Refine
Year of publication
- 2018 (172) (remove)
Institute
- Fachbereich Medizintechnik und Technomathematik (60)
- IfB - Institut für Bioengineering (36)
- Fachbereich Elektrotechnik und Informationstechnik (35)
- INB - Institut für Nano- und Biotechnologien (23)
- Fachbereich Luft- und Raumfahrttechnik (19)
- Fachbereich Chemie und Biotechnologie (18)
- Fachbereich Energietechnik (15)
- Fachbereich Maschinenbau und Mechatronik (10)
- Fachbereich Bauingenieurwesen (8)
- Solar-Institut Jülich (4)
Language
- English (172) (remove)
Document Type
- Article (84)
- Conference Proceeding (65)
- Part of a Book (15)
- Book (3)
- Doctoral Thesis (2)
- Working Paper (2)
- Conference Poster (1)
Keywords
- Energy efficiency (2)
- Engineering optimization (2)
- MINLP (2)
- Pump System (2)
- Serious Game (2)
- Water (2)
- Agility (1)
- Antarctica (1)
- Awareness (1)
- Bioeconomy (1)
Sleep scoring is a necessary and time-consuming task in sleep studies. In animal models (such as mice) or in humans, automating this tedious process promises to facilitate long-term studies and to promote sleep biology as a data-driven f ield. We introduce a deep neural network model that is able to predict different states of consciousness (Wake, Non-REM, REM) in mice from EEG and EMG recordings with excellent scoring results for out-of-sample data. Predictions are made on epochs of 4 seconds length, and epochs are classified as artifactfree or not. The model architecture draws on recent advances in deep learning and in convolutional neural networks research. In contrast to previous approaches towards automated sleep scoring, our model does not rely on manually defined features of the data but learns predictive features automatically. We expect deep learning models like ours to become widely applied in different fields, automating many repetitive cognitive tasks that were previously difficult to tackle.
For pelvic floor disorders that cannot be treated with non-surgical procedures, minimally invasive surgery has become a more frequent and safer repair procedure. More than 20 million prosthetic meshes are implanted each year worldwide. The simple selection of a single synthetic mesh construction for any level and type of pelvic floor dysfunctions without adopting the design to specific requirements increase the risks for mesh related complications. Adverse events are closely related to chronic foreign body reaction, with enhanced formation of scar tissue around the surgical meshes, manifested as pain, mesh erosion in adjacent structures (with organ tissue cut), mesh shrinkage, mesh rejection and eventually recurrence. Such events, especially scar formation depend on effective porosity of the mesh, which decreases discontinuously at a critical stretch when pore areas decrease making the surgical reconstruction ineffective that further augments the re-operation costs. The extent of fibrotic reaction is increased with higher amount of foreign body material, larger surface, small pore size or with inadequate textile elasticity. Standardized studies of different meshes are essential to evaluate influencing factors for the failure and success of the reconstruction. Measurements of elasticity and tensile strength have to consider the mesh anisotropy as result of the textile structure. An appropriate mesh then should show some integration with limited scar reaction and preserved pores that are filled with local fat tissue. This chapter reviews various tissue reactions to different monofilament mesh implants that are used for incontinence and hernia repairs and study their mechanical behavior. This helps to predict the functional and biological outcomes after tissue reinforcement with meshes and permits further optimization of the meshes for the specific indications to improve the success of the surgical treatment.
The terms bioeconomy and biorefineries are used for a variety of processes and developments. This short introduction is intended to provide a delimitation and clarification of the terminology as well as a classification of current biorefinery concepts. The basic process diagrams of the most important biorefinery types are shown.
To train end users how to interact with digital systems is indispensable to ensure a strong computer security. 'Competence Developing Game'-based approaches are particularly suitable for this purpose because of their motivation-and simulation-aspects. In this paper the Competence Developing Game 'GHOST' for cybersecurity awareness trainings and its underlying patterns are described. Accordingly, requirements for an 'Competence Developing Game' based training are discussed. Based on these requirements it is shown how a game can fulfill these requirements. A supplementary game interaction design and a corresponding evaluation study is shown. The combination of training requirements and interaction design is used to create a 'Competence Developing Game'-based training concept. A part of these concept is implemented into a playable prototype that serves around one hour of play respectively training time. This prototype is used to perform an evaluation of the game and training aspects of the awareness training. Thereby, the quality of the game aspect and the effectiveness of the training aspect are shown.
An amperometric bi-enzyme biosensor based on substrate recycling principle for the amplification of the sensor signal has been developed for the detection of adrenaline in blood. Adrenaline can be used as biomarker verifying successful adrenal venous sampling procedure. The adrenaline biosensor has been realized via modification of a galvanic oxygen sensor with a bi-enzyme membrane combining a genetically modified laccase and a pyrroloquinoline quinone-dependent glucose dehydrogenase. The measurement conditions such as pH value and temperature were optimized to enhance the sensor performance. A high sensitivity and a low detection limit of about 0.5–1 nM adrenaline have been achieved in phosphate buffer at pH 7.4, relevant for measurements in blood samples. The sensitivity of the biosensor to other catecholamines such as noradrenaline, dopamine and dobutamine has been studied. Finally, the sensor has been successfully applied for the detection of adrenaline in human blood plasma.
Human-induced pluripotent stem cell-derived cardiomyocytes (hiPS-CM) today are widely used for the investigation of normal electromechanical cardiac function, of cardiac medication and of mutations. Computational models are thus established that simulate the behavior of this kind of cells. This section first motivates the modeling of hiPS-CM and then presents and discusses several modeling approaches of microscopic and macroscopic constituents of human-induced pluripotent stem cell-derived and mature human cardiac tissue. The focus is led on the mapping of the computational results one can achieve with these models onto mature human cardiomyocyte models, the latter being the real matter of interest. Model adaptivity is the key feature that is discussed because it opens the way for modeling various biological effects like biological variability, medication, mutation and phenotypical expression. We compare the computational with experimental results with respect to normal cardiac function and with respect to inotropic and chronotropic drug effects. The section closes with a discussion on the status quo of the specificity of computational models and on what challenges have to be solved to reach patient-specificity.