Dräger (Germany)

HighlightsWhat are the main findings?Combining domain-generalized pre-training with test-time adaptation substantially enhances segmentation accuracy for medical images in out-of-domain scenarios.Generalizing image descriptors and intensity augmentation during training and adaptation outperformed complex existing methods in several scenarios.What is the implication of the main finding?We provide a powerful tool to achieve high segmentation accuracy out-of-domain, even without access to source data.Our efficient method is applicable to many top public 2D and 3D segmentation models trained on large, private datasets.Applying pre-trained medical deep learning segmentation models to out-of-domain images often yields predictions of insufficient quality. In this study, we propose using a robust generalizing descriptor, along with augmentation, to enable domain-generalized pre-training and test-time adaptation, thereby achieving high-quality segmentation in unseen domains. In this study, five different publicly available datasets, including 3D CT and MRI images, are used to evaluate segmentation performance in out-of-domain scenarios. The settings include abdominal, spine, and cardiac imaging. Domain-generalized pre-training on source data is used to obtain the best initial performance in the target domain. We introduce a combination of the generalizing SSC descriptor and GIN intensity augmentation for optimal generalization. Segmentation results are subsequently optimized at test time, where we propose adapting the pre-trained models for every unseen scan using a consistency scheme with the augmentation–descriptor combination. The proposed generalized pre-training and subsequent test-time adaptation improve model performance significantly in CT to MRI cross-domain prediction for abdominal (+46.2 and +28.2 Dice), spine (+72.9), and cardiac (+14.2 and +55.7 Dice) scenarios (p < 0.001). Our method enables the optimal, independent use of source and target data, successfully bridging domain gaps with a compact and efficient methodology.

The efficient direct integration of real-time medical device data is a promising approach to improve patient care enabling a direct and eminent intervention. This study presents a comprehensive approach for integrating real-time medical device data into clinical environments using the HL7® FHIR® standards and IEEE 11073 Service-Oriented Device Connectivity (SDC). The study proposes a conceptual framework and an opensource proof-of-concept implementation for real-time data integration within the Medical Data Integration Center (MeDIC) at UKSH. Key components include a selective recording mechanism to mitigate storage issues and ensure accurate data capture. Our robust network architecture utilizes Kafka brokers for seamless data transfer in isolated networks. The study demonstrates the selective capturing of real-time data within a clinical setting to enable medical device data for a down-stream processing and analysis.

We conducted a field study at a music festival to examine effects of naturally occurring sleep deficits and substance intoxication on false memory formation and susceptibility to suggestion, using two paradigms. In a misinformation task, sleep restriction was associated with increased levels of susceptibility to suggestion and false alarms to nonsuggestive questions for a virtual reality eyewitness scenario. Use of tetrahydrocannabinol and amphetamines was also related to increased false alarms to nonsuggestive questions, indicating that such questions might be sensitive to use of those drugs. In an implantation paradigm, neither sleep restriction, substances, nor exposure to fake social media content increased the likelihood of falsely believing or recollecting a purportedly experienced festival event. However, some people came to falsely believe (13%, n = 30) or falsely remember (6%, n = 14) the false suggested event. Findings indicate that some established lab effects can also be observed in a less controlled environment.

Health data collected by wearables and apps can be useful as part of patient-generated health data (PGHD) or personal health data for medical diagnosis or general health monitoring. Mobile health apps are more and more accepted, generate evidence and might be increasingly used in personal medicine. Data retrieved from wearables and apps are mostly not following a medical data standard and cannot be retrieved from the vendors in a straightforward way. The present work started the implementation of a Digital Health Convener and described the process to collect data from several wearables – starting with Fitbit data – and transforms this data to standardized JSON files following the Open mHealth (OmH) IEEE and the HL7 FHIR standard. The project achieved is provided as open source and can be extended and used in future projects to generate OmH and FHIR conform PGHD.

Hintergrund Derzeit werden Eltern schon im Rahmen der U2 über das Risiko des plötzlichen Kindstodes und dessen vermeidbare Risikofaktoren wie u.a. die Bauchlage (BL) aufgeklärt. Während der postnatalen Zeit im Krankenhaus kommt die BL in der klinischen Versorgung Frühgeborener (FG) allerdings häufig bis ca. 1 Woche vor Entlassung zum Einsatz. Diese Studie soll genau diesen Übergang von der aus respiratorischen Gründen häufig genutzten BL während des Krankenhausaufenthalts zu der für die Häuslichkeit empfohlenen Rückenlage (RL) genauer untersuchen. In einer Vorarbeit konnte gezeigt werden, dass mit Hilfe eines Bewegungssensors typische Atemmuster während des Schlafs Neugeborener detektiert werden können.

Die vorliegende Dissertation umfaßt die röntgenstrukturanalytische Untersuchung der Antibiotika Mersacidin, Actagardin und Ala(0)-Actagardin, die eine eigenständige Untergruppe der sogenannten Lantibiotika des Typs B bilden. Lantibiotika sind Oligopeptide, deren charakteristisches Merkmal schwefelhaltige Ringstrukturen innerhalb der Peptidkette sind. Hierbei handelt es sich um zyklische Thioether, die eine Klasse von Nicht-Standardaminosäuren, die sogenannten Lanthionine, enthalten. Der hohe Anteil an ungewöhnlichen Aminosäuren mit zum Teil invertierter Konfiguration der stereogenen Alpha-Kohlenstoffatome macht die Lantibiotika zu interessanten neuen Therapeutika im Kampf gegen bakterielle Infektionen.Erstmals konnte die zuvor nur hypothetische Existenz einer Kationbindungstasche in den Lantibiotika durch Kristallisation und Strukturaufklärung eines Calciumkomplexes von Actagardin bewiesen werden. Anhand der Ergebnisse dieser Arbeit wird in Kombination mit in der Literatur veröffentlichter biochemischer Daten ein möglicher Wirkmechanismus der antibiotischen Aktivität der Lantibiotika gegenüber Gram-positiven Bakterien vorgeschlagen.

Das Design eines Produktes gibt immer Hinweise auf seine Funktionalität, was es kann oder nicht kann, was es aushält oder nicht aushält. Die menschliche Wahrnehmung solcher Eigenschaften und Qualitäten kann daher wesentlich dafür sein, wie das Produkt gehandhabt und benutzt wird. Qualität wird durch formale Eigenschaften, Form und Detailgestaltung wahrgenommen. Diese muss mit der tatsächlichen Robustheit eines Produktes übereinstimmen, um den beabsichtigten Gebrauch und Einsatz des Produktes zu provozieren. Ein Missverhältnis zwischen erlebter und tatsächlicher Qualität eines Produkts kann zu Ausfällen, hohen Wartungskosten, geringerer Benutzerzufriedenheit oder geringerem Markenwert führen. In der (interdisziplinären) kollaborativen Produktentwicklung basieren die Anforderungen und die Produktbewertung meist auf quantitativen Maßen. Während die Definition und Bewertung der tatsächlichen Qualität und Robustheit in der Produktentwicklung gut etabliert ist, fehlt noch eine standardisierte Methode zur Definition und Bewertung der erlebten Qualität eines Produkts. In diesem Beitrag wird ein Ansatz vorgestellt, durch den die Übereinstimmung zwischen tatsächlicher und erlebter Qualität durch eine objektive Skala festgelegt bzw. kontrolliert werden kann. Dazu wurde eine semantisch-differentielle Skala entwickelt, die relevante Aspekte der haptischen, visuellen und akustischen Eigenschaften von Produkten erfasst. Es kann sowohl in der Anforderungsdefinition, als auch in den Evaluationsphasen und der Produktentwicklung des Designprozesses eingesetzt werden. Dadurch kann kontrolliert werden, ob die Designziele erreicht wurden oder nicht.

BackgroundIn acute respiratory distress syndrome (ARDS), non-ventilated perfused regions coexist with non-perfused ventilated regions within lungs. The number of unmatched regions might reflect ARDS severity and affect the risk of ventilation-induced lung injury. Despite pathophysiological relevance, unmatched ventilation and perfusion are not routinely assessed at the bedside. The aims of this study were to quantify unmatched ventilation and perfusion at the bedside by electrical impedance tomography (EIT) investigating their association with mortality in patients with ARDS and to explore the effects of positive end-expiratory pressure (PEEP) on unmatched ventilation and perfusion in subgroups of patients with different ARDS severity based on PaO2/FiO2 and compliance.MethodsProspective observational study in 50 patients with mild (36%), moderate (46%), and severe (18%) ARDS under clinical ventilation settings. EIT was applied to measure the regional distribution of ventilation and perfusion using central venous bolus of saline 5% during end-inspiratory pause. We defined unmatched units as the percentage of only ventilated units plus the percentage of only perfused units.ResultsPercentage of unmatched units was significantly higher in non-survivors compared to survivors (32[27–47]% vs. 21[17–27]%, p < 0.001). Percentage of unmatched units was an independent predictor of mortality (OR 1.22, 95% CI 1.07–1.39, p = 0.004) with an area under the ROC curve of 0.88 (95% CI 0.79–0.97, p < 0.001). The percentage of ventilation to the ventral region of the lung was higher than the percentage of ventilation to the dorsal region (32 [27–38]% vs. 18 [13–21]%, p < 0.001), while the opposite was true for perfusion (28 [22–38]% vs. 36 [32–44]%, p < 0.001).Higher percentage of only perfused units was correlated with lower dorsal ventilation (r = − 0.486, p < 0.001) and with lower PaO2/FiO2 ratio (r = − 0.293, p = 0.039).ConclusionsEIT allows bedside assessment of unmatched ventilation and perfusion in mechanically ventilated patients with ARDS. Measurement of unmatched units could identify patients at higher risk of death and could guide personalized treatment.

Abstract Respiratory support is a key element of modern medical care, ranging from oxygen therapy to full ventilatory support. A central component of mechanical ventilation is the control of the resulting pneumatic quantities such as pressure and flow. In this article the use of robust model predictive control for pressure-controlled mechanical ventilation is proposed, with the goal of increasing the safety of the patient by considering physiological safety constraints. The uncertainty in the estimation of physiological model parameters as well as model uncertainties are considered as disturbances to the system, which are taken into account through the proposed robust model predictive control framework. The practical applicability of this control approach is illustrated in an implementation on a research demonstrator of the ventilation unit from an anaesthesia workstation.

The continuous suspension (ConSus) technology is a new coming approach that promises performance advantages over industrially used abrasive waterjet (AWJ) systems in production scenarios. However, it currently lacks practical experience and process parametrization is typically still empirical of either using a 2-phase abrasive suspension jet (ASJ) or a 3-phase injection AWJ tool. A practical comparison of respective AWJ tool characteristics of prior work piece contact is needed for fundamental parametrization. An experimental approach was used to compare the respective AWJ systems. The abrasive material was compared before and after going through either an injection AWJ head or a ConSus ASJ system. Typical abrasive materials and mesh sizes for AWJ machining were systematically investigated. This paper shows that current AWJ cutting systems yield different results when using equal abrasive material. ConSus shows almost no significant effect on abrasive size distribution while commonly used injector systems shifts a tremendous portion of the original abrasive to smaller grain sizes. Therefore, the results enhance process understanding and revision of existing process models for future applications.