with an emphasis on countries such as Brazil, Colombia, Venezuela, and India. In general, the CMIP6 models predicted greater risks of invasion and dissemination of CBF until 2030 and 2050 in the southern regions of the USA, European Union, and China, which are some of the world's largest orange producers. Knowledge of the current situation and future propagation paths of the pest serve as tools to improve the strategic government policies employed in CBF's regulation, commercialization, inspection, combat, and phytosanitary management.

Improvement of pulsars detection using dataset balancing methods and symbolic classification ensemble

Highly accurate detection of pulsars is mandatory. With the application of machine learning (ML) algorithms, the detection of pulsars can certainly be improved if the dataset is balanced. In this paper, the publicly available dataset (HTRU2) is highly imbalanced so various balancing methods were applied. The balanced dataset was used in genetic programming symbolic classifier (GPSC) to obtain symbolic expressions (SEs) that can detect pulsars with high classification accuracy. To find the optimal combination of GPSC hyperparameters the random hyperparameter search (RHS) method was developed and applied. The GPSC was trained using 5-fold cross-validation so after each training a total of 5 SEs were obtained. The best set of SEs are selected based on their classification performance and all of them are applied on the original dataset. The best classification accuracy (ACC), the area under receiver operating characteristic (AUC), precision, recall, and f1-score were achieved in the case of the dataset balanced with the AllKNN method i.e. all mean evaluation metric values are equal to 0.995. The ensemble consisted of 25 SEs that achieved the ACC=0.978, AUC=0.9452 , Precision=0.905, Recall=0.9963, and F1−Score=0.94877, on the original dataset.

A dynamic online nomogram for predicting the heterogeneity trajectories of frailty among elderly gastric cancer survivors

BackgroundFrailty is very common among older people with gastric cancer and seriously affects their prognosis. The development of frailty is continuous and dynamic, increasing the difficulty and burden of care. ObjectivesThe aims of this study were to delineate the developmental trajectory of frailty in older people with gastric cancer 1 year after surgery, identify heterogeneous frailty trajectories, and further explore their predictors to construct a nomogram for prediction. DesignWe conducted a prospective longitudinal observation study. Clinical evaluation and data collection were performed at discharge, and at 1, 3, 6, and 12 months. Setting and participantsThis study was conducted in a tertiary hospital and 381 gastric cancer patients over 60 years who underwent radical gastrectomy completed the 1-year follow-up. MethodsA growth mixture model (GMM) was used to delineate the frailty trajectories, and identify heterogeneous trajectories. A regression model was performed to determine their predictors and further construct a nomogram based on the predictors. Bootstrap with 1000 resamples was used for internal validation of nomogram, a receiver operating characteristic (ROC) curve to evaluate discrimination, calibration curves to evaluate calibration and decision curve analysis (DCA) to evaluate the clinical value. ResultsGMM identified three classes of frailty trajectories: “frailty improving”, “frailty persisting” and “frailty deteriorating”. The latter two were referred to as heterogeneous frailty trajectories. Regression analysis showed 8 independent predictors of heterogeneous frailty trajectories and a nomogram was constructed based on these predictors. The area under ROC curve (AUC) of the nomogram was 0.731 (95 % CI = 0.679–0.781), the calibration curves demonstrated that probabilities predicted by the nomogram agreed well with the actual observation with a mean absolute error of 0.025, and the DCA of nomogram indicated that the net benefits were higher than that of the other eight single factors. ConclusionsOlder gastric cancer patients have heterogeneous frailty trajectories of poor prognosis during one-year postoperative survival. Therefore, early assessment to predict the occurrence of heterogeneous frailty trajectories is essential to improve the outcomes of elderly gastric cancer patients. Scientific and effective frailty interventions should be further explored in the future to improve the prognosis of older gastric cancer patients. Contribution of the paper statementsThis study constructed a static and dynamic online nomogram with good discrimination and calibration, which can help to screen high-risk patients, implement preoperative risk stratification easily and promote the rational allocation of medical resources greatly. RegistrationClinicalTrials.gov (Number: NCT05982899). Tweetable abstractOur findings identified three frailty trajectories and constructed a nomogram to implement preoperative risk stratification and improve patient outcomes.

Diagnostic performance of coronary computed tomography angiography stenosis score for coronary stenosis

BackgroundCoronary computed tomography angiography stenosis score (CCTA-SS) is a proposed diagnosis score that considers the plaque characteristics, myocardial function, and the diameter reduction rate of the lesions. This study aimed to evaluate the diagnostic performance of the CCTA-SS in seeking coronary artery disease (CAD).MethodsThe 228 patients with suspected CAD who underwent CCTA and invasive coronary angiography (ICA) procedures were under examination. The diagnostic performance was evaluated with the receiver operating curve (ROC) for CCTA-SS in detecting CAD (defined as a diameter reduction of ≥ 50%) and severe CAD (defined as a diameter reduction of ≥ 70%).ResultsThe area under ROC (AUC) of CCTA-SS was 0.909 (95% CI: 0.864–0.943), which was significantly higher than that of CCTA (AUC: 0.826; 95% CI: 0.771–0.873; P = 0.0352) in diagnosing of CAD with a threshold of 50%. The optimal cutoff point of CCTA-SS was 51% with a sensitivity of 90.66%, specificity of 95.65%, positive predictive value of 98.80%, negative predictive value of 72.13%, and accuracy of 91.67%, whereas the optimal cutoff point of CCTA was 55%, and the corresponding values were 87.36%, 93.48%, 98.15%, 65.15%, and 88.60%, respectively. With a threshold of 70%, the performance of CCTA-SS with an AUC of 0.927 (95% CI: 0.885–0.957) was significantly higher than that of CCTA with an AUC of 0.521 (95% CI: 0.454–0.587) (P < 0.0001).ConclusionsCCTA-SS significantly improved the diagnostic accuracy of coronary stenosis, including CAD and severe CAD, compared with CCTA.

Machine Learning Identifies Metabolic Dysfunction-Associated Steatotic Liver Disease in Patients With Diabetes Mellitus.

The presence of metabolic dysfunction-associated steatotic liver disease (MASLD) in patients with diabetes mellitus (DM) is associated with a high risk of cardiovascular disease, but is often underdiagnosed. To develop machine learning (ML) models for risk assessment of MASLD occurrence in patients with DM. Feature selection determined the discriminative parameters, utilized to classify DM patients as those with and without MASLD. The performance of the multiple logistic regression model was quantified by sensitivity, specificity, and percentage of correctly classified patients, and receiver operating characteristic (ROC) curve analysis. Decision curve analysis (DCA) assessed the model's net benefit for alternative treatments. We studied 2000 patients with DM (mean age 58.85 ± 17.37 years; 48% women). Eight parameters: age, body mass index, type of DM, alanine aminotransferase, aspartate aminotransferase, platelet count, hyperuricaemia, and treatment with metformin were identified as discriminative. The experiments for 1735 patients show that 744/991 (75.08%) and 586/744 (78.76%) patients with/without MASLD were correctly identified (sensitivity/specificity: 0.75/0.79). The area under ROC (AUC) was 0.84 (95% CI, 0.82-0.86), while DCA showed a higher clinical utility of the model, ranging from 30% to 84% threshold probability. Results for 265 test patients confirm the model's generalizability (sensitivity/specificity: 0.80/0.74; AUC: 0.81 [95% CI, 0.76-0.87]), whereas unsupervised clustering identified high-risk patients. A ML approach demonstrated high performance in identifying MASLD in patients with DM. This approach may facilitate better risk stratification and cardiovascular risk prevention strategies for high-risk patients with DM at risk of MASLD.

Interpretable machine learning model to predict surgical difficulty in laparoscopic resection for rectal cancer.

Laparoscopic total mesorectal excision (LaTME) is standard surgical methods for rectal cancer, and LaTME operation is a challenging procedure. This study is intended to use machine learning to develop and validate prediction models for surgical difficulty of LaTME in patients with rectal cancer and compare these models' performance. We retrospectively collected the preoperative clinical and MRI pelvimetry parameter of rectal cancer patients who underwent laparoscopic total mesorectal resection from 2017 to 2022. The difficulty of LaTME was defined according to the scoring criteria reported by Escal. Patients were randomly divided into training group (80%) and test group (20%). We selected independent influencing features using the least absolute shrinkage and selection operator (LASSO) and multivariate logistic regression method. Adopt synthetic minority oversampling technique (SMOTE) to alleviate the class imbalance problem. Six machine learning model were developed: light gradient boosting machine (LGBM); categorical boosting (CatBoost); extreme gradient boost (XGBoost), logistic regression (LR); random forests (RF); multilayer perceptron (MLP). The area under receiver operating characteristic curve (AUROC), accuracy, sensitivity, specificity and F1 score were used to evaluate the performance of the model. The Shapley Additive Explanations (SHAP) analysis provided interpretation for the best machine learning model. Further decision curve analysis (DCA) was used to evaluate the clinical manifestations of the model. A total of 626 patients were included. LASSO regression analysis shows that tumor height, prognostic nutrition index (PNI), pelvic inlet, pelvic outlet, sacrococcygeal distance, mesorectal fat area and angle 5 (the angle between the apex of the sacral angle and the lower edge of the pubic bone) are the predictor variables of the machine learning model. In addition, the correlation heatmap shows that there is no significant correlation between these seven variables. When predicting the difficulty of LaTME surgery, the XGBoost model performed best among the six machine learning models (AUROC=0.855). Based on the decision curve analysis (DCA) results, the XGBoost model is also superior, and feature importance analysis shows that tumor height is the most important variable among the seven factors. This study developed an XGBoost model to predict the difficulty of LaTME surgery. This model can help clinicians quickly and accurately predict the difficulty of surgery and adopt individualized surgical methods.

Development of an evidence-based diagnostic algorithm for infection in patients with transcutaneous osseointegration following amputation.

Introduction: Transcutaneous osseointegration following amputation (TOFA) confers better mobility and quality of life for most patients versus socket prosthesis rehabilitation. Peri-TOFA infection remains the most frequent complication and lacks an evidence-based diagnostic algorithm. This study's objective was to investigate preoperative factors associated with positive intraoperative cultures among patients suspected of having peri-TOFA infection in order to create an evidence-based diagnostic algorithm. Methods: We conducted a retrospective study of 83 surgeries (70 patients) performed to manage suspected lower-extremity peri-TOFA infection at a specialty orthopedic practice and tertiary referral hospital in a major urban center. The diagnosis of infection was defined as positive intraoperative cultures. Preoperative patient history (fevers, subjective pain, increased drainage), physician examination findings (local cellulitis, purulent discharge, implant looseness), and laboratory data (white blood cell count, C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and external swab culture) were evaluated for association with subsequent positive intraoperative cultures using regression and area under receiver-operator curve (AUC) modeling. Results: Peri-implant limb pain (highly correlated with infection), ESR (highly correlated against infection), positive preoperative swab (moderately correlated with infection), gross implant motion (moderately correlated against infection), and erythema or cellulitis of the transcutaneous region (mildly correlated with infection) were variables included in the best AUC model, which achieved an 85 % positive predictive value. Other clinical findings and laboratory values (notably CRP and WBC) were non-predictive of infection. Conclusions: This seminal investigation to develop a preoperative diagnostic algorithm for peri-TOFA infection suggests that the clinical examination remains paramount. Further evaluation of a wider spectrum of clinical, laboratory, and imaging data, consistently and routinely collected with prospective data techniques in larger cohorts of patients, is necessary to create a robust predictive algorithm.

An inflammation-based model for identifying severe acute pancreatitis: a single-center retrospective study

Background and aimsSevere acute pancreatitis (SAP) is potentially lethal. Considering the role of inflammation in the progression of acute pancreatitis (AP), this study aims to develop a model based on inflammatory indexes for identifying the presence of SAP.MethodsOverall, 253 patients with AP who were consecutively admitted between July 2018 and November 2020 were screened, of whom 60 had SAP. Systemic immune-inflammation index (SII), neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR), neutrophil-to-platelet ratio (NPR), systemic inflammation response index (SIRI), platelet-to-albumin ratio (PAR), C-reactive protein-to-albumin ratio (CAR), C-reactive protein-to-lymphocyte ratio (CLR), and triglyceride glucose (TyG) index were calculated. Multivariate logistic regression analyses were performed to identify independent risk factors of SAP. Then, inflammation-based models were established. Receiver operating characteristics (ROC) curve analyses were performed. Area under ROC curve (AUROC) was calculated.ResultsDiabetes mellitus, fatty liver, high white blood cell count (WBC), C-reactive protein (CRP), red blood cell distribution width (RDW), procalcitonin (PCT), SII, NLR, NPR, CAR, CLR, and TyG index, and a low LMR were significantly associated with SAP. Considering the collinearity among these variables, 10 multivariate logistic regression analyses were separately performed. Finally, four independent inflammation-based models were established. Of them, the best one, which was calculated as follows: 1.204*fatty liver (yes = 1; no = 0) + 0.419*PCT + 0.005*CLR - 2.629, had an AUROC of 0.795 with a specificity of 73.4% and a sensitivity of 71.7%.ConclusionThe inflammation-based model consisting of fatty liver, PCT, and CLR has a good diagnostic performance for SAP.

Enhanced prediction of abnormal glucose tolerance using an extended non-invasive risk score incorporating routine renal biochemistry

IntroductionType 2 diabetes is preventable in subjects with impaired glucose tolerance based on 2-hour plasma glucose (2hPG) during 75 g oral glucose tolerance test (OGTT). We incorporated routine biochemistry to...

Development and validation of machine learning models to predict the need for haemostatic therapy in acute upper gastrointestinal bleeding.

Acute upper gastrointestinal bleeding (AUGIB) is a major cause of morbidity and mortality. This presentation however is not universally high risk as only 20-30% of bleeds require urgent haemostatic therapy. Nevertheless, the current standard of care is for all patients admitted to an inpatient bed to undergo endoscopy within 24 h for risk stratification which is invasive, costly and difficult to achieve in routine clinical practice. To develop novel non-endoscopic machine learning models for AUGIB to predict the need for haemostatic therapy by endoscopic, radiological or surgical intervention. A retrospective cohort study. We analysed data from patients admitted with AUGIB to hospitals from 2015 to 2020 (n = 970). Machine learning models were internally validated to predict the need for haemostatic therapy. The performance of the models was compared to the Glasgow-Blatchford score (GBS) using the area under receiver operating characteristic (AUROC) curves. The random forest classifier [AUROC 0.84 (0.80-0.87)] had the best performance and was superior to the GBS [AUROC 0.75 (0.72-0.78), p < 0.001] in predicting the need for haemostatic therapy in patients with AUGIB. A GBS cut-off of ⩾12 was associated with an accuracy, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of 0.74, 0.49, 0.81, 0.41 and 0.85, respectively. The Random Forrest model performed better with an accuracy, sensitivity, specificity, PPV and NPV of 0.82, 0.54, 0.90, 0.60 and 0.88, respectively. We developed and validated a machine learning algorithm with high accuracy and specificity in predicting the need for haemostatic therapy in AUGIB. This could be used to risk stratify high-risk patients to urgent endoscopy.

A convolutional neural network model using intraprostatic patterns of [F18]DCFPyL uptake in PSMA PET images to predict synchronous metastases.

46 Background: [F18]DCFPyL (PyL) is a PSMA targeted imaging agent that provides whole-body staging of prostate cancer. Image analysis of the primary tumor using deep learning algorithms might offer additional insight into disease biology, including co-existing metastatic disease. We developed convolutional neural network (CNN) models using inputs from the entire prostate on PyL images along with auto-segmented hot spots within the of primary prostate cancer tumor to predict synchronous metastases and compared the models against established models built from clinico-pathologic data. Methods: 91 Veterans with de novo prostate cancer were imaged with PyL PSMA PET/CT for initial staging (46% with metastatic disease). The PyL images of the prostate were analyzed using aPROMISE, which autosegments, localizes, and quantifies disease on PSMA PET. The segmentations of the prostate were used to map the PyL PET image of the prostate. Both the entire prostate, as well as aPROMISE defined hot spots were used as inputs for the CNN, where, according to attention map analysis, the hotspot information helps the network understand the location and extent of tumors. The CNN model architecture was based on SqueezeNet v2. The dataset was split into training, validation and test sets using stratified random sampling. The area under ROC curve (AUC) was computed to determine the performance of the model in predicting the presence of metastases and the test predictions were compared with ground truth (M1). Training was repeated 50 times and the best performing experiment was identified. Prediction scores from UCSF-CAPRA and UCLA PSMA risk calculator were used as comparators. Results: The best CNN model had an AUC of 0.89 for prediction of metastatic disease (median 0.72, ICR 0.64 and 0.8). For comparison, the AUC from the UCSF-CAPRA score and UCLA-PSMA risk calculator (any upstaging on PET), which rely on clinicopathologic information, had AUCs of 0.729 and 0.754 in this dataset, respectively. Conclusions: The CNN based model using PyL imaging demonstrates that synchronous metastases can be predicted from intraprostatic PyL uptake patterns alone with a predictive accuracy in this dataset that is at least comparable to published prediction models based on clinicopathologic features. This study raises the hypothesis that PyL CNN-based models could be developed to prognosticate metastatic progression.