Deep embedded clustering generalisability and adaptation for integrating mixed datatypes: two critical care cohorts.

We validated a Deep Embedded Clustering (DEC) model and its adaptation for integrating mixed datatypes (in this study, numerical and categorical variables). Deep Embedded Clustering (DEC) is a promising technique capable of managing extensive sets of variables and non-linear relationships. Nevertheless, DEC cannot adequately handle mixed datatypes. Therefore, we adapted DEC by replacing the autoencoder with an X-shaped variational autoencoder (XVAE) and optimising hyperparameters for cluster stability. We call this model "X-DEC". We compared DEC and X-DEC by reproducing a previous study that used DEC to identify clusters in a population of intensive care patients. We assessed internal validity based on cluster stability on the development dataset. Since generalisability of clustering models has insufficiently been validated on external populations, we assessed external validity by investigating cluster generalisability onto an external validation dataset. We concluded that both DEC and X-DEC resulted in clinically recognisable and generalisable clusters, but X-DEC produced much more stable clusters.

Machine learning for risk stratification in the emergency department (MARS-ED) study protocol for a randomized controlled pilot trial on the implementation of a prediction model based on machine learning technology predicting 31-day mortality in the emergency department.

BACKGROUND: Many prediction models have been developed to help identify emergency department (ED) patients at high risk of poor outcome. However, these models often underperform in clinical practice and their actual clinical impact has hardly ever been evaluated. We aim to perform a clinical trial to investigate the clinical impact of a prediction model based on machine learning (ML) technology. METHODS: The study is a prospective, randomized, open-label, non-inferiority pilot clinical trial. We will investigate the clinical impact of a prediction model based on ML technology, the RISKINDEX, which has been developed to predict the risk of 31-day mortality based on the results of laboratory tests and demographic characteristics. In previous studies, the RISKINDEX was shown to outperform internal medicine specialists and to have high discriminatory performance. Adults patients (18 years or older) will be recruited in the ED. All participants will be randomly assigned to the control group or the intervention group in a 1:1 ratio. Participants in the control group will receive care as usual in which the study team asks the attending physicians questions about their clinical intuition. Participants in the intervention group will also receive care as usual, but in addition to asking the clinical impression questions, the study team presents the RISKINDEX to the attending physician in order to assess the extent to which clinical treatment is influenced by the results. DISCUSSION: This pilot clinical trial investigates the clinical impact and implementation of an ML based prediction model in the ED. By assessing the clinical impact and prognostic accuracy of the RISKINDEX, this study aims to contribute valuable insights to optimize patient care and inform future research in the field of ML based clinical prediction models. TRIAL REGISTRATION: ClinicalTrials.gov NCT05497830. Machine Learning for Risk Stratification in the Emergency Department (MARS-ED). Registered on August 11, 2022. URL: https://clinicaltrials.gov/study/NCT05497830 .

Explainable Machine Learning Models for Rapid Risk Stratification in the Emergency Department: A Multicenter Study.

BACKGROUND: Risk stratification of patients presenting to the emergency department (ED) is important for appropriate triage. Diagnostic laboratory tests are an essential part of the workup and risk stratification of these patients. Using machine learning, the prognostic power and clinical value of these tests can be amplified greatly. In this study, we applied machine learning to develop an accurate and explainable clinical decision support tool model that predicts the likelihood of 31-day mortality in ED patients (the RISKINDEX). This tool was developed and evaluated in four Dutch hospitals. METHODS: Machine learning models included patient characteristics and available laboratory data collected within the first 2 h after ED presentation, and were trained using 5 years of data from consecutive ED patients from the Maastricht University Medical Center (Maastricht), Meander Medical Center (Amersfoort), and Zuyderland Medical Center (Sittard and Heerlen). A sixth year of data was used to evaluate the models using area under the receiver-operating-characteristic curve (AUROC) and calibration curves. The Shapley additive explanations (SHAP) algorithm was used to obtain explainable machine learning models. RESULTS: The present study included 266 327 patients with 7.1 million laboratory results available. Models show high diagnostic performance with AUROCs of 0.94, 0.98, 0.88, and 0.90 for Maastricht, Amersfoort, Sittard and Heerlen, respectively. The SHAP algorithm was utilized to visualize patient characteristics and laboratory data patterns that underlie individual RISKINDEX predictions. CONCLUSIONS: Our clinical decision support tool has excellent diagnostic performance in predicting 31-day mortality in ED patients. Follow-up studies will assess whether implementation of these algorithms can improve clinically relevant end points.

Head-to-head comparison of 19 prediction models for short-term outcome in medical patients in the emergency department: a retrospective study.

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 of the discriminatory performance of several prediction models in a large cohort of ED patients. METHODS: In this retrospective study, we selected prediction models that aim to predict poor outcome and we included adult medical ED patients. Primary outcome was 31-day mortality, secondary outcomes were 1-day mortality, 7-day mortality, and a composite endpoint of 31-day mortality and admission to intensive care unit (ICU).The discriminatory performance of the prediction models was assessed using an area under the receiver operating characteristic curve (AUC). Finally, the prediction models with the highest performance to predict 31-day mortality were selected to further examine calibration and appropriate clinical cut-off points. RESULTS: We included 19 prediction models and applied these to 2185 ED patients. Thirty-one-day mortality was 10.6% (231 patients), 1-day mortality was 1.4%, 7-day mortality was 4.4%, and 331 patients (15.1%) met the composite endpoint. The RISE UP and COPE score showed similar and very good discriminatory performance for 31-day mortality (AUC 0.86), 1-day mortality (AUC 0.87), 7-day mortality (AUC 0.86) and for the composite endpoint (AUC 0.81). Both scores were well calibrated. Almost no patients with RISE UP and COPE scores below 5% had an adverse outcome, while those with scores above 20% were at high risk of adverse outcome. Some of the other prediction models (i.e. APACHE II, NEWS, WPSS, MEWS, EWS and SOFA) showed significantly higher discriminatory performance for 1-day and 7-day mortality than for 31-day mortality. CONCLUSIONS: Head-to-head validation of 19 prediction models in medical ED patients showed that the RISE UP and COPE score outperformed other models regarding 31-day mortality.

Diurnal Variations in Natriuretic Peptide Levels: Clinical Implications for the Diagnosis of Acute Heart Failure.

BACKGROUND: Current guidelines recommend interpreting concentrations of NPs (natriuretic peptides) irrespective of the time of presentation to the emergency department. We hypothesized that diurnal variations in NP concentration may affect their diagnostic accuracy for acute heart failure. METHODS: In a secondary analysis of a multicenter diagnostic study enrolling patients presenting with acute dyspnea to the emergency department and using central adjudication of the final diagnosis by 2 independent cardiologists, the diagnostic accuracy for acute heart failure of BNP (B-type NP), NT-proBNP (N-terminal pro-B-type NP), and MR-proANP (midregional pro-atrial NP) was compared among 1577 daytime presenters versus 908 evening/nighttime presenters. In a validation study, the presence of a diurnal rhythm in BNP and NT-proBNP concentrations was examined by hourly measurements in 44 stable individuals. RESULTS: Among patients adjudicated to have acute heart failure, BNP, NT-proBNP, and MR-proANP concentrations were comparable among daytime versus evening/nighttime presenters (all P=nonsignificant). Contrastingly, among patients adjudicated to have other causes of dyspnea, evening/nighttime presenters had lower BNP (median, 44 [18-110] versus 74 [27-168] ng/L; P<0.01) and NT-proBNP (median, 212 [72-581] versus 297 [102-902] ng/L; P<0.01) concentrations versus daytime presenters. This resulted in higher diagnostic accuracy as quantified by the area under the curve of BNP and NT-proBNP among evening/nighttime presenters (0.97 [95% CI, 0.95-0.98] and 0.95 [95% CI, 0.93-0.96] versus 0.94 [95% CI, 0.92-0.95] and 0.91 [95% CI, 0.90-0.93]) among daytime presenters (both P<0.01). These differences were not observed for MR-proANP. Diurnal variation of BNP and NT-proBNP with lower evening/nighttime concentration was confirmed in 44 stable individuals (P<0.01). CONCLUSIONS: BNP and NT-proBNP, but not MR-proANP, exhibit a diurnal rhythm that results in even higher diagnostic accuracy among evening/nighttime presenters versus daytime presenters. REGISTRATION: URL: https://www. CLINICALTRIALS: gov; Unique identifiers: NCT01831115, NCT02091427, and NCT02210897.

Machine learning-based glucose prediction with use of continuous glucose and physical activity monitoring data: The Maastricht Study.

BACKGROUND: Closed-loop insulin delivery systems, which integrate continuous glucose monitoring (CGM) and algorithms that continuously guide insulin dosing, have been shown to improve glycaemic control. The ability to predict future glucose values can further optimize such devices. In this study, we used machine learning to train models in predicting future glucose levels based on prior CGM and accelerometry data. METHODS: We used data from The Maastricht Study, an observational population-based cohort that comprises individuals with normal glucose metabolism, prediabetes, or type 2 diabetes. We included individuals who underwent >48h of CGM (n = 851), most of whom (n = 540) simultaneously wore an accelerometer to assess physical activity. A random subset of individuals was used to train models in predicting glucose levels at 15- and 60-minute intervals based on either CGM data or both CGM and accelerometer data. In the remaining individuals, model performance was evaluated with root-mean-square error (RMSE), Spearman's correlation coefficient (rho) and surveillance error grid. For a proof-of-concept translation, CGM-based prediction models were optimized and validated with the use of data from individuals with type 1 diabetes (OhioT1DM Dataset, n = 6). RESULTS: Models trained with CGM data were able to accurately predict glucose values at 15 (RMSE: 0.19mmol/L; rho: 0.96) and 60 minutes (RMSE: 0.59mmol/L, rho: 0.72). Model performance was comparable in individuals with type 2 diabetes. Incorporation of accelerometer data only slightly improved prediction. The error grid results indicated that model predictions were clinically safe (15 min: >99%, 60 min >98%). Our prediction models translated well to individuals with type 1 diabetes, which is reflected by high accuracy (RMSEs for 15 and 60 minutes of 0.43 and 1.73 mmol/L, respectively) and clinical safety (15 min: >99%, 60 min: >91%). CONCLUSIONS: Machine learning-based models are able to accurately and safely predict glucose values at 15- and 60-minute intervals based on CGM data only. Future research should further optimize the models for implementation in closed-loop insulin delivery systems.

Cardiac Troponin T: The Impact of Posttranslational Modifications on Analytical Immunoreactivity in Blood up to the Excretion in Urine.

Cardiac troponin T (cTnT) is a sensitive and specific biomarker for detecting cardiac muscle injury. Its concentration in blood can be significantly elevated outside the normal reference range under several pathophysiological conditions. The classical analytical method in routine clinical analysis to detect cTnT in serum or plasma is a single commercial immunoassay, which is designed to quantify the intact cTnT molecule. The targeted epitopes are located in the central region of the cTnT molecule. However, in blood cTnT exists in different biomolecular complexes and proteoforms: bound (to cardiac troponin subunits or to immunoglobulins) or unbound (as intact protein or as proteolytic proteoforms). While proteolysis is a principal posttranslational modification (PTM), other confirmed PTMs of the proteoforms include N-terminal initiator methionine removal, N-acetylation, O-phosphorylation, O-(N-acetyl)-glucosaminylation, N(ɛ)-(carboxymethyl)lysine modification and citrullination. The immunoassay probably detects several of those cTnT biomolecular complexes and proteoforms, as long as they have the centrally targeted epitopes in common. While analytical cTnT immunoreactivity has been studied predominantly in blood, it can also be detected in urine, although it is unclear in which proteoform cTnT immunoreactivity is present in urine. This review presents an overview of the current knowledge on the pathophysiological lifecycle of cTnT. It provides insight into the impact of PTMs, not only on the analytical immunoreactivity, but also on the excretion of cTnT in urine as one of the waste routes in that lifecycle. Accordingly, and after isolating the proteoforms from urine of patients suffering from proteinuria and acute myocardial infarction, the structures of some possible cTnT proteoforms are reconstructed using mass spectrometry and presented.

Greater daily glucose variability and lower time in range assessed with continuous glucose monitoring are associated with greater aortic stiffness: The Maastricht Study.

AIMS: CVD is the main cause of morbidity and mortality in individuals with diabetes. It is currently unclear whether daily glucose variability contributes to CVD. Therefore, we investigated whether glucose variability is associated with arterial measures that are considered important in CVD pathogenesis. METHODS: We included participants of The Maastricht Study, an observational population-based cohort, who underwent at least 48 h of continuous glucose monitoring (CGM) (n = 853; age: 59.9 ± 8.6 years; 49% women, 23% type 2 diabetes). We studied the cross-sectional associations of two glucose variability indices (CGM-assessed SD [SDCGM] and CGM-assessed CV [CVCGM]) and time in range (TIRCGM) with carotid-femoral pulse wave velocity (cf-PWV), carotid distensibility coefficient, carotid intima-media thickness, ankle-brachial index and circumferential wall stress via multiple linear regression. RESULTS: Higher SDCGM was associated with higher cf-PWV after adjusting for demographics, cardiovascular risk factors and lifestyle factors (regression coefficient [B] per 1 mmol/l SDCGM [and corresponding 95% CI]: 0.413 m/s [0.147, 0.679], p = 0.002). In the model additionally adjusted for CGM-assessed mean sensor glucose (MSGCGM), SDCGM and MSGCGM contributed similarly to cf-PWV (respective standardised regression coefficients [st.ßs] and 95% CIs of 0.065 [-0.018, 0.167], p = 0.160; and 0.059 [-0.043, 0.164], p = 0.272). In the fully adjusted models, both higher CVCGM (B [95% CI] per 10% CVCGM: 0.303 m/s [0.046, 0.559], p = 0.021) and lower TIRCGM (B [95% CI] per 10% TIRCGM: -0.145 m/s [-0.252, -0.038] p = 0.008) were statistically significantly associated with higher cf-PWV. Such consistent associations were not observed for the other arterial measures. CONCLUSIONS: Our findings show that greater daily glucose variability and lower TIRCGM are associated with greater aortic stiffness (cf-PWV) but not with other arterial measures. If corroborated in prospective studies, these results support the development of therapeutic agents that target both daily glucose variability and TIRCGM to prevent CVD.

A comparison of machine learning models versus clinical evaluation for mortality prediction in patients with sepsis.

INTRODUCTION: Patients with sepsis who present to an emergency department (ED) have highly variable underlying disease severity, and can be categorized from low to high risk. Development of a risk stratification tool for these patients is important for appropriate triage and early treatment. The aim of this study was to develop machine learning models predicting 31-day mortality in patients presenting to the ED with sepsis and to compare these to internal medicine physicians and clinical risk scores. METHODS: A single-center, retrospective cohort study was conducted amongst 1,344 emergency department patients fulfilling sepsis criteria. Laboratory and clinical data that was available in the first two hours of presentation from these patients were randomly partitioned into a development (n = 1,244) and validation dataset (n = 100). Machine learning models were trained and evaluated on the development dataset and compared to internal medicine physicians and risk scores in the independent validation dataset. The primary outcome was 31-day mortality. RESULTS: A number of 1,344 patients were included of whom 174 (13.0%) died. Machine learning models trained with laboratory or a combination of laboratory + clinical data achieved an area-under-the ROC curve of 0.82 (95% CI: 0.80-0.84) and 0.84 (95% CI: 0.81-0.87) for predicting 31-day mortality, respectively. In the validation set, models outperformed internal medicine physicians and clinical risk scores in sensitivity (92% vs. 72% vs. 78%;p<0.001,all comparisons) while retaining comparable specificity (78% vs. 74% vs. 72%;p>0.02). The model had higher diagnostic accuracy with an area-under-the-ROC curve of 0.85 (95%CI: 0.78-0.92) compared to abbMEDS (0.63,0.54-0.73), mREMS (0.63,0.54-0.72) and internal medicine physicians (0.74,0.65-0.82). CONCLUSION: Machine learning models outperformed internal medicine physicians and clinical risk scores in predicting 31-day mortality. These models are a promising tool to aid in risk stratification of patients presenting to the ED with sepsis.

Identification and Characterization of Cardiac Troponin T Fragments in Serum of Patients Suffering from Acute Myocardial Infarction.

BACKGROUND: Cardiac troponin T (cTnT) is the preferred biomarker for the diagnosis of acute myocardial infarction (AMI). It has been suggested that cTnT is present predominantly in fragmented forms in human serum following AMI. In this study, we have used a targeted mass spectrometry assay and epitope mapping using Western blotting to confirm this hypothesis. METHODS: cTnT was captured from the serum of 12 patients diagnosed with AMI using an immunoprecipitation technique employing the M11.7 catcher antibody and fractionated with SDS-PAGE. Coomassie-stained bands of 4 patients at 37, 29, and 16 kDa were excised from the gel, digested with trypsin, and analyzed on a Q Exactive instrument set on targeted Selected Ion Monitoring mode with data-dependent tandem mass spectrometry (MS/MS) for identification. Western blotting employing 3 different antibodies was used for epitope mapping. RESULTS: Ten cTnT peptides of interest were targeted. By using MS/MS, all of these peptides were identified in the 37-kDa, intact, cTnT band. In the 29- and 16-kDa fragment bands, 8 and 4 cTnT-specific peptides were identified, respectively. Some of these peptides were "semitryptic," meaning that their C-termini were not formed by trypsin cleavage. The C-termini of these semitryptic peptides represent the C-terminal end of the cTnT molecules present in these bands. These results were confirmed independently by epitope mapping. CONCLUSIONS: Using LC-MS, we have succeeded in positively identifying the 29- and 16-kDa fragment bands as cTnT-derived products. The amino acid sequences of the 29- and 16-kDa fragments are Ser79-Trp297 and Ser79-Gln199, respectively.

Cardiac troponin T degradation in serum is catalysed by human thrombin.

Cardiac troponin T (cTnT) has been shown to be present in fragmented forms in human serum after acute myocardial infarction (AMI). While calpain-1 and caspase-3 have been identified as intracellular proteases able to cleave the N-terminus of cTnT, it is still unclear which proteases are responsible for the extensive and progressive cTnT fragmentation observed in serum of AMI-patients. In this pilot study we have investigated the possibility that human thrombin may be involved in this process. Purified human cTnT was spiked in unprocessed and deproteinated serum in the presence or absence of either purified human thrombin or PPACK thrombin inhibitor. After immunoprecipitation, SDS-PAGE and Western blotting we observed an increase in cTnT fragmentation when purified thrombin was added to deproteinated serum. Consequently, the addition of thrombin inhibitor to unprocessed serum resulted in a decrease of cTnT fragmentation. Our results suggest that multiple enzymes are involved in cTnT degradation, and that thrombin plays an important role.