Intensive care clinicians are presented with large quantities of patient information and measurements from a multitude of monitoring systems. The limited ability of humans to process such complex information hinders physicians to readily recognize and act on early signs of patient deterioration. We used machine learning to develop an early warning system for circulatory failure based on a high-resolution ICU database with 240 patientyears of data. This automatic system predicts 90.0% of circulatory failure events (prevalence 3.1%), with 81.8% identified more than two hours in advance, resulting in an area under the receiver operating characteristic curve of 94.0% and area under the precision recall curve of 63.0%. The model was externally validated in a large independent patient cohort.

As data science and artificial intelligence continue to rapidly gain traction, the publication of freely available ICU datasets has become invaluable to propel data-driven clinical research. In this guide for clinicians and researchers, we aim to: 1) systematically search and identify all publicly available adult clinical ICU datasets, 2) compare their characteristics, data quality, and richness and critically appraise their strengths and weaknesses, and 3) provide researchers with suggestions, which datasets are appropriate for answering their clinical question.DATA SOURCES: A systematic search was performed in Pubmed, ArXiv, MedRxiv, and BioRxiv. STUDY SELECTION:We selected all studies that reported on publicly available adult patient-level intensive care datasets.DATA EXTRACTION: A total of four publicly available, adult, critical care, patientlevel databases were included (Amsterdam University Medical Center data base [AmsterdamUMCdb], eICU Collaborative Research Database eICU CRD], High time-resolution intensive care unit dataset [HiRID], and Medical Information Mart for Intensive Care-IV). Databases were compared using a priori defined categories, including demographics, patient characteristics, and data richness. The study protocol and search strategy were prospectively registered. DATA SYNTHESIS:Four ICU databases fulfilled all criteria for inclusion and were queried using SQL (PostgreSQL version 12; PostgreSQL Global Development Group) and analyzed using R (R Foundation for Statistical Computing, Vienna, Austria). The number of unique patient admissions varied between 23,106 (AmsterdamUMCdb) and 200,859 (eICU-CRD). Frequency of laboratory values and vital signs was highest in HiRID, for example, 5.2 (±3.4) lactate values per day and 29.7 (±10.2) systolic blood pressure values per hour. Treatment intensity varied with vasopressor and ventilatory support in 69.0% and 83.0% of patients in AmsterdamUMCdb versus 12.0% and 21.0% in eICU-CRD, respectively. ICU mortality ranged from 5.5% in eICU-CRD to 9.9% in AmsterdamUMCdb. CONCLUSIONS:We identified four publicly available adult clinical ICU datasets. Sample size, severity of illness, treatment intensity, and frequency of reported parameters differ markedly between the databases. This should guide clinicians and researchers which databases to best answer their clinical questions.

Ventilator-associated pneumonia (VAP) is a frequent complication of mechanical ventilation and is associated with substantial morbidity and mortality. Accurate diagnosis of VAP relies in part on subjective diagnostic criteria. Surveillance according to ventilator-associated event (VAE) criteria may allow quick and objective benchmarking. Our objective was to create an automated surveillance tool for VAE tiers I and II on a large data collection, evaluate its diagnostic accuracy and retrospectively determine the yearly baseline VAE incidence. We included all consecutive intensive care unit admissions of patients with mechanical ventilation at Bern University Hospital, a tertiary referral center, from January 2008 to July 2016. Data was automatically extracted from the patient data management system and automatically processed. We created and implemented an application able to automatically analyze respiratory and relevant medication data according to the Centers for Disease Control protocol for VAE-surveillance. In a subset of patients, we compared the accuracy of automated VAE surveillance according to CDC criteria to a gold standard (a composite of automated and manual evaluation with mediation for discrepancies) and evaluated the evolution of the baseline incidence. The study included 22′442 ventilated admissions with a total of 37′221 ventilator days. 592 ventilator-associated events (tier I) occurred; of these 194 (34%) were of potentially infectious origin (tier II). In our validation sample, automated surveillance had a sensitivity of 98% and specificity of 100% in detecting VAE compared to the gold standard. The yearly VAE incidence rate ranged from 10.1–22.1 per 1000 device days and trend showed a decrease in the yearly incidence rate ratio of 0.96 (95% CI, 0.93–1.00, p = 0.03). This study demonstrated that automated VAE detection is feasible, accurate and reliable and may be applied on a large, retrospective sample and provided insight into long-term institutional VAE incidences. The surveillance tool can be extended to other centres and provides VAE incidences for performing quality control and intervention studies.

The development of respiratory failure is common among patients in intensive care units (ICU). Large data quantities from ICU patient monitoring systems make timely and comprehensive analysis by clinicians difficult but are ideal for automatic processing by machine learning algorithms. Early prediction of respiratory system failure could alert clinicians to patients at risk of respiratory failure and allow for early patient reassessment and treatment adjustment. We propose an early warning system that predicts moderate/severe respiratory failure up to 8 hours in advance. Our system was trained on HiRID-II, a data-set containing more than 60,000 admissions to a tertiary care ICU. An alarm is typically triggered several hours before the beginning of respiratory failure. Our system outperforms a clinical baseline mimicking traditional clinical decision-making based on pulse-oximetric oxygen saturation and the fraction of inspired oxygen. To provide model introspection and diagnostics, we developed an easy-to-use web browser-based system to explore model input data and predictions visually.

No abstract

The recent success of machine learning methods applied to time series collected from Intensive Care Units (ICU) exposes the lack of standardized machine learning benchmarks for developing and comparing such methods. While raw datasets, such as MIMIC-IV or eICU, can be freely accessed on Physionet, the choice of tasks and pre-processing is often chosen ad-hoc for each publication, limiting comparability across publications. In this work, we aim to improve this situation by providing a benchmark covering a large spectrum of ICU-related tasks. Using the HiRID dataset, we define multiple clinically relevant tasks developed in collaboration with clinicians. In addition, we provide a reproducible end-to-end pipeline to construct both data and labels. Finally, we provide an in-depth analysis of current state-of-the-art sequence modeling methods, highlighting some limitations of deep learning approaches for this type of data. With this benchmark, we hope to give the research community the possibility of a fair comparison of their work.

In the CNS, amino acid (AA) neurotransmitters and neurotransmitter precursors are subject to tight homeostatic control mediated by blood-brain barrier (BBB) solute carrier amino acid transporters (AATs). Since the BBB is composed of multiple closely apposed cell types and opportunities for human in vivo studies are limited, we used in vitro and computational approaches to investigate human BBB AAT activity and regulation. Quantitative real-time PCR (qPCR) of the human BBB endothelial cell model hCMEC/D3 (D3) was used to determine expression of selected AAT, tight junction (TJ), and signal transduction (ST) genes under various culture conditions. L-leucine uptake data were interrogated with a computational model developed by our group for calculating AAT activity in complex cell cultures. This approach is potentially applicable to in vitro cell culture drug studies where multiple “receptors” may mediate observed responses. Of 7 Leu AAT genes expressed by D3 only the activity of SLC7A5-SLC3A2/LAT1-4F2HC (LAT1), SLC43A2/LAT4 (LAT4) and sodium-dependent AATs, SLC6A15/B0AT2 (B0AT2), and SLC7A7/y+LAT1 (y+LAT1) were calculated to be required for Leu uptake. Therefore, D3 Leu transport may be mediated by a potentially physiologically relevant functional cooperation between the known BBB AAT, LAT1 and obligatory exchange (y+LAT1), facilitative diffusion (LAT4), and sodium symporter (B0AT2) transporters.

Dynamic assessment of mortality risk in the intensive care unit (ICU) can be used to stratify patients, inform about treatment effectiveness or serve as part of an early-warning system. Static risk scoring systems, such as APACHE or SAPS, have recently been supplemented with data-driven approaches that track the dynamic mortality risk over time. Recent works have focused on enhancing the information delivered to clinicians even further by producing full survival distributions instead of point predictions or fixed horizon risks. In this work, we propose a non-parametric ensemble model, Weighted Resolution Survival Ensemble (WRSE), tailored to estimate such dynamic individual survival distributions. Inspired by the simplicity and robustness of ensemble methods, the proposed approach combines a set of binary classifiers spaced according to a decay function reflecting the relevance of short-term mortality predictions. Models and baselines are evaluated under weighted calibration and discrimination metrics for individual survival distributions which closely reflect the utility of a model in ICU practice. We show competitive results with state-of-the-art probabilistic models, while greatly reducing training time by factors of 2-9x.