A5007378283- Machine Learning in Healthcare
- Sepsis Diagnosis and Treatment
- Acute Myocardial Infarction Research
- Emergency and Acute Care Studies
- Artificial Intelligence in Healthcare
- Venous Thromboembolism Diagnosis and Management
- Signaling Pathways in Disease
- Statistical Methods in Epidemiology
- Atrial Fibrillation Management and Outcomes
- Diabetes Management and Research
- Cardiac Imaging and Diagnostics
- Heart Failure Treatment and Management
- Biosimilars and Bioanalytical Methods
- Insurance, Mortality, Demography, Risk Management
- Peptidase Inhibition and Analysis
- Coronary Interventions and Diagnostics
- Diabetes and associated disorders
- Diabetes Management and Education
- Cardiovascular Health and Disease Prevention
- Medical Coding and Health Information
- Biotin and Related Studies
- Long-Term Effects of COVID-19
- Clinical Reasoning and Diagnostic Skills
- Blood properties and coagulation
- Pancreatic function and diabetes
Maastricht University
2018-2024
Maastricht University Medical Centre
2016-2024
University Medical Center
2018
Patients with sepsis who present to an emergency department (ED) have highly variable underlying disease severity, and can be categorized from low high risk. Development of a risk stratification tool for these patients is important appropriate triage early treatment. The aim this study was develop machine learning models predicting 31-day mortality in presenting the ED compare internal medicine physicians clinical scores.A single-center, retrospective cohort conducted amongst 1,344...
Background Closed-loop insulin delivery systems, which integrate continuous glucose monitoring (CGM) and algorithms that continuously guide dosing, have been shown to improve glycaemic control. The ability predict future values can further optimize such devices. In this study, we used machine learning train models in predicting levels based on prior CGM accelerometry data. Methods We data from Maastricht Study, an observational population‐based cohort comprises individuals with normal...
Abstract Aims CVD is the main cause of morbidity and mortality in individuals with diabetes. It currently unclear whether daily glucose variability contributes to CVD. Therefore, we investigated associated arterial measures that are considered important pathogenesis. Methods We included participants The Maastricht Study, an observational population-based cohort, who underwent at least 48 h continuous monitoring (CGM) ( n = 853; age: 59.9 ± 8.6 years; 49% women, 23% type 2 diabetes). studied...
Abstract We validated a Deep Embedded Clustering (DEC) model and its adaptation for integrating mixed datatypes (in this study, numerical categorical variables). is promising technique capable of managing extensive sets variables non-linear relationships. Nevertheless, DEC cannot adequately handle datatypes. Therefore, we adapted by replacing the autoencoder with an X-shaped variational (XVAE) optimising hyperparameters cluster stability. call “X-DEC”. compared X-DEC reproducing previous...
Abstract Background Many prediction models have been developed to help identify emergency department (ED) patients at high risk of poor outcome. However, these often underperform in clinical practice and their actual impact has hardly ever evaluated. We aim perform a trial investigate the model based on machine learning (ML) technology. Methods The study is prospective, randomized, open-label, non-inferiority pilot trial. will ML technology, RISK INDEX , which predict 31-day mortality...
Abstract BACKGROUND Cardiac troponin T (cTnT) is the preferred biomarker for diagnosis of acute myocardial infarction (AMI). It has been suggested that cTnT present predominantly in fragmented forms human serum following AMI. In this study, we have used a targeted mass spectrometry assay and epitope mapping using Western blotting to confirm hypothesis. METHODS was captured from 12 patients diagnosed with AMI an immunoprecipitation technique employing M11.7 catcher antibody fractionated...
The CoLab score was developed and externally validated to rule out COVID-19 among suspected patients presenting at the emergency department. We hypothesized a within-patient decrease in over time an intensive care unit (ICU) cohort. Such would create opportunity potentially need for isolation when infection is overcome. Using linear mixed-effects models, data from Maastricht Intensive Care COVID (MaastrICCht) cohort were used investigate association between score. Models adjusted sex, APACHE...
Cardiac troponin T (cTnT) is key in diagnosing myocardial infarction (MI) but also elevated end-stage renal disease (ESRD) patients. Specific larger cTnT proteoforms were identified for the acute phase of MI, while serum ESRD patients solely small fragments found. However, others allocated this to a pre-analytic effect due abundant thrombin generation serum. Therefore, we investigated various anticoagulation methods on composition and concentration compared MI
Figure 1 Bland-Altman plot of absolute hs-cTnT concentration differences before and after biotin depletion in 572 patients.The open dots are individual data points.The black line represents the median difference (0.00, 95% CI: -0.02 to 0.00) grey is reference line.
Background: Current guidelines recommend interpreting concentrations of NPs (natriuretic peptides) irrespective the time presentation to emergency department. We hypothesized that diurnal variations in NP concentration may affect their diagnostic accuracy for acute heart failure. Methods: In a secondary analysis multicenter study enrolling patients presenting with dyspnea department and using central adjudication final diagnosis by 2 independent cardiologists, failure BNP (B-type NP),...
Introduction Prediction models for identifying emergency department (ED) patients at high risk of poor outcome are often not externally validated. We aimed to perform a head-to-head comparison the discriminatory performance several prediction in large cohort ED patients.
Abstract Background Risk stratification of patients presenting to the emergency department (ED) is important for appropriate triage. Diagnostic laboratory tests are an essential part workup and risk these patients. Using machine learning, prognostic power clinical value can be amplified greatly. In this study, we applied learning develop accurate explainable decision support tool model that predicts likelihood 31-day mortality in ED (the RISKINDEX). This was developed evaluated four Dutch...
The role of cardiac troponins (cTn) have become increasingly important in diagnosing myocardial infarction (MI), especially patients without electrocardiogram abnormalities (1).
Non-Hispanic white (White) populations are overrepresented in medical studies. Potential healthcare disparities can happen when machine learning models, used diabetes technologies, trained on data from primarily White patients. We aimed to evaluate algorithmic fairness glucose predictions. This study utilized continuous monitoring (CGM) 101 and 104 Black participants with type 1 collected by the JAEB Center for Health Research, US. Long short-term memory (LSTM) deep models were 11 datasets...
Abstract Background Interest in prediction models, including machine learning (ML) based on laboratory data has increased tremendously. Uncertainty measurements and predictions such are inherently intertwined. This study developed a framework for assessing the impact of biological analytical variation uncertainty categorical models. Methods Practical application was demonstrated renal function loss (Chronic Kidney Disease Epidemiology Collaboration [CKD-EPI] equation) 31-day mortality...
Abstract Introduction Patients with sepsis who present to an emergency department (ED) have highly variable underlying disease severity, and can be categorized from low high risk. Development of a risk stratification tool for these patients is important appropriate triage early treatment. The aim this study was develop machine learning models predicting 31-day mortality in presenting the ED compare internal medicine physicians clinical scores. Methods A single-center, retrospective cohort...
Topic: 33. Bleeding disorders (congenital and acquired) Background: Hypodysfibrinogenemia is a rare hereditary fibrinogen disorder characterized by quantitative qualitative defects. These defects can cause thrombotic hemorrhagic phenotypes. Unfortunately, predicting the phenotype in specific patient often not possible with routine coagulation tests. Aims: To characterize genetic profile of family hypodysfibrinogenemia to investigate ability innovative tests predict bleeding and/or phenotypes...
Abstract Background Risk stratification of patients presenting to the emergency department (ED) is important for appropriate triage. Diagnostic laboratory tests are an essential part work-up and risk these patients. Using machine learning, prognostic power clinical value can be amplified greatly. In this study, we applied learning develop accurate explainable decision support tool model that predicts likelihood 31-day mortality in ED (the RISK INDEX ). This was developed evaluated four Dutch...