Refine
Year of publication
Document Type
- Article (11)
- Part of Periodical (2)
- Monograph/Edited Volume (1)
- Conference Proceeding (1)
- Postprint (1)
Keywords
- Bildband (1)
- Bottom-up effects (1)
- Brandenburg (1)
- COVID-19 (1)
- Carbon cycling (1)
- Clinical predictive modeling (1)
- Cochlea-Implantat (1)
- Denkmal (1)
- Denkmalpflege (1)
- EHR (1)
Institute
- Institut für Biochemie und Biologie (4)
- Institut für Physik und Astronomie (3)
- Extern (2)
- Mathematisch-Naturwissenschaftliche Fakultät (2)
- Department Linguistik (1)
- Department Sport- und Gesundheitswissenschaften (1)
- Hasso-Plattner-Institut für Digital Engineering GmbH (1)
- Hasso-Plattner-Institut für Digital Engineering gGmbH (1)
- Institut für Chemie (1)
- Interdisziplinäres Zentrum für Musterdynamik und Angewandte Fernerkundung (1)
In semi-arid savannas, unsustainable land use can lead to degradation of entire landscapes, e.g. in the form of shrub encroachment. This leads to habitat loss and is assumed to reduce species diversity. In BIOTA phase 1, we investigated the effects of land use on population dynamics on farm scale. In phase 2 we scale up to consider the whole regional landscape consisting of a diverse mosaic of farms with different historic and present land use intensities. This mosaic creates a heterogeneous, dynamic pattern of structural diversity at a large spatial scale. Understanding how the region-wide dynamic land use pattern affects the abundance of animal and plant species requires the integration of processes on large as well as on small spatial scales. In our multidisciplinary approach, we integrate information from remote sensing, genetic and ecological field studies as well as small scale process models in a dynamic region-wide simulation tool. <hr> Interdisziplinäres Zentrum für Musterdynamik und Angewandte Fernerkundung Workshop vom 9. - 10. Februar 2006.
Background Anxiety and depressive disorders share common features of mood dysfunctions. This has stimulated interest in transdiagnostic dimensional research as proposed by the Research Domain Criteria (RDoC) approach by the National Institute of Mental Health (NIMH) aiming to improve the understanding of underlying disease mechanisms. The purpose of this study was to investigate the processing of RDoC domains in relation to disease severity in order to identify latent disorder-specific as well as transdiagnostic indicators of disease severity in patients with anxiety and depressive disorders.
Methods Within the German research network for mental disorders, 895 participants (n = 476 female, n = 602 anxiety disorder, n = 257 depressive disorder) were recruited for the Phenotypic, Diagnostic and Clinical Domain Assessment Network Germany (PD-CAN) and included in this cross-sectional study. We performed incremental regression models to investigate the association of four RDoC domains on disease severity in patients with affective disorders: Positive (PVS) and Negative Valance System (NVS), Cognitive Systems (CS) and Social Processes (SP).
Results The results confirmed a transdiagnostic relationship for all four domains, as we found significant main effects on disease severity within domain-specific models (PVS: & beta; = -0.35; NVS: & beta; = 0.39; CS: & beta; = -0.12; SP: & beta; = -0.32). We also found three significant interaction effects with main diagnosis showing a disease-specific association.
Limitations The cross-sectional study design prevents causal conclusions. Further limitations include possible outliers and heteroskedasticity in all regression models which we appropriately controlled for.
Conclusion Our key results show that symptom burden in anxiety and depressive disorders is associated with latent RDoC indicators in transdiagnostic and disease-specific ways.
Despite advances in machine learning-based clinical prediction models, only few of such models are actually deployed in clinical contexts. Among other reasons, this is due to a lack of validation studies. In this paper, we present and discuss the validation results of a machine learning model for the prediction of acute kidney injury in cardiac surgery patients initially developed on the MIMIC-III dataset when applied to an external cohort of an American research hospital. To help account for the performance differences observed, we utilized interpretability methods based on feature importance, which allowed experts to scrutinize model behavior both at the global and local level, making it possible to gain further insights into why it did not behave as expected on the validation cohort. The knowledge gleaned upon derivation can be potentially useful to assist model update during validation for more generalizable and simpler models. We argue that interpretability methods should be considered by practitioners as a further tool to help explain performance differences and inform model update in validation studies.
Background:
COVID-19 has infected millions of people worldwide and is responsible for several hundred thousand fatalities. The COVID-19 pandemic has necessitated thoughtful resource allocation and early identification of high-risk patients. However, effective methods to meet these needs are lacking.
Objective:
The aims of this study were to analyze the electronic health records (EHRs) of patients who tested positive for COVID-19 and were admitted to hospitals in the Mount Sinai Health System in New York City; to develop machine learning models for making predictions about the hospital course of the patients over clinically meaningful time horizons based on patient characteristics at admission; and to assess the performance of these models at multiple hospitals and time points.
Methods:
We used Extreme Gradient Boosting (XGBoost) and baseline comparator models to predict in-hospital mortality and critical events at time windows of 3, 5, 7, and 10 days from admission. Our study population included harmonized EHR data from five hospitals in New York City for 4098 COVID-19-positive patients admitted from March 15 to May 22, 2020. The models were first trained on patients from a single hospital (n=1514) before or on May 1, externally validated on patients from four other hospitals (n=2201) before or on May 1, and prospectively validated on all patients after May 1 (n=383). Finally, we established model interpretability to identify and rank variables that drive model predictions.
Results:
Upon cross-validation, the XGBoost classifier outperformed baseline models, with an area under the receiver operating characteristic curve (AUC-ROC) for mortality of 0.89 at 3 days, 0.85 at 5 and 7 days, and 0.84 at 10 days. XGBoost also performed well for critical event prediction, with an AUC-ROC of 0.80 at 3 days, 0.79 at 5 days, 0.80 at 7 days, and 0.81 at 10 days. In external validation, XGBoost achieved an AUC-ROC of 0.88 at 3 days, 0.86 at 5 days, 0.86 at 7 days, and 0.84 at 10 days for mortality prediction. Similarly, the unimputed XGBoost model achieved an AUC-ROC of 0.78 at 3 days, 0.79 at 5 days, 0.80 at 7 days, and 0.81 at 10 days. Trends in performance on prospective validation sets were similar. At 7 days, acute kidney injury on admission, elevated LDH, tachypnea, and hyperglycemia were the strongest drivers of critical event prediction, while higher age, anion gap, and C-reactive protein were the strongest drivers of mortality prediction.
Conclusions:
We externally and prospectively trained and validated machine learning models for mortality and critical events for patients with COVID-19 at different time horizons. These models identified at-risk patients and uncovered underlying relationships that predicted outcomes.
In this pilot study, we describe a high-pressure incubation system allowing multiple subsampling of a pressurized culture without decompression. The system was tested using one piezophilic (Photobacterium profundum), one piezotolerant (Colwellia maris) bacterial strain and a decompressed sample from the Mediterranean deep sea (3044 m) determining bacterial community composition, protein production (BPP) and cell multiplication rates (BCM) up to 27 MPa. The results showed elevation of BPP at high pressure was by a factor of 1.5 +/- 1.4 and 3.9 +/- 2.3 for P. profundum and C. maris, respectively, compared to ambient-pressure treatments and by a factor of 6.9 +/- 3.8 fold in the field samples. In P. profundum and C. maris, BCM at high pressure was elevated (3.1 +/- 1.5 and 2.9 +/- 1.7 fold, respectively) compared to the ambient-pressure treatments. After 3 days of incubation at 27 MPa, the natural bacterial deep-sea community was dominated by one phylum of the genus Exiguobacterium, indicating the rapid selection of piezotolerant bacteria. In future studies, our novel incubation system could be part of an isopiestic pressure chain, allowing more accurate measurement of bacterial activity rates which is important both for modeling and for predicting the efficiency of the oceanic carbon pump.
Das Herbsttreffen Patholinguistik wird seit 2007 jährlich vom Verband für Patholinguistik e.V. (vpl) durchgeführt. Das 7. Herbsttreffen mit dem Schwerpunktthema "Hören – Zuhören – Dazugehören: Sprachtherapie bei Hörstörungen und Cochlea-Implantat" fand am 16.11.2013 in Potsdam statt. Der vorliegende Tagungsband beinhaltet die sechs Vorträge zum Schwerpunktthema aus verschiedenen Perspektiven: der medizinischen, der therapeutischen, der wissenschaftlichen sowie der von Betroffenen. Weiterhin sind die Beiträge der Posterpräsentationen zu Themen der sprachtherapeutischen Forschung und Praxis abgedruckt.