Enhancing Robustness of Machine Learning Integration With Routine Laboratory Blood Tests to Predict Inpatient Mortality After Intracerebral Hemorrhage

Objective: The accurate evaluation of outcomes at a personalized level in patients with intracerebral hemorrhage (ICH) is critical clinical implications. This study aims to evaluate how machine learning integrates with routine laboratory tests and electronic health records (EHRs) data to predict inpatient mortality after ICH.Methods: In this machine learning-based prognostic study, we included 1,835 consecutive patients with acute ICH between October 2010 and December 2018. The model building process incorporated five pre-implant ICH score variables (clinical features) and 13 out of 59 available routine laboratory parameters. We assessed model performance according to a range of learning metrics, such as the mean area under the receiver operating characteristic curve [AUROC]. We also used the Shapley additive explanation algorithm to explain the prediction model.Results: Machine learning models using laboratory data achieved AUROCs of 0.71–0.82 in a split-by-year development/testing scheme. The non-linear eXtreme Gradient Boosting model yielded the highest prediction accuracy. In the held-out validation set of development cohort, the predictive model using comprehensive clinical and laboratory parameters outperformed those using clinical alone in predicting in-hospital mortality (AUROC [95% bootstrap confidence interval], 0.899 [0.897–0.901] vs. 0.875 [0.872–0.877]; P <0.001), with over 81% accuracy, sensitivity, and specificity. We observed similar performance in the testing set.Conclusions: Machine learning integrated with routine laboratory tests and EHRs could significantly promote the accuracy of inpatient ICH mortality prediction. This multidimensional composite prediction strategy might become an intelligent assistive prediction for ICH risk reclassification and offer an example for precision medicine.

Urinary Orosomucoid (α-1-Acid Glycoprotein) As A Single Potential Biomarker Through PET Assay for Diagnosis of Early Onset Sepsis in Neonates: A Review

Neonatal sepsis is a condition in which bacteria are present in an infant’s sterile body fluids. It is considered one of the most common causes of infant death, with nearly one million deaths per birthday and approximately 2 million deaths in the first week of life. To aid in the early diagnosis of neonatal sepsis, a potential new biomarker for early neonatal sepsis called orosomucoid (ORM) or α1-glycoprotein (α1AGP) in urine is being evaluated because of its greater accuracy than current diagnostic tools. Combined with particle turbidity analysis (PET), neonatal sepsis can be diagnosed in an immediate, sensitive, specific and non-invasive manner. The early local increase in urinary ORM in sepsis suggests that it could be a new promising marker of sepsis and an important part of routine laboratory and clinical practice.

Success and Challenges Associated with Large Scale Collaborative Surveillance for Carbapenemase Genes in Gram Negative Bacteria

The emergence and spread of antimicrobial resistance, especially in Gram negative bacteria has led to significant morbidity and increased cost of healthcare. Large surveillance studies such as the one performed by the Antibiotic Resistance Laboratory Network are immensely valuable in understanding the scope of resistance mechanisms especially among carbapenemase producing Gram negative bacteria. However, the routine laboratory detection of carbapenemases in these bacteria remain challenging and require further optimization.

Long-term within- and between-subject biological variation of 29 routine laboratory measurands in athletes

Objectives Within- and between-subject biological variation (BV) estimates have many applications in laboratory medicine. However, robust high-quality BV estimates are lacking for many populations, such as athletes. This study aimed to deliver BV estimates of 29 routine laboratory measurands derived from a Biological Variation Data Critical Appraisal Checklist compliant design in a population of high-endurance athletes. Methods Eleven samples per subject were drawn from 30 triathletes monthly, during a whole sport season. Serum samples were measured in duplicate for proteins, liver enzymes, lipids and kidney-related measurands on an Advia2400 (Siemens Healthineers). After outlier and homogeneity analysis, within-subject (CVI) and between-subject (CVG) biological variation estimates were delivered (CV-ANOVA and log-ANOVA, respectively) and a linear mixed model was applied to analyze the effect of exercise and health related variables. Results Most CVI estimates were similar or only slightly higher in athletes compared to those reported for the general population, whereas two- to three-fold increases were observed for amylase, ALT, AST and ALP. No effect of exercise and health related variables were observed on the CVI estimates. For seven measurands, data were not homogeneously distributed and BV estimates were therefore not reported. Conclusions The observation of higher CVI estimates in athletes than what has been reported for the general population may be related to physiological stress over time caused by the continuous practice of exercise. The BV estimates derived from this study could be applied to athlete populations from disciplines in which they exercise under similar conditions of intensity and duration.

Evaluation of an Integrated Smart Sensor System for Real-Time Characterization and Digitalization of Postoperative Abdominal Drain Output: A Pilot Study

Background: For centuries, surgeons have relied on surgical drains during postoperative care. Despite all advances in modern medicine and the area of digitalization, as of today, most if not all assessment of abdominal secretions excreted via surgical drains are carried out manually. We here introduce a novel integrated Smart Sensor System ( Smart Drain) that allows for real-time characterization and digitalization of postoperative abdominal drain output at the patient’s bedside. Methods: A prototype of the Smart Drain was developed using a sophisticated spectrometer for assessment of drain output. The prototype measures 10 × 6 × 6 cm and therefore easily fits at the bedside. At the time of measurement with our Smart Drain, the drain output was additionally sent off to be analyzed in our routine laboratory for typical markers of interest in abdominal surgery such as bilirubin, lipase, amylase, triglycerides, urea, protein, and red blood cells. A total of 45 samples from 19 patients were included. Results: The measurements generated were found to correlate with conventional laboratory measurements for bilirubin (r = .658, P = .000), lipase (r = .490, P = .002), amylase (r = .571, P = .000), triglycerides (r = .803, P = .000), urea (r = .326, P = .033), protein (r = .387, P = .012), and red blood cells (r = .904, P = .000). Conclusions: To our best knowledge, for the first time we describe a device using a sophisticated spectrometer that allows for real-time characterization and digitalization of postoperative abdominal drain output at the patient’s bedside.

Using artificial intelligence in a primary care setting to identify patients at risk for cancer: a risk prediction model based on routine laboratory tests

Objectives To evaluate the ability of an artificial intelligence (AI) model to predict the risk of cancer in patients referred from primary care based on routine blood tests. Results obtained with the AI model are compared to results based on logistic regression (LR). Methods An analytical profile consisting of 25 predefined routine laboratory blood tests was introduced to general practitioners (GPs) to be used for patients with non-specific symptoms, as an additional tool to identify individuals at increased risk of cancer. Consecutive analytical profiles ordered by GPs from November 29th 2011 until March 1st 2020 were included. AI and LR analysis were performed on data from 6,592 analytical profiles for their ability to detect cancer. Cohort I for model development included 5,224 analytical profiles ordered by GP’s from November 29th 2011 until the December 31st 2018, while 1,368 analytical profiles included from January 1st 2019 until March 1st 2020 constituted the “out of time” validation test Cohort II. The main outcome measure was a cancer diagnosis within 90 days. Results The AI model based on routine laboratory blood tests can provide an easy-to use risk score to predict cancer within 90 days. Results obtained with the AI model were comparable to results from the LR model. In the internal validation Cohort IB, the AI model provided slightly better results than the LR analysis both in terms of the area under the receiver operating characteristics curve (AUC) and PPV, sensitivity/specificity while in the “out of time” validation test Cohort II, the obtained results were comparable. Conclusions The AI risk score may be a valuable tool in the clinical decision-making. The score should be further validated to determine its applicability in other populations.

Predicting psoriasis using routine laboratory tests with random forest

Psoriasis is a chronic inflammatory skin disease that affects approximately 125 million people worldwide. It has significant impacts on both physical and emotional health-related quality of life comparable to other major illnesses. Accurately prediction of psoriasis using biomarkers from routine laboratory tests has important practical values. Our goal is to derive a powerful predictive model for psoriasis disease based on only routine hospital tests. We collected a data set including 466 psoriasis patients and 520 healthy controls with 81 variables from only laboratory routine tests, such as age, total cholesterol, HDL cholesterol, blood pressure, albumin, and platelet distribution width. In this study, Boruta feature selection method was applied to select the most relevant features, with which a Random Forest model was constructed. The model was tested with 30 repetitions of 10-fold cross-validation. Our classification model yielded an average accuracy of 86.9%. 26 notable features were selected by Boruta, among which 15 features are confirmed from previous studies, and the rest are worth further investigations. The experimental results demonstrate that the machine learning approach has good potential in predictive modeling for the psoriasis disease given the information only from routine hospital tests.