Thermodynamically Stable Cationic Dimers in Carboxyl-Functionalized Ionic Liquids: The Paradoxical Case of "Anti-Electrostatic" Hydrogen Bonding.

We show that carboxyl-functionalized ionic liquids (ILs) form doubly hydrogen-bonded cationic dimers (c+=c+) despite the repulsive forces between ions of like charge and competing hydrogen bonds between cation and anion (c+-a-). This structural motif as known for formic acid, the archetype of double hydrogen bridges, is present in the solid state of the IL 1-(carboxymethyl)pyridinium bis(trifluoromethylsulfonyl)imide [HOOC-CH2-py][NTf2]. By means of quantum chemical calculations, we explored different hydrogen-bonded isomers of neutral (HOOC-(CH2)n-py+)2(NTf2-)2, single-charged (HOOC-(CH2)n-py+)2(NTf2-), and double-charged (HOOC- (CH2)n-py+)2 complexes for demonstrating the paradoxical case of "anti-electrostatic" hydrogen bonding (AEHB) between ions of like charge. For the pure doubly hydrogen-bonded cationic dimers (HOOC- (CH2)n-py+)2, we report robust kinetic stability for n = 1-4. At n = 5, hydrogen bonding and dispersion fully compensate for the repulsive Coulomb forces between the cations, allowing for the quantification of the two equivalent hydrogen bonds and dispersion interaction in the order of 58.5 and 11 kJmol-1, respectively. For n = 6-8, we calculated negative free energies for temperatures below 47, 80, and 114 K, respectively. Quantum cluster equilibrium (QCE) theory predicts the equilibria between cationic monomers and dimers by considering the intermolecular interaction between the species, leading to thermodynamic stability at even higher temperatures. We rationalize the H-bond characteristics of the cationic dimers by the natural bond orbital (NBO) approach, emphasizing the strong correlation between NBO-based and spectroscopic descriptors, such as NMR chemical shifts and vibrational frequencies.

[Artificial Intelligence: Infrastructures and Prerequisites at European Level]. / Künstliche Intelligenz: Infrastrukturen und Voraussetzungen auf europäischer Ebene.

The application of artificial intelligence (AI) is often associated with the use of large amounts of data for the construction of AI models and algorithms. This data should ideally comply with the FAIR Data principles, i.e. being findable, accessible, interoperable and reusable. However, the handling of health data poses a particular challenge in this context. In this article, we highlight the challenges of the data usage for AI in medicine using the example of anaesthesia and intensive care medicine. We discuss the current situation but also the obstacles for a wider application of AI in medicine in Europe and give suggestions how to solve the different issues. The article covers different subjects like data protection, research data infrastructures and approval of medical products. Finally, this article shows how it can nevertheless be possible to establish a secure and at the same time effective handling of data for use in AI at the European level despite its unneglectable difficulties.

[Usage of Artificial Intelligence in the Combat against the COVID-19 Pandemic]. / Nutzung von künstlicher Intelligenz zur Bekämpfung der COVID-19-Pandemie.

The COVID-19 pandemic is a global health emergency of historic dimension. In this situation, researchers worldwide wanted to help manage the pandemic by using artificial intelligence (AI). This narrative review aims to describe the usage of AI in the combat against COVID-19. The addressed aspects encompass AI algorithms for analysis of thoracic X-rays or CTs, prediction models for severity and outcome of the disease, AI applications in development of new drugs and vaccines as well as forecasting models for spread of the virus. The review shows, which approaches were pursued, and which were successful.

Biometric covariates and outcome in COVID-19 patients: are we looking close enough?

BACKGROUND: The impact of biometric covariates on risk for adverse outcomes of COVID-19 disease was assessed by numerous observational studies on unstratified cohorts, which show great heterogeneity. However, multilevel evaluations to find possible complex, e.g. non-monotonic multi-variate patterns reflecting mutual interference of parameters are missing. We used a more detailed, computational analysis to investigate the influence of biometric differences on mortality and disease evolution among severely ill COVID-19 patients. METHODS: We analyzed a group of COVID-19 patients requiring Intensive care unit (ICU) treatment. For further analysis, the study group was segmented into six subgroups according to Body mass index (BMI) and age. To link the BMI/age derived subgroups with risk factors, we performed an enrichment analysis of diagnostic parameters and comorbidities. To suppress spurious patterns, multiple segmentations were analyzed and integrated into a consensus score for each analysis step. RESULTS: We analyzed 81 COVID-19 patients, of whom 67 required mechanical ventilation (MV). Mean mortality was 35.8%. We found a complex, non-monotonic interaction between age, BMI and mortality. A subcohort of patients with younger age and intermediate BMI exhibited a strongly reduced mortality risk (p < 0.001), while differences in all other groups were not significant. Univariate impacts of BMI or age on mortality were missing. Comparing MV with non-MV patients, we found an enrichment of baseline CRP, PCT and D-Dimers within the MV group, but not when comparing survivors vs. non-survivors within the MV patient group. CONCLUSIONS: The aim of this study was to get a more detailed insight into the influence of biometric covariates on the outcome of COVID-19 patients with high degree of severity. We found that survival in MV is affected by complex interactions of covariates differing to the reported covariates, which are hidden in generic, non-stratified studies on risk factors. Hence, our study suggests that a detailed, multivariate pattern analysis on larger patient cohorts reflecting the specific disease stages might reveal more specific patterns of risk factors supporting individually adapted treatment strategies.

Future Medical Artificial Intelligence Application Requirements and Expectations of Physicians in German University Hospitals: Web-Based Survey.

BACKGROUND: The increasing development of artificial intelligence (AI) systems in medicine driven by researchers and entrepreneurs goes along with enormous expectations for medical care advancement. AI might change the clinical practice of physicians from almost all medical disciplines and in most areas of health care. While expectations for AI in medicine are high, practical implementations of AI for clinical practice are still scarce in Germany. Moreover, physicians' requirements and expectations of AI in medicine and their opinion on the usage of anonymized patient data for clinical and biomedical research have not been investigated widely in German university hospitals. OBJECTIVE: This study aimed to evaluate physicians' requirements and expectations of AI in medicine and their opinion on the secondary usage of patient data for (bio)medical research (eg, for the development of machine learning algorithms) in university hospitals in Germany. METHODS: A web-based survey was conducted addressing physicians of all medical disciplines in 8 German university hospitals. Answers were given using Likert scales and general demographic responses. Physicians were asked to participate locally via email in the respective hospitals. RESULTS: The online survey was completed by 303 physicians (female: 121/303, 39.9%; male: 173/303, 57.1%; no response: 9/303, 3.0%) from a wide range of medical disciplines and work experience levels. Most respondents either had a positive (130/303, 42.9%) or a very positive attitude (82/303, 27.1%) towards AI in medicine. There was a significant association between the personal rating of AI in medicine and the self-reported technical affinity level (H4=48.3, P<.001). A vast majority of physicians expected the future of medicine to be a mix of human and artificial intelligence (273/303, 90.1%) but also requested a scientific evaluation before the routine implementation of AI-based systems (276/303, 91.1%). Physicians were most optimistic that AI applications would identify drug interactions (280/303, 92.4%) to improve patient care substantially but were quite reserved regarding AI-supported diagnosis of psychiatric diseases (62/303, 20.5%). Of the respondents, 82.5% (250/303) agreed that there should be open access to anonymized patient databases for medical and biomedical research. CONCLUSIONS: Physicians in stationary patient care in German university hospitals show a generally positive attitude towards using most AI applications in medicine. Along with this optimism comes several expectations and hopes that AI will assist physicians in clinical decision making. Especially in fields of medicine where huge amounts of data are processed (eg, imaging procedures in radiology and pathology) or data are collected continuously (eg, cardiology and intensive care medicine), physicians' expectations of AI to substantially improve future patient care are high. In the study, the greatest potential was seen in the application of AI for the identification of drug interactions, assumedly due to the rising complexity of drug administration to polymorbid, polypharmacy patients. However, for the practical usage of AI in health care, regulatory and organizational challenges still have to be mastered.

Cyclic Octamer of Hydroxyl-functionalized Cations with Net Charge Q=+8e Kinetically Stabilized by a 'Molecular Island' of Cooperative Hydrogen Bonds.

Cyclic octamers are well-known structural motifs in chemistry, biology and physics. These include covalently bound cyclic octameric sulphur, cylic octa-alkanes, cyclo-octameric peptides as well as hydrogen-bonded ring clusters of alcohols. In this work, we show that even calculated cyclic octamers of hydroxy-functionalized pyridinium cations with a net charge Q=+8e are kinetically stable. Eight positively charged cations are kept together by hydrogen bonding despite the strong Coulomb repulsive forces. Sufficiently long hydroxy-octyl chains prevent "Coulomb explosion" by increasing the distance between the positive charges at the pyridinium rings, reducing the Coulomb repulsion and thus strengthen hydrogen bonds between the OH groups. The eightfold positively charged cyclic octamer shows spectroscopic properties similar to those obtained for hydrogen-bonded neutral cyclic octamers of methanol. Thus, the area of the hydrogen bonded OH ring represents a 'molecular island' within an overall cationic environment. Although not observable, the spectroscopic properties and the correlated NBO parameters of the calculated cationic octamer support the detection of smaller cationic clusters in ionic liquids, which we observed despite the competition with ion pairs wherein attractive Coulomb forces enhance hydrogen bonding between cation and anion.

Future Mobile Device Usage, Requirements, and Expectations of Physicians in German University Hospitals: Web-Based Survey.

BACKGROUND: The use of mobile devices in hospital care constantly increases. However, smartphones and tablets have not yet widely become official working equipment in medical care. Meanwhile, the parallel use of private and official devices in hospitals is common. Medical staff use smartphones and tablets in a growing number of ways. This mixture of devices and how they can be used is a challenge to persons in charge of defining strategies and rules for the usage of mobile devices in hospital care. OBJECTIVE: Therefore, we aimed to examine the status quo of physicians' mobile device usage and concrete requirements and their future expectations of how mobile devices can be used. METHODS: We performed a web-based survey among physicians in 8 German university hospitals from June to October 2019. The online survey was forwarded by hospital management personnel to physicians from all departments involved in patient care at the local sites. RESULTS: A total of 303 physicians from almost all medical fields and work experience levels completed the web-based survey. The majority regarded a tablet (211/303, 69.6%) and a smartphone (177/303, 58.4%) as the ideal devices for their operational area. In practice, physicians are still predominantly using desktop computers during their worktime (mean percentage of worktime spent on a desktop computer: 56.8%; smartphone: 12.8%; tablet: 3.6%). Today, physicians use mobile devices for basic tasks such as oral (171/303, 56.4%) and written (118/303, 38.9%) communication and to look up dosages, diagnoses, and guidelines (194/303, 64.0%). Respondents are also willing to use mobile devices for more advanced applications such as an early warning system (224/303, 73.9%) and mobile electronic health records (211/303, 69.6%). We found a significant association between the technical affinity and the preference of device in medical care (χs2=53.84, P<.001) showing that with increasing self-reported technical affinity, the preference for smartphones and tablets increases compared to desktop computers. CONCLUSIONS: Physicians in German university hospitals have a high technical affinity and positive attitude toward the widespread implementation of mobile devices in clinical care. They are willing to use official mobile devices in clinical practice for basic and advanced mobile health uses. Thus, the reason for the low usage is not a lack of willingness of the potential users. Challenges that hinder the wider adoption of mobile devices might be regulatory, financial and organizational issues, and missing interoperability standards of clinical information systems, but also a shortage of areas of application in which workflows are adapted for (small) mobile devices.

[Respiratory Insufficiency: State of the Art - Diagnosis and Therapy]. / Respiratorisches Versagen: State of the Art ­ Diagnose und Therapie.

The term "respiratory insufficiency" (RI) describes the inability of an organism to maintain the gas exchange between the ambient air and its peripheral organs. This causes a hypoxia and hypercapnia. The mechanisms that lead to RI are either an impaired gas exchange in the lung tissue or an alveolar hypoventilation caused by an insufficient ventilatory pump. Thus the RI is divided into a hypoxic or a hypercapnic RI. The diagnostic procedure for RI contains several examinations. One key aspect of the exploration is the identification of potential reversible and thus correctable reasons of the RI. The goal of the therapy is to maintain the oxygen supply for the peripheral organs and the elimination of CO2. It covers supportive and causal therapeutic interventions. Wherever possible, therapy should resolve the cause of the RI. The non-invasive ventilation (NIV) is the therapy of choice in severe cases of a hypercapnic RI. It relieves the exhausted respiratory muscles and improves respiratory situation. It can be used for hypoxic RI as well, but frequently invasive ventilation is required. If NIV is not able to improve the patient's condition, invasive ventilation (IV) is applied.

Development of a machine learning model for prediction of the duration of unassisted spontaneous breathing in patients during prolonged weaning from mechanical ventilation.

PURPOSE: Treatment of patients undergoing prolonged weaning from mechanical ventilation includes repeated spontaneous breathing trials (SBTs) without respiratory support, whose duration must be balanced critically to prevent over- and underload of respiratory musculature. This study aimed to develop a machine learning model to predict the duration of unassisted spontaneous breathing. MATERIALS AND METHODS: Structured clinical data of patients from a specialized weaning unit were used to develop (1) a classifier model to qualitatively predict an increase of duration, (2) a regressor model to quantitatively predict the precise duration of SBTs on the next day, and (3) the duration difference between the current and following day. 61 features, known to influence weaning, were included into a Histogram-based gradient boosting model. The models were trained and evaluated using separated data sets. RESULTS: 18.948 patient-days from 1018 individual patients were included. The classifier model yielded an ROC-AUC of 0.713. The regressor models displayed a mean absolute error of 2:50 h for prediction of absolute durations and 2:47 h for day-to-day difference. CONCLUSIONS: The developed machine learning model showed informed results when predicting the spontaneous breathing capacity of a patient in prolonged weaning, however lacking prognostic quality required for direct translation to clinical use.

Diaphragm ultrasound in patients with prolonged weaning from mechanical ventilation.

Background: Several publications have examined diaphragmatic ultrasound using two-dimensional (2D) parameters in the context of weaning from mechanical ventilation (MV) and extubation. However, the studied cohorts had rather short duration of ventilation. Examinations on patients with prolonged weaning after long-term ventilation were missing. It was the aim of this study to assess of the diaphragm and peripheral musculature of patients undergoing prolonged weaning creating a chronological sequence of ultrasonic parameters during the course of weaning. Methods: This study was carried out as a monocentric, prospective observational cross-sectional study. Patients in prolonged weaning who were transferred to a specialized weaning unit were eligible for inclusion if they were ventilated invasively by means of an endotracheal tube or tracheal cannula and if their expected treatment period was at least 5 days. Diaphragmatic function and one representative peripheral muscle were examined in 50 patients between March 2020 and April 2021. The 2D sonographic parameters of diaphragm and diaphragmatic function consisted of diaphragmatic thickness (Tdi) at the end of inspiration and expiration, the fractional thickening (FT) and the diaphragmatic excursion. Additionally, the M. quadriceps femoris was sonographically assessed at two locations. The difference of measurements between the first and the last measuring timepoint were examined using the Wilcoxon signed-rank test. For a longer chronological sequence, the Friedman's rank sum test with subsequent Wilcoxon-Nemenyi-McDonald-Thompson test for multiple comparisons was carried out. Results: Fifty patients with prolonged weaning were included. The median duration of MV before transfer to the weaning unit was 11.5 [interquartile range (IQR) 10] days. Forty-one patients could be assessed over the full course of weaning, with 38 successfully weaned. Within these 41 patients, the sonographic parameters of the diaphragm slightly increased over the course of weaning indicating an increase in thickness and mobility. Especially parameters which represented an active movement reached statistical significance, i.e., inspiratory Tdi when assessed under spontaneous breathing [begin 3.41 (0.99) vs. end 3.43 (1.31) mm; P=0.01] and diaphragmatic excursion [begin 0.7 (0.8) vs. end 0.9 (0.6) cm; P=0.01]. The presence of positive end-expiratory pressure (PEEP) and pressure support did not influence the sonographic parameters significantly. The M. quadriceps femoris, in contrast, decreased slightly but constantly over the time [lower third: begin 1.36 (0.48) vs. end 1.28 (0.36) cm; P=0.054]. Conclusions: The present study is the first one to longitudinally analyse diaphragmatic ultrasound in patients with prolonged weaning. Sonographic assessment showed that Tdi and excursion increased over the course of prolonged weaning, while the diameter of a representative peripheral muscle decreased. However, the changes are rather small, and data show a wide dispersion. To allow a potential, standardized use of diaphragm ultrasound for diagnostic decision support in prolonged weaning, further studies in this specific patient group are required.

A hybrid modeling framework for generalizable and interpretable predictions of ICU mortality across multiple hospitals.

The development of reliable mortality risk stratification models is an active research area in computational healthcare. Mortality risk stratification provides a standard to assist physicians in evaluating a patient's condition or prognosis objectively. Particular interest lies in methods that are transparent to clinical interpretation and that retain predictive power once validated across diverse datasets they were not trained on. This study addresses the challenge of consolidating numerous ICD codes for predictive modeling of ICU mortality, employing a hybrid modeling approach that integrates mechanistic, clinical knowledge with mathematical and machine learning models . A tree-structured network connecting independent modules that carry clinical meaning is implemented for interpretability. Our training strategy utilizes graph-theoretic methods for data analysis, aiming to identify the functions of individual black-box modules within the tree-structured network by harnessing solutions from specific max-cut problems. The trained model is then validated on external datasets from different hospitals, demonstrating successful generalization capabilities, particularly in binary-feature datasets where label assessment involves extrapolation.

Effects of structured protocolized physical therapy on the duration of mechanical ventilation in patients with prolonged weaning.

PURPOSE: 20% of patients with mechanical ventilation (MV) have a prolonged, complex weaning process, often experiencing a condition of ICU-acquired weakness (ICUAW), with a severe decrease in muscle function and restricted long-term prognosis. We aimed to analyze a protocolized, systematic approach of physiotherapy in prolonged weaning patients and hypothesized that the duration of weaning from MV would be shortened. METHODS: ICU patients with prolonged weaning were included before (group 1) and after (group 2) introduction of a quality control measure of a structured and protocolized physiotherapy program. Primary endpoint was the tested dynamometric handgrip strength and the Surgical Intensive Care Unit Optimal Mobilization Score (SOMS). Secondary endpoints were weaning success rate, ventilator-free days, hospital mortality, the prevalence of ICUAW, infections and delirium. RESULTS: 106 patients were included. Both the SOMS and the handgrip test were significantly improved after introducing the program. Despite no differences in weaning success rates at discharge, the total length of MV was significantly shorter in group 2, which also had lower prevalence of infection and higher probability of survival. CONCLUSIONS: Protocolized, systematic physiotherapy resulted in an improvement of the clinical outcome in patients with prolonged weaning. Results were objectifiable with the SOMS and the handgrip test.

Spectroscopic Evidence for Doubly Hydrogen-Bonded Cationic Dimers in the Solid, Liquid, and Gaseous Phases of Carboxyl-Functionalized Ionic Liquids.

Intermolecular interactions determine whether matter sticks together, gases condense into liquids, or liquids freeze into solids. The most prominent example is hydrogen bonding in water, responsible for the anomalous properties in the liquid phase and polymorphism in ice. The physical properties are also exceptional for ionic liquids (ILs), wherein a delicate balance of Coulomb interactions, hydrogen bonds, and dispersion interactions results in a broad liquid range and the vaporization of ILs as ion pairs. In this study, we show that strong, local, and directional hydrogen bonds govern the structures and arrangements in the solid, liquid, and gaseous phases of carboxyl-functionalized ILs. For that purpose, we explored the H-bonded motifs by X-ray diffraction and attenuated total reflection (ATR) infrared (IR) spectroscopy in the solid state, by ATR and transmission IR spectroscopy in the liquid phase, and by cryogenic ion vibrational predissociation spectroscopy (CIVPS) in the gaseous phase at low temperature. The analysis of the CO stretching bands reveals doubly hydrogen-bonded cationic dimers (c═c), resembling the archetype H-bond motif known for carboxylic acids. The like-charge doubly hydrogen-bonded ion pairs are present in the crystal structure of the IL, survive phase transition into the liquid state, and are still present in the gaseous phase even in (2,1) complexes wherein one counterion is removed and repulsive Coulomb interaction increased. The interpretation of the vibrational spectra is supported by quantum chemical methods. These observations have implications for the fundamental nature of the hydrogen bond between ions of like charge.

Computational Simulation of Virtual Patients Reduces Dataset Bias and Improves Machine Learning-Based Detection of ARDS from Noisy Heterogeneous ICU Datasets.

Goal: Machine learning (ML) technologies that leverage large-scale patient data are promising tools predicting disease evolution in individual patients. However, the limited generalizability of ML models developed on single-center datasets, and their unproven performance in real-world settings, remain significant constraints to their widespread adoption in clinical practice. One approach to tackle this issue is to base learning on large multi-center datasets. However, such heterogeneous datasets can introduce further biases driven by data origin, as data structures and patient cohorts may differ between hospitals. Methods: In this paper, we demonstrate how mechanistic virtual patient (VP) modeling can be used to capture specific features of patients' states and dynamics, while reducing biases introduced by heterogeneous datasets. We show how VP modeling can be used for data augmentation through identification of individualized model parameters approximating disease states of patients with suspected acute respiratory distress syndrome (ARDS) from observational data of mixed origin. We compare the results of an unsupervised learning method (clustering) in two cases: where the learning is based on original patient data and on data derived in the matching procedure of the VP model to real patient data. Results: More robust cluster configurations were observed in clustering using the model-derived data. VP model-based clustering also reduced biases introduced by the inclusion of data from different hospitals and was able to discover an additional cluster with significant ARDS enrichment. Conclusions: Our results indicate that mechanistic VP modeling can be used to significantly reduce biases introduced by learning from heterogeneous datasets and to allow improved discovery of patient cohorts driven exclusively by medical conditions.

Hamiltonian adaptive resolution simulation for molecular liquids.

Adaptive resolution schemes allow the simulation of a molecular fluid treating simultaneously different subregions of the system at different levels of resolution. In this work we present a new scheme formulated in terms of a global Hamiltonian. Within this approach equilibrium states corresponding to well-defined statistical ensembles can be generated making use of all standard molecular dynamics or Monte Carlo methods. Models at different resolutions can thus be coupled, and thermodynamic equilibrium can be modulated keeping each region at desired pressure or density without disrupting the Hamiltonian framework.

Analysis of Chest X-ray for COVID-19 Diagnosis as a Use Case for an HPC-Enabled Data Analysis and Machine Learning Platform for Medical Diagnosis Support.

The COVID-19 pandemic shed light on the need for quick diagnosis tools in healthcare, leading to the development of several algorithmic models for disease detection. Though these models are relatively easy to build, their training requires a lot of data, storage, and resources, which may not be available for use by medical institutions or could be beyond the skillset of the people who most need these tools. This paper describes a data analysis and machine learning platform that takes advantage of high-performance computing infrastructure for medical diagnosis support applications. This platform is validated by re-training a previously published deep learning model (COVID-Net) on new data, where it is shown that the performance of the model is improved through large-scale hyperparameter optimisation that uncovered optimal training parameter combinations. The per-class accuracy of the model, especially for COVID-19 and pneumonia, is higher when using the tuned hyperparameters (healthy: 96.5%; pneumonia: 61.5%; COVID-19: 78.9%) as opposed to parameters chosen through traditional methods (healthy: 93.6%; pneumonia: 46.1%; COVID-19: 76.3%). Furthermore, training speed-up analysis shows a major decrease in training time as resources increase, from 207 min using 1 node to 54 min when distributed over 32 nodes, but highlights the presence of a cut-off point where the communication overhead begins to affect performance. The developed platform is intended to provide the medical field with a technical environment for developing novel portable artificial-intelligence-based tools for diagnosis support.

Novel architecture for gated recurrent unit autoencoder trained on time series from electronic health records enables detection of ICU patient subgroups.

Electronic health records (EHRs) are used in hospitals to store diagnoses, clinician notes, examinations, lab results, and interventions for each patient. Grouping patients into distinct subsets, for example, via clustering, may enable the discovery of unknown disease patterns or comorbidities, which could eventually lead to better treatment through personalized medicine. Patient data derived from EHRs is heterogeneous and temporally irregular. Therefore, traditional machine learning methods like PCA are ill-suited for analysis of EHR-derived patient data. We propose to address these issues with a new methodology based on training a gated recurrent unit (GRU) autoencoder directly on health record data. Our method learns a low-dimensional feature space by training on patient data time series, where the time of each data point is expressed explicitly. We use positional encodings for time, allowing our model to better handle the temporal irregularity of the data. We apply our method to data from the Medical Information Mart for Intensive Care (MIMIC-III). Using our data-derived feature space, we can cluster patients into groups representing major classes of disease patterns. Additionally, we show that our feature space exhibits a rich substructure at multiple scales.

Developing an Artificial Intelligence-Based Representation of a Virtual Patient Model for Real-Time Diagnosis of Acute Respiratory Distress Syndrome.

Acute Respiratory Distress Syndrome (ARDS) is a condition that endangers the lives of many Intensive Care Unit patients through gradual reduction of lung function. Due to its heterogeneity, this condition has been difficult to diagnose and treat, although it has been the subject of continuous research, leading to the development of several tools for modeling disease progression on the one hand, and guidelines for diagnosis on the other, mainly the "Berlin Definition". This paper describes the development of a deep learning-based surrogate model of one such tool for modeling ARDS onset in a virtual patient: the Nottingham Physiology Simulator. The model-development process takes advantage of current machine learning and data-analysis techniques, as well as efficient hyperparameter-tuning methods, within a high-performance computing-enabled data science platform. The lightweight models developed through this process present comparable accuracy to the original simulator (per-parameter R2 > 0.90). The experimental process described herein serves as a proof of concept for the rapid development and dissemination of specialised diagnosis support systems based on pre-existing generalised mechanistic models, making use of supercomputing infrastructure for the development and testing processes and supported by open-source software for streamlined implementation in clinical routines.

[Weaning from invasive ventilation : Challenges in the clinical routine]. / Weaning von invasiver Beatmung : Herausforderungen im klinischen Alltag.

Modern intensive care medicine is caught between the conflicting demands of an efficient but also increasingly more technical intensive care treatment with numerous therapeutic options and, at the same time, an ageing society with increasing morbidity. This is reflected, among other things, in an increasing number of ventilated patients in intensive care units and an increasing proportion of patients for whom ventilation cannot easily be discontinued. Weaning from a ventilator, which can account for more than 50% of the total ventilation time, therefore plays a central role in this process. This main topic article presents the need for strategically wise and holistic actions to minimize the consequences of invasive mechanical ventilation for patients. An attempt is made to shed more light on individual aspects of the ventilation weaning process with high relevance for clinical practice. Especially for prolonged weaning from ventilation, many more concepts are needed than simply ending ventilation.

A training strategy for hybrid models to break the curse of dimensionality.

Mechanistic/data-driven hybrid modeling is a key approach when the mechanistic details of the processes at hand are not sufficiently well understood, but also inferring a model purely from data is too complex. By the integration of first principles into a data-driven approach, hybrid modeling promises a feasible data demand alongside extrapolation. In this work, we introduce a learning strategy for tree-structured hybrid models to perform a binary classification task. Given a set of binary labeled data, the challenge is to use them to develop a model that accurately assesses labels of new unlabeled data. Our strategy employs graph-theoretic methods to analyze the data and deduce a function that maps input features to output labels. Our focus here is on data sets represented by binary features in which the label assessment of unlabeled data points is always extrapolation. Our strategy shows the existence of small sets of data points within given binary data for which knowing the labels allows for extrapolation to the entire valid input space. An implementation of our strategy yields a notable reduction of training-data demand in a binary classification task compared with different supervised machine learning algorithms. As an application, we have fitted a tree-structured hybrid model to the vital status of a cohort of COVID-19 patients requiring intensive-care unit treatment and mechanical ventilation. Our learning strategy yields the existence of patient cohorts for whom knowing the vital status enables extrapolation to the entire valid input space of the developed hybrid model.