External Validation Of Model Research Articles

Derivation and External Validation of the Ottawa Bloodstream Infection Model for Acutely Ill Adults.

Knowing the probability that patients have a bloodstream infection (BSI) could influence the ordering of blood cultures and interpretation of their preliminary results. Many previous BSI probability models have limited applicability and accuracy. This study used currently recommended modeling techniques and a large sample to derive and validate the Ottawa BSI Model. At a tertiary care teaching hospital, we retrieved a random sample of 4180 adults having blood cultures in our emergency department or during the initial 48h of the encounter. Variable selection was based on clinical experience and a systematic review of previous model performance. Model performance was measured in a temporal external validation group of 4680 patients. A total of 327 derivation patients had a BSI (8.0%). BSI risk increased with increased number of culture sets (2 sets: adjusted odds ratio [aOR] 1.52 [1.10-2.11]; 3 sets: 1.99 [0.86-4.58]); with indwelling catheter (aOR 2.07 [1.34-3.20); with increasing temperature, heart rate, and neutrophil-lymphocyte ratio; and with decreasing systolic blood pressure, platelet count, urea-creatinine ratio, and estimated glomerular filtration rate. In the temporal external validation group, model discrimination was good (c-statistic 0.71 [0.69-0.74]) and calibration was very good (integrated calibration index .016 [.010-.024]). Exclusion of validation patients with acute SARS-CoV-2 infection improved discrimination slightly (c-statistic 0.73 [0.69-0.76]). The Ottawa BSI Model uses commonly available data to return an expected BSI probability for acutely ill patients. However, it cannot exclude BSI and its complexity requires computational assistance to use.

Development and external validation of the multichannel deep learning model based on unenhanced CT for differentiating fat-poor angiomyolipoma from renal cell carcinoma: a two-center retrospective study.

There are undetectable levels of fat in fat-poor angiomyolipoma. Thus, it is often misdiagnosed as renal cell carcinoma. We aimed to develop and evaluate a multichannel deep learning model for differentiating fat-poor angiomyolipoma (fp-AML) from renal cell carcinoma (RCC). This two-center retrospective study included 320 patients from the First Affiliated Hospital of Sun Yat-Sen University (FAHSYSU) and 132 patients from the Sun Yat-Sen University Cancer Center (SYSUCC). Data from patients at FAHSYSU were divided into a development dataset (n = 267) and a hold-out dataset (n = 53). The development dataset was used to obtain the optimal combination of CT modality and input channel. The hold-out dataset and SYSUCC dataset were used for independent internal and external validation, respectively. In the development phase, models trained on unenhanced CT images performed significantly better than those trained on enhanced CT images based on the fivefold cross-validation. The best patient-level performance, with an average area under the receiver operating characteristic curve (AUC) of 0.951 ± 0.026 (mean ± SD), was achieved using the "unenhanced CT and 7-channel" model, which was finally selected as the optimal model. In the independent internal and external validation, AUCs of 0.966 (95% CI 0.919-1.000) and 0.898 (95% CI 0.824-0.972), respectively, were obtained using the optimal model. In addition, the performance of this model was better on large tumors (≥ 40mm) in both internal and external validation. The promising results suggest that our multichannel deep learning classifier based on unenhanced whole-tumor CT images is a highly useful tool for differentiating fp-AML from RCC.

External Validation of the IMPROD-MRI Volumetric Model to Predict the Utility of Systematic Biopsies at the Time of Targeted Biopsy.

Background: The aim of this study was to validate externally a nomogram that relies on MRI volumetric parameters and clinical data to determine the need for a standard biopsy in addition to a target biopsy for men with suspicious prostate MRI findings. Methods: We conducted a retrospective analysis of a prospectively maintained database of 469 biopsy-naïve men who underwent prostate biopsies. These biopsies were guided by pre-biopsy multiparametric Magnetic Resonance Imaging (mpMRI) and were performed at two different institutions. We included men with a PIRADSsv 2.1 score from 3 to 5. Each patient underwent both an MRI-ultrasound fusion biopsy of identified MRI-suspicious lesions and a systematic biopsy according to our protocol. The lesion volume percentage was determined as the proportion of cancer volume on MRI relative to the entire prostate volume. The study's outcomes were iPCa (Gleason Grade Group 1) and csPCa (Gleason Grade Group > 1). We evaluated the model's performance using AUC decision curve analyses and a systematic analysis of model-derived probability cut-offs in terms of the potential to avoid diagnosing iPCa and to accurately diagnose csPCa. Results: The nomogram includes age, PSA value, prostate volume, PIRADSsv 2.1 score, percentage of MRI-suspicious lesion volume, and lesion location. AUC was determined to be 0.73. By using various nomogram cut-off thresholds (ranging from 5% to 30%), it was observed that 19% to 58% of men could potentially avoid undergoing standard biopsies. In this scenario, the model might miss 0% to 10% of diagnosis of csPCa and could prevent identifying 6% to 31% of iPCa cases. These results are in line with findings from the multi-institutional external validation study based on the IMPROD trial (n = 122) and the MULTI-IMPROD trial (n = 262). According to DCA, the use of this nomogram led to an increased overall net clinical benefit when the threshold probability exceeded 10%. Conclusions: This study supports the potential value of a model relying on MRI volumetric measurements for selecting individuals with clinical suspicion of prostate cancer who would benefit from undergoing a standard biopsy in addition to a targeted biopsy.

Call to Improve the Quality of Prediction Tools for Intrahepatic Cholangiocarcinoma Resection: A Critical Appraisal, Systematic Review, and External Validation Study.

To conduct a systematic review, critical appraisal, and external validation of survival prediction tools for patients undergoing intrahepatic cholangiocarcinoma (iCCA) resection. Despite the development of several survival prediction tools in recent years for patients undergoing iCCA resections, there is a lack of critical appraisal and external validation of these models. We conducted a systematic review and critical appraisal of survival and recurrence prediction models for patients undergoing curative-intent iCCA resections. Studies were evaluated based on their model design, risk of bias, reporting, performance, and validation results. We identified the best model and externally validated it using our institution's data. This review included a total of 31 studies, consisting of 26 studies with original prediction tools and 5 studies that only conducted external validations. Among the 26, 54% of the studies conducted internal validations, 46% conducted external validations, and only 1 study scored a low risk of bias. Harrell's C-statistics ranged from 0.67 to 0.76 for internal validation and from 0.64 to 0.75 for external validation. Only 81% of the studies reported model calibration. Our external validation of the best model (Intrahepatic Cholangiocarcinoma [ICC]-Metroticket) estimated Harrell's and Uno's C-statistics of 0.67 (95% CI: 0.56-0.77) and Uno's time-dependent area under the receiver operating characteristic curve (AUC) of 0.71 (95% CI: 0.53-0.88), with a Brier score of 0.20 (95% CI: 0.15-0.26) and good calibration plots. Many prediction models have been published in recent years, but their quality remains poor, and minimal methodological quality improvement has been observed. The ICC-Metroticket was selected as the best model (Uno's time-dependent AUC of 0.71) for 5-year overall survival prediction in patients undergoing curative-intent iCCA resection.

Mechanistic Skin Modeling of Plasma Concentrations of Sunscreen Active Ingredients Following Facial Application

Sunscreen products constitute two distinct categories. Recreational sunscreens protect against high-intensity, episodic sun exposure, often applied over the entire body. In contrast, facial sunscreen products are designed for sub-erythemal, low-intensity daily sun exposure. Such different exposures necessitate distinctive product safety assessments. Building on earlier methods for predicting dermal disposition, a mechanistic model was developed to simulate plasma concentrations of seven organic sunscreen active ingredients: avobenzone, ensulizole, homosalate, octinoxate, octisalate, octocrylene, and oxybenzone, following facial application. In vitro permeation testing (IVPT) was performed with two different vehicles using a subset of the UV filters. These IVPT results, in addition to previously published IVPT data and published in vivo Maximal Usage Trial (MUsT) data for the UV filters, were used to train the mechanistic dermal model via a Bayesian Markov chain Monte Carlo (MCMC) method. An external validation of the trained model with real-world in vivo datasets demonstrated that the model's predicted UV filter plasma concentrations align well with experimental measurements and capture the observed inter-individual variability. Predictions of steady-state UV filter plasma concentrations under facial application scenarios at 5% concentration and at the maximal allowable concentrations were then generated by the trained model. Oxybenzone had the greatest predicted plasma concentration following facial application. Homosalate and octisalate predictions had high uncertainty associated with the absence of data. Several application scenarios pertaining to avobenzone, ensulizole, octocrylene and octinoxate were identified in which median plasma concentration levels were at 0.5 ng/ml or below when applied in the recreational or facial product. Model limitations include uncertainty in vehicle/water partitioning, formulation metamorphosis, and UV filter systemic clearance, all of which can be refined with additional data. For UV filters, limiting exposure to facial application reduces human safety concerns based on FDA established thresholds.

Development and external validation of a machine learning model for prediction of survival in undifferentiated pleomorphic sarcoma.

Machine learning (ML) algorithms to predict cancer survival have recently been reported for a number of sarcoma subtypes, but none have investigated undifferentiated pleomorphic sarcoma (UPS). ML is a powerful tool that has the potential to better prognosticate UPS. The Surveillance, Epidemiology, and End Results (SEER) database was queried for cases of histologically confirmed undifferentiated pleomorphic sarcoma (UPS) (n = 665). Patient, tumor, and treatment characteristics were recorded, and ML models were developed to predict 1-, 3-, and 5-year survival. The best performing ML model was externally validated using an institutional cohort of UPS patients (n = 151). All ML models performed best at the 1-year time point and worst at the 5-year time point. On internal validation within the SEER cohort, the best models had c-statistics of 0.67-0.69 at the 5-year time point. The Multi-Layer Perceptron Neural Network (MLP) model was the best performing model and used for external validation. Similarly, the MLP model performed best at 1-year and worst at 5-year on external validation with c-statistics of 0.85 and 0.81, respectively. The MLP model was well calibrated on external validation. The MLP model has been made publicly available at https://rachar.shinyapps.io/ups_app/ . Machine learning models perform well for survival prediction in UPS, though this sarcoma subtype may be more difficult to prognosticate than other subtypes. Future studies are needed to further validate the machine learning approach for UPS prognostication.

Escitalopram Personalized Dosing: A Population Pharmacokinetics Repository Method.

Escitalopram (SCIT) represents a first-line antidepressant and antianxiety medication. Pharmacokinetic studies of SCIT have demonstrated considerable interindividual variability, emphasizing the need for personalized dosing. Accordingly, we aimed to create a repository of parametric population pharmacokinetic (PPK) models of SCIT to facilitate model-informed precision dosing. In November 2022, we searched PubMed, Embase, and Web of Science for published PPK models and identified eight models. All the structural models reported in the literature were either one- or two-compartment models. In order to investigate the variances in model performance, the parameters of all PPK models were derived from the literature published. A representative virtual population, characterized by an age of 30, a body weight of 70 kg, and a BMI of 23 kg/m2, was generated for the purpose of replicating these models. To accomplish this, the rxode2 package in the R programming language was employed. Subsequently, we compared simulated concentration-time profiles and evaluated the impact of covariates on clearance. The most significant covariates were CYP2C19 phenotype, weight, and age, indicating that dosing regimens should be tailored accordingly. Additionally, among Chinese psychiatric patients, SCIT showed nearly double the exposure compared to other populations, specifically when considering the same CYP2C19 population restriction, which is a knowledge gap that needs further investigation. Furthermore, this repository of parametric PPK models for SCIT has a wide range of potential applications, like design miss or delay dose remedy strategies and external PPK model validation.

Deep learning classification of shoulder fractures on plain radiographs of the humerus, scapula and clavicle.

In this study, we present a deep learning model for fracture classification on shoulder radiographs using a convolutional neural network (CNN). The primary aim was to evaluate the classification performance of the CNN for proximal humeral fractures (PHF) based on the AO/OTA classification system. Secondary objectives included evaluating the model's performance for diaphyseal humerus, clavicle, and scapula fractures. The training dataset consisted of 6,172 examinations, including 2-7 radiographs per examination. The overall area under the curve (AUC) for fracture classification was 0.89, indicating good performance. For PHF classification, 12 out of 16 classes achieved an AUC of 0.90 or greater. Additionally, the CNN model had excellent overall AUC for diaphyseal humerus fractures (0.97), clavicle fractures (0.96), and good AUC for scapula fractures (0.87). Despite the limitations of the study, such as the reliance on ground truth labels provided by students with limited radiographic assessment experience, our findings are in concordance with previous studies, further consolidating CNN as potent fracture classifiers in plain radiographs. The inclusion of multiple radiographs with different views from each examination, as well as the generally unselected nature of the sample, contributed to the overall generalizability of the study. This is the fifth study published by our group on AI in orthopaedic radiographs, which has consistently shown promising results. The next challenge for the orthopaedic research community will be to transfer these results from the research setting into clinical practice. External validation of the CNN model should be conducted in the future before it is considered for use in a clinical setting.

Opioid burden in patients with inoperable pancreatic adenocarcinoma and the development of a multivariable risk prediction model for opioid use: A retrospective cohort study

IntroductionRisk prediction models to guide patient selection for early pre-emptive endoscopic ultrasound guided coeliac plexus neurolysis are lacking. This study aimed to determine in patients with inoperable pancreatic cancer: (1) opioid burden, (2) the relationship between opioid use and all-cause mortality, (3) risk factors for opioid use, and (4) develop and internally validate a risk prediction model for opioid use at three months. MethodsThis was a single-centre retrospective cohort study of patients with confirmed pancreatic cancer. Cox proportional hazard regression estimated the association between opioid use at baseline and all-cause mortality. Logistic regression estimated the associations between clinical and radiological variables with opioid use by three months. Two risk prediction models were developed for opioid use (clinical and clinical-radiological). Model discrimination and calibration was assessed. ResultsIn total, 383 patients with inoperable pancreatic cancer were included. Prevalence of pain ranged between 37% and 47% at three monthly intervals in the first year of diagnosis. Opioid use at baseline was associated with poorer survival. Age, pain at presentation, performance status, tumour distance from the right ganglion, the anterior-posterior and the latero-lateral tumour dimensions were independent risk factors for the opioid use at three months. The Area Under Curve (AUC) for the clinical and clinical-radiological models was 0.81 and 0.84, respectively. Models were well calibrated. ConclusionsOpioid use is prevalent in patients with pancreatic cancer, associated with poor prognosis, and can be predicted based on clinical and radiological variables. External validation of this predictive model is required.

Transportability of bacterial infection prediction models for critically ill patients.

Bacterial infections (BIs) are common, costly, and potentially life-threatening in critically ill patients. Patients with suspected BIs may require empiric multidrug antibiotic regimens and therefore potentially be exposed to prolonged and unnecessary antibiotics. We previously developed a BI risk model to augment practices and help shorten the duration of unnecessary antibiotics to improve patient outcomes. Here, we have performed a transportability assessment of this BI risk model in 2 tertiary intensive care unit (ICU) settings and a community ICU setting. We additionally explored how simple multisite learning techniques impacted model transportability. Patients suspected of having a community-acquired BI were identified in 3 datasets: Medical Information Mart for Intensive Care III (MIMIC), Northwestern Medicine Tertiary (NM-T) ICUs, and NM "community-based" ICUs. ICU encounters from MIMIC and NM-T datasets were split into 70/30 train and test sets. Models developed on training data were evaluated against the NM-T and MIMIC test sets, as well as NM community validation data. During internal validations, models achieved AUROCs of 0.78 (MIMIC) and 0.81 (NM-T) and were well calibrated. In the external community ICU validation, the NM-T model had robust transportability (AUROC 0.81) while the MIMIC model transported less favorably (AUROC 0.74), likely due to case-mix differences. Multisite learning provided no significant discrimination benefit in internal validation studies but offered more stability during transport across all evaluation datasets. These results suggest that our BI risk models maintain predictive utility when transported to external cohorts. Our findings highlight the importance of performing external model validation on myriad clinically relevant populations prior to implementation.

A clinical model to predict the progression of knee osteoarthritis: data from Dryad

BackgroundKnee osteoarthritis (KOA) is a multifactorial, slow-progressing, non-inflammatory degenerative disease primarily affecting synovial joints. It is usually induced by advanced age and/or trauma and eventually leads to irreversible destruction of articular cartilage and other tissues of the joint. Current research on KOA progression has limited clinical application significance. In this study, we constructed a prediction model for KOA progression based on multiple clinically relevant factors to provide clinicians with an effective tool to intervene in KOA progression.MethodThis study utilized the data set from the Dryad database which included patients with Kellgren–Lawrence (KL) grades 2 and 3. The KL grades was determined as the dependent variable, while 15 potential predictors were identified as independent variables. Patients were randomized into training set and validation set. The training set underwent LASSO analysis, model creation, visualization, decision curve analysis and internal validation using R language. The validation set is externally validated and F1-score, precision, and recall are computed.ResultA total of 101 patients with KL2 and 94 patients with KL3 were selected. We randomly split the data set into a training set and a validation set by 8:2. We filtered “BMI”, “TC”, “Hypertension treatment”, and “JBS3 (%)” to build the prediction model for progression of KOA. Nomogram used to visualize the model in R language. Area under ROC curve was 0.896 (95% CI 0.847–0.945), indicating high discrimination. Mean absolute error (MAE) of calibration curve = 0.041, showing high calibration. MAE of internal validation error was 0.043, indicating high model calibration. Decision curve analysis showed high net benefit. External validation of the metabolic syndrome column-line graph prediction model was performed by the validation set. The area under the ROC curve was 0.876 (95% CI 0.767–0.984), indicating that the model had a high degree of discrimination. Meanwhile, the calibration curve Mean absolute error was 0.113, indicating that the model had a high degree of calibration. The F1 score is 0.690, the precision is 0.667, and the recall is 0.714. The above metrics represent a good performance of the model.ConclusionWe found that KOA progression was associated with four variable predictors and constructed a predictive model for KOA progression based on the predictors. The clinician can intervene based on the nomogram of our prediction model.Key informationThis study is a clinical predictive model of KOA progression. KOA progression prediction model has good credibility and clinical value in the prevention of KOA progression.

Development and external validation of a diagnostic model for cardiometabolic-based chronic disease : results from the China health and retirement longitudinal study (CHARLS)

BackgroundCardiovascular disease(CVD) is the leading cause of death in the world. Cardiometabolic-based chronic disease (CMBCD) model is presented that provides a basis for sustainable and early, evidence-based therapeutic targeting to mitigate the ravagest and development of CVD. CMBCD include dysglycemia, hypertension, and/or dyslipidemia progressing to downstream CVD events.ObjectivesThe objective of our research was to develop and externally validate a diagnostic model of CMBCD.MethodsDesign: Multivariable logistic regression of a cohort for 9,463 participants aged at least 45 years were drawn from the 2018 wave of the China Health and Retirement Longitudinal Study (CHARLS). Setting: The 2018 wave of the CHARLS. Participants:Diagnostic model development: Totally 6,218 participants whose individual ID < 250,000,000,000. External validation: Totally 3,245 participants whose individual ID > 250,000,000,000. Outcomes: CMBCD .ResultsCMBCD occurred in 25.5%(1,584/6,218)of individuals in the development data set and 26.2%(850 /3,245)of individuals in the validation data set. The strongest predictors of CMBCD were age, general health status, location of residential address, smoking, housework ability, pain, and exercise tolerance. We developed a diagnostic model of CMBCD. Discrimination was the ability of the diagnostic model to differentiate between people who with and without CMBCD. This measure was quantified by calculating the area under the receiver operating characteristic(ROC) curve(AUC).The AUC was 0.6199 ± 0.0083, 95% confidence interval(CI) = 0.60372 ~ 0.63612. We constructed a nomograms using the development database based on age, general health status, location of residential address, smoking, housework ability, pain, and exercise tolerance. The AUC was 0.6033 ± 0.0116, 95% CI = 0.58066 ~ 0.62603 in the validation data set.ConclusionsWe developed and externally validated a diagnostic model of CMBCD. Discrimination, calibration, and decision curve analysis were satisfactory.

Development and external validation of a mixed-effects deep learning model to diagnose COVID-19 from CT imaging.

The automatic analysis of medical images has the potential improve diagnostic accuracy while reducing the strain on clinicians. Current methods analyzing 3D-like imaging data, such as computerized tomography imaging, often treat each image slice as individual slices. This may not be able to appropriately model the relationship between slices. Our proposed method utilizes a mixed-effects model within the deep learning framework to model the relationship between slices. We externally validated this method on a data set taken from a different country and compared our results against other proposed methods. We evaluated the discrimination, calibration, and clinical usefulness of our model using a range of measures. Finally, we carried out a sensitivity analysis to demonstrate our methods robustness to noise and missing data. In the external geographic validation set our model showed excellent performance with an AUROC of 0.930 (95%CI: 0.914, 0.947), with a sensitivity and specificity, PPV, and NPV of 0.778 (0.720, 0.828), 0.882 (0.853, 0.908), 0.744 (0.686, 0.797), and 0.900 (0.872, 0.924) at the 0.5 probability cut-off point. Our model also maintained good calibration in the external validation dataset, while other methods showed poor calibration. Deep learning can reduce stress on healthcare systems by automatically screening CT imaging for COVID-19. Our method showed improved generalizability in external validation compared to previous published methods. However, deep learning models must be robustly assessed using various performance measures and externally validated in each setting. In addition, best practice guidelines for developing and reporting predictive models are vital for the safe adoption of such models.

Clinical-radiomics-based treatment decision support for KIT Exon 11 deletion in gastrointestinal stromal tumors: a multi-institutional retrospective study.

gastrointestinal stromal tumors (GISTs) with KIT exon 11 deletions have more malignant clinical outcomes. A radiomics model was constructed for the preoperative prediction of KIT exon 11 deletion in GISTs. Overall, 126 patients with GISTs who underwent preoperative enhanced CT were included. GISTs were manually segmented using ITK-SNAP in the arterial phase (AP) and portal venous phase (PVP) images of enhanced CT. Features were extracted using Anaconda (version 4.2.0) with PyRadiomics. Radiomics models were constructed by LASSO. The clinical-radiomics model (combined model) was constructed by combining the clinical model with the best diagnostic effective radiomics model. ROC curves were used to compare the diagnostic effectiveness of radiomics model, clinical model, and combined model. Diagnostic effectiveness among radiomics model, clinical model and combine model were analyzed in external cohort (n=57). Statistics were carried out using R 3.6.1. The Radscore showed favorable diagnostic efficacy. Among all radiomics models, the AP-PVP radiomics model exhibited excellent performance in the training cohort, with an AUC of 0.787 (95% CI: 0.687-0.866), which was verified in the test cohort (AUC=0.775, 95% CI: 0.608-0.895). Clinical features were also analyzed. Among the radiomics, clinical and combined models, the combined model showed favorable diagnostic efficacy in the training (AUC=0.863) and test cohorts (AUC=0.851). The combined model yielded the largest AUC of 0.829 (95% CI, 0.621-0.950) for the external validation of the combined model. GIST patients could be divided into high or low risk subgroups of recurrence and mortality by the Radscore. The radiomics models based on enhanced CT for predicting KIT exon 11 deletion mutations have good diagnostic performance.

Development, External Validation, and Visualization of Machine Learning Models for Predicting Occurrence of Acute Kidney Injury after Cardiac Surgery.

Cardiac surgery-associated acute kidney injury (CSA-AKI) is a major complication that results in short- and long-term mortality among patients. Here, we adopted machine learning algorithms to build prediction models with the overarching goal of identifying patients who are at a high risk of such unfavorable kidney outcomes. A total of 1686 patients (development cohort) and 422 patients (validation cohort), with 126 pre- and intra-operative variables, were recruited from the First Medical Centre and the Sixth Medical Centre of Chinese PLA General Hospital in Beijing, China, respectively. Analyses were performed using six machine learning techniques, namely K-nearest neighbor, logistic regression, decision tree, random forest (RF), support vector machine, and neural network, and the APPROACH score, a previously established risk score for CSA-AKI. For model tuning, optimal hyperparameter was achieved by using GridSearch with 5-fold cross-validation from the scikit-learn library. Model performance was externally assessed via the receiver operating characteristic (ROC) and decision curve analysis (DCA). Explainable machine learning was performed using the Python SHapley Additive exPlanation (SHAP) package and Seaborn library, which allow the calculation of marginal contributory SHAP value. 637 patients (30.2%) developed CSA-AKI within seven days after surgery. In the external validation, the RF classifier exhibited the best performance among the six machine learning techniques, as shown by the ROC curve and DCA, while the traditional APPROACH risk score showed a relatively poor performance. Further analysis found no specific causative factor contributing to the development of CSA-AKI; rather, the development of CSA-AKI appeared to be a complex process resulting from a complex interplay of multiple risk factors. The SHAP summary plot illustrated the positive or negative contribution of RF-top 20 variables and extrapolated risk of developing CSA-AKI at individual levels. The Seaborn library showed the effect of each single feature on the model output of the RF prediction. Efficient machine learning approaches were successfully established to predict patients with a high probability of developing acute kidney injury after cardiac surgery. These findings are expected to help clinicians to optimize treatment strategies and minimize postoperative complications. The study protocol was registered at the ClinicalTrials Registration System (https://www.clinicaltrials.gov/, #NCT04966598) on July 26, 2021.

External validation and revision of Penn incisional hernia prediction model: A large-scale retrospective cohort of abdominal operations

BackgroundIncisional hernia (IH) manifests in 10%–15% of abdominal surgeries and patients at elevated risk of this complication should be identified for prophylactic intervention. This study aimed to externally validate the Penn hernia risk calculator. MethodsThe Ramathibodi abdominal surgery cohort was constructed by linking relevant hospital databases from 2010 to 2021. Penn hernia risk scores were calculated according to the original model which was externally validated using a seven-step approach. An updated model which included four additional predictor variables (i.e., age, immunosuppressive medication, ostomy reversal, and transfusion) added to those of the three original predictors (i.e., body mass index, chronic liver disease, and open surgery) was also evaluated. The area under the receiver operating characteristic curve (AUC) was estimated, and calibration performance was compared using the Hosmer–Lemeshow goodness-of-fit method for the observed/expected (O/E) ratio. ResultsA total of 12,155 abdominal operations were assessed. The original Penn model yielded fair discrimination with an AUC (95% confidence interval (CI)) of 0.645 (0.607, 0.683). The updated model that included the additional predictor variables achieved an acceptable AUC (95% CI) of 0.733 (0.698, 0.768) with the O/E ratio of 0.968 (0.848, 1.088). ConclusionThe updated model achieved improved discrimination and calibration performance, and should be considered for the identification of high-risk patients for further hernia prevention strategy.

Predicting models for arm impairment: External validation of the Scandinavian models and identification of new predictors in post-acute stroke settings.

Post-stroke arm impairment at rehabilitation admission as predictor of discharge arm impairment was consistently reported as extremely useful. Several models for acute prediction exist (e.g. the Scandinavian), though lacking external validation and larger time-window admission assessments. (1) use the 33 Fugl-Meyer Assessment-Upper Extremity (FMA-UE) individual items to predict total FMA-UE score at discharge of patients with ischemic stroke admitted to rehabilitation within 90 days post-injury, (2) use eight individual items (seven from the Scandinavian study plus the top predictor item from objective 1) to predict mild impairment (FMA-UE≥48) at discharge and (3) adjust the top three models from objective 2 with known confounders. This was an observational study including 287 patients (from eight settings) admitted to rehabilitation (2009-2020). We applied regression models to candidate predictors, reporting adjusted R2, odds ratios and ROC-AUC using 10-fold cross-validation. We achieved good predictive power for the eight item-level models (AUC: 0.70-0.82) and for the three adjusted models (AUC: 0.85-0.88). We identified finger mass flexion as new item-level top predictor (AUC:0.88) and time to admission (OR = 0.9(0.9;1.0)) as only common significant confounder. Scandinavian item-level predictors are valid in a different context, finger mass flexion outperformed known predictors, days-to-admission predict discharge mild arm impairment.

External Validation of a Machine Learning Model to Predict 6-Month Mortality for Patients With Advanced Solid Tumors

This prognostic study performed external validation of a machine learning model to predict 6-month mortality among patients with advanced solid tumors.

A rapid estimation of pigment content in virgin olive oils by a solvent‐free method

AbstractIn this work, a fast method was proposed for estimating the virgin olive oils (VOOs) carotenoids and chlorophylls concentration using color measurement. The pigment content by conventional spectrophotometry method and CIELAB color (L*, a*, and b*) at different degree of sample thickness (from 5 to 50 mm) of one hundred VOOs were measured. Oil carotenoids and chlorophylls content were correlated with the color parameters for the different oil thickness studied to design the prediction models of the new method. The best regression coefficients (R2) were obtained for multiple linear regression model using the three independent variables (L*, a*, and b*) together measured at 5 mm of oil thickness. The R2 were 0.9679, 0.9515, and 0.9644 for predicting carotenoids, chlorophylls, and total pigments, respectively. External validation of these prediction models was satisfactory (relative error &lt; 0.1). Therefore, this new solvent‐free colorimetric method is a useful method for determination of carotenoids and chlorophylls content in VOOs.Practical applications: The simple colorimetric method developed in this study offers a fast and accurate alternative to current methods published in the literature to estimate the pigment content in VOOs. It is a rapid (less than 1 min) and cheap method, with the advantage of ease of operation, no sample pretreatment and solvent‐free, thus environmentally friendly. This methodology can potentially be used by trained “nonprofessional analytical skilled” people in small laboratories or olive oil mills with limited technical facilities. Therefore, the technique is highly plausible as an alternative to determine the pigment content in VOOs. Finally, future works with this methodology could be carried out to online control of VOOs pigments content in the oil extraction process.

External validation and update of the J-ACCESS model in an Italian cohort of patients undergoing stress myocardial perfusion imaging

BackgroundCardiovascular risk models are based on traditional risk factors and investigations such as imaging tests. External validation is important to determine reproducibility and generalizability of a prediction model. We performed an external validation of t the Japanese Assessment of Cardiac Events and Survival Study by Quantitative Gated SPECT (J-ACCESS) model, developed from a cohort of patients undergoing stress myocardial perfusion imaging. MethodsWe included 3623 patients with suspected or known coronary artery disease undergoing stress single-photon emission computer tomography (SPECT) myocardial perfusion imaging at our academic center between January 2001 and December 2019. ResultsIn our study population, the J-ACCESS model underestimated the risk of major adverse cardiac events (cardiac death, nonfatal myocardial infarction, and severe heart failure requiring hospitalization) within three-year follow-up. The recalibrations and updated of the model slightly improved the initial performance: C-statistics increased from 0.664 to 0.666 and Brier score decreased from 0.075 to 0.073. Hosmer–Lemeshow test indicated a logistic regression fit only for the calibration slope (P = .45) and updated model (P = .22). In the update model, the intercept, diabetes, and severity of myocardial perfusion defects categorized coefficients were comparable with J-ACCESS. ConclusionThe external validation of the J-ACCESS model as well as recalibration models have a limited value for predicting of three-year major adverse cardiac events in our patients. The performance in predicting risk of the updated model resulted superimposable to the calibration slope model.