Speech Analysis for the Automatic Detection and Monitoring of Parkinson's Disease
23.09.2019 von 10:00 bis 11:00
Parkinson's disease (PD) is second most common neurodegenerative disorder worldwide. It affects the control of muscles and limbs in the body and typically has negative impact on the speech production. Other motor activities like handwriting and gait are also affected. This talk will start with a general description of several neurodegenerative disorders including PD, Alzheimer's, and Aphasia. Typical speech disorders suffered by PD patients will be discussed and a methodology to automatically model those symptoms is also introduced. The suitability of such a methodology for the automatic detection and monitoring of PD is also discussed. In the final part of the talk, an extension of the methodology is presented considering other bio signals like handwriting and gait.
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
Inducing transitions between healthy and pathological brain states: insights from semi-empirical computational modeling
12.09.2019 von 16:00 bis 17:00
Human consciousness is correlated with the complexity of brain dynamics. The response of brain activity to external electromagnetic perturbations is a robust marker of the level of consciousness, both in physiological and pathological states. However, it is difficult to assess this response, and different approximations are adopted in practice. In my talk I will show how the development of semi-empirical models (i.e. fusion of dynamical systems with information obtained from structural and functional neuroimaging) allows to build "in silico brains" for the rehearsal of the response against different external perturbations. This freedom for virtual experimentation allows to investigate potential solutions to a much more difficult problem than diagnosis: the induction of transitions towards the state of ordinary wakefulness, one of the greatest challenges of contemporary neurology. While most of the examples provided in my talk will concern states of reduced awareness, the developed methodology can also be applied to other neurological and psychiatric conditions.
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
Ferroelectricity in methylammonium lead iodide perovskite solar cells
21.10.2019 von 16:00 bis 17:00
Among the emerging photovoltaic technologies, perovskite solar cells stand out with remarkable power conversion efficiencies (PCEs) and low-cost solution processability, rivaling established technologies. Currently, the scientific community controversially discusses the importance of the ferroic properties for the exceptional performance of MAPbI3 light-harvesting layers.
In this work, we performed a comprehensive AFM study including Piezoresponse Force Microscopy (PFM) and Kelvin Probe Force Microscopy (KPFM). On large flat crystals, we find 90 nm wide ferroelectric domains of alternating in-plane polarization. EBSD mapping allowed for the spatially resolved correlation of the ferroelectric patterns and the crystal orientation within the MAPbI3 thin-films. Electrical simulations provide insight into the working principle of ferroelectric MAPbI3 solar cells. Poling experiments elucidate the impact of the ferroelectric microstructure on macroscopic device properties.
Altogether, these investigations provide micro-structural target properties for MAPbI3 thin-film deposition and outline pathways forward for more efficient, eco-friendly and lead-free perovskite solar cells.
Holger Röhm1,2,Tobias Leonhard1,2, Alexander D. Schulz1,2, Susanne Wagner2,3,
Michael J. Hoffmann2,3 and Alexander Colsmann1,2
1 Light Technology Institute, Karlsruhe Institute of Technology,
Karlsruhe, 76131, Germany.
2 Material Research Center for Energy Systems, Karlsruhe Institute of Technology, Karlsruhe, 76131, Germany.
3 Institute for Applied Materials – Ceramic Materials and Technologies,
Karlsruhe Institute of Technology, Karlsruhe, 76131, Germany.
alexander.colsmann@kit.edu
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
Magnetic Particle Imaging - Promises and Challenges of an Emerging Imaging Modality
13.01.2020 von 16:00 bis 17:00
Magnetic Particle Imaging (MPI) is a recently invented three-dimensional imaging method that quantitatively measures the spatial distribution of a tracer based on magnetic nanoparticles. The modality promises a high sensitivity and high spatial as well as temporal resolution. There is a high potential of MPI to improve interventional and image-guided surgical procedures because, today, established medical imaging modalities typically excel in only one or two of these important imaging properties. MPI makes use of the non-linear magnetization characteristics of the magnetic nanoparticles.
For this purpose, two magnetic fields are created and superimposed, a static selection field and an oscillatory drive field. If, for instance, SPIONs, i.e. superparamagnetic iron-oxide nanoparticles, are subjected to the oscillatory magnetic field, the particles will react with a non-linear magnetization response, which can be measured with an appropriate pick-up coil arrangement. Due to the non-linearity of the particles' magnetization characteristics, the received signal consists of the fundamental excitation frequency as well as of harmonics, i.e. oscillations with multiples of the fundamental frequency.
After separation of the fundamental signal, the nanoparticle concentration can be estimated based on the harmonics. The spatial coding is realized with the static selection field that produces a field-free point. Essentially, reconstruction in MPI is the solution of an inverse problem, where, based on the measured induction voltages in the pick-up coils, the spatial distribution of the nanoparticles can be estimated. The relation between the measured voltages and the desired tracer distribution is established by the MPI system function.
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
Label-free particle and cell sorting based on inherent physical properties
11.11.2019 von 16:00 bis 17:00
Cell sorting is a key tool in medicine and biology to find and characterize cells of interest. Whereas standard immune-labelling is a powerful technique combined with for example fluorescence activated cell sorting (FACS), it does require expensive and sensitive equipment and reagents. Instead, we rely on the inherent physical properties of cells to avoid any need for labelling, and by implementing our sorting schemes on microfluidics platforms we make the preparation and analysis of a sample simpler, more widely accessible and cheaper.
In our laboratory we pioneered morphology, density and deformability-based sorting using deterministic lateral displacement (DLD) [1, 2] [3]. We characterized the mechanical and morphological properties of different variants of red blood cells as well as cancer cells and bacteria. We demonstrated the sorting of bacterial chains of different lengths[4], which is relevant due to the dependence of the morphology on virulence. We have been looking into nanoscale particles with our aims focused on extracellular vesicles. By combining the DLD with electrokinetics, we can tune the separation and sort particles that are close to 100nm [5].
Overall, we try to develop novel sorting schemes that rely on the inherent properties of the particles. This way we not only simplify the sorting but also open up for novel sorting parameters that are not accessible using standard approaches.
1. Beech, J.P., et al., Sorting cells by size, shape and deformability. Lab on a Chip, 2012. 12(6): p. 1048-1051.
2. Holm, S., J.P. Beech, and J.O. Tegenfeldt. Combined Density and Size-Based Sorting in Deterministic Lateral Displacement Devices. in microTAS2013. 2013. Freiburg, Germany: Society for Chemistry and Micro-Nano Systems.
3. Holm, S.H., et al., Simplifying microfluidic separation devices towards field-detection of blood parasites. Analytical Methods, 2016. 8: p. 3291-3300.
4. Beech, J.P., et al., Separation of pathogenic bacteria by chain length. Analytica Chimica Acta, 2018. 1000: p. 223-231.
5. Beech, J.P., et al., Active Posts in Deterministic Lateral Displacement Devices. Advanced Materials Technologies, 2019. 4(9): p. 1900339.
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
Unconscious Bias – Unbewusste Vorurteile
06.11.2019 von 16:00 bis 18:00
Jeden Tag treffen wir Entscheidungen und glauben, dass diese sachlich begründet sind. Tatsächlich gilt dies nur für die wenigsten unserer Entscheidungen. Diese sind vielmehr überwiegend ein Produkt unseres Bauchgefühls, basierend auf den Vorgaben unseres Unterbewusstseins. Wenn wir die Kompetenz und Leistung von Menschen bewerten, beurteilen wir niemals nur die Leistung, sondern immer auch den Menschen. Und so kommt es, dass wir die Leistung von Männern und Frauen, von schlanken und dicken Menschen, von Einheimischen und Ausländern ganz unterschiedlich bewerten. In unsere Bewertung fließen jede Menge Aspekte mit hinein, die in der Leistungsbeurteilung nichts zu suchen haben.
Der interaktive Vortrag „Unconscious Bias“ klärt darüber auf, und beschäftigt sich mit folgenden Themen
- Was sind unbewusste Vorurteile und woher kommen sie?
- Welche unbewussten Vorurteile bestehen im Berufsleben?
- Wie wirken sich diese aus?
- Was kann man gegen die unbewussten Vorurteile tun?
|
 |
| |
Dr. Nicola Byok |
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
An Algorithmic Method of Partial Derivatives
15.11.2019 von 14:15 bis 15:00
We give applications of the following problem to parameterized algorithms: Given a d × d matrix A, whose entries are linear forms, and a differential operator T of degree d represented by a (skew) arithmetic circuit, decide if the partial derivative of det(A) by T vanishes. This approach leads to improvements and significant simplifications of previous parameterized algorithms for unweighted problems. This question is connected to recent approaches in parameterized algorithms involving the exterior algebra and Waring rank, and raises natural questions in commutative algebra that could yield further improvements.
This is joint work with Kevin Pratt from CMU.
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
Silicon photonics for electronic photonic integration
02.12.2019 von 16:00 bis 17:00
Abstract:
Already since the advent of cable telephony there is a remarkable growth of network traffic on a global scale. A similarity to Moore’s law of microelectronics has been pointed out. In fact, this permanent growth would have been impossible without the fascinating advance of IC technology. The technologies required to sustain this unprecedented growth are photonic and electronic integration. Optical solutions intimately integrated with electronics are required to satisfy the demand of bandwidth, energy-per-bit, and cost. The power horse of electronic photonic integration is silicon photonics, in particular photonic CMOS or BiCMOS technologies. Going beyond the concept of present More-than-Moore technologies, photonic BiCMOS combines 2 high-performance technology sub-modules (high –performance HBTs and photonics) with a baseline CMOS process. The added complexity pays off with a full set of photonic features fit for broad-band optoelectronic system integration.
The talk shall provide a short introduction to the field of silicon photonics. We shall then present IHP’s perspective on a photonic More-than-Moore technology, joining most advanced HBT technology with high-speed photonic devices in photonic BiCMOS. The final part of the presentation will present examples of photonic electronic integration for a range of applications.
Bio:
Lars Zimmermann leads the silicon photonics team at IHP, the Leibniz Institute für innovative Mikroelektronik in Frankfurt (Oder) and is professor at TU Berlin in the field of silicon photonics. He coordinates the cooperation of IHP and TU Berlin in the frame of the Joint Lab Silicon Photonics. His current work is dedicated to high-performance photonic-electronic integration for optical communications and for nonlinear optical signal processing. Lars Zimmermann undertook his postgraduate studies at Katholieke Universiteit Leuven, Belgium. In Leuven, he was affiliated with IMEC where he worked for 5 years and received a PhD degree in 2003. His scientific work at IMEC dealt with the development of extended short-wave infrared detector arrays and sensor assembly processes. In 2004, Lars Zimmermann moved to TU Berlin. In Berlin, he worked for 5 years on silicon-based optical motherboard technology, realizing early hybrid assemblies of silicon waveguides with lasers, semiconductor optical amplifiers, and detectors. In 2008, Lars Zimmermann moved to IHP, directing IHP’s silicon photonics developments. In 2018, he re-joined TU Berlin.
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
Control and planning for autonomous systems under uncertainty - fusing predictive control and learning with guarantees
Dynamic multi-frequency analysis: looking into the dynamic behaviour of electrochemical systems
13.11.2019 von 12:00 bis 13:00
Aquarium
Dynamic multi-frequency analysis (DMFA) is a tool that allows acquiring dynamic impedance spectra of electrochemical systems in a large range of frequencies, from 1 MHz down to few Hz, with high time-resolved precision [1,2]. Starting from the physical definition of impedance and admittance, and using Volterra series, it has been shown how the concept of dynamic impedance is an extension of the concept of stationary impedance [3]. Here an overlook in the application of dynamic multi-frequency analysis to several electrochemical systems will be shown, among which electron transfer to a redox couple, hydrogen evolution reaction, ion intercalation in Prussian blue analogues and lithium manganese oxide, silicon electro-oxidation. In particular, several aspects of modeling and fitting the dynamic impedance spectra will be addressed. In addition to this, a short overview on the analysis of the dynamic non-linear frequency response in electrochemical systems will be addressed, and first experimental results on electron transfer to a redox couple will be shown.
Literature:
[1] A. Battistel, G. Du and F. La Mantia, Electroanal. 2016, 28, 2346.
[2] D. Koster, G. Du, A. Battistel, and F. La Mantia, Electrochim. Acta 2017, 246, 553
[3] A. Battistel, and F. La Mantia, Electrochim. Acta 2019, 304, 513.
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
Survey of state estimation for (bio)chemical systems - A personal perspective
Advanced functional materials for sustainable energy applications
16.12.2019 von 16:00 bis 17:00
Aquarium-D-036
Research on phase transitions in advanced functional materials is widespread and continues to grow. Scientific interest aside, they are attractive for a wide range of current and future technologies that include computation, medical instrumentation, and energy conversion and storage. During this talk, I will present my work on ferromagnetic, ferroelectric and ferroelastic phase transitions that permit large thermal changes to be driven by changes in magnetic field, electric field or stress field. The resulting magnetocaloric, electrocaloric and mechanocaloric (barocaloric and elastocaloric) effects promise new cooling technologies that are energy efficient and environmentally friendly. The materials that I study include multiferroic materials that possess strong magnetoelectric coupling, which are interesting also for applications in sensors, actuators and data storage.
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
Innovative soft magnetic materials for electrical drives
20.01.2020 von 16:00 bis 17:00
Aquarium
In electric drive technology and mechatronics the market for variable-speed drives is continuously increasing. These electrical machines are requested to provide maximum drive power whereas only minimum installation space or little weight is required. In addition, there is often the demand for high energy efficiency.
Applications range from electric traction drives for battery electric vehicles, electric aircrafts, fans and compressors to medical technology. The enhancement of mechanical power is usually accomplished by increasing the speed. As a result, the supply frequency of the three-phase system must be significantly elevated compared to drives that work directly on the 50 Hz supply grid. However, this has a strong impact on the power flow within the machine. The increase in the current heat losses in the winding - due to current skin effects at higher frequencies - can be sufficiently counteracted by a suitable winding design. On the contrary, core losses increase intrinsically with the frequency. Hence thinner electrical sheets with reduced losses are used. However this procedure is limited to supply frequencies of 1 kHz due to material composition.
An innovative alternative to standardized electrical sheets is the use of soft magnetic composites (SMC). In addition to a significant reduction of eddy current losses, SMC is a magnetically isotropic material. In contrast to common radial flux machines, the guidance of the magnetic flux can be extended to the third space dimension without additional efforts and costs. As a result, it is possible to develop novel motor designs and manufacturing processes which provide numerous advantages.
During the colloquium talk, current SMC products are presented and compared with conventional soft magnetic materials. In addition, the opportunities and limits of these new materials for electrical drives are demonstrated by means of prototypes and their testing. Finally some indications are given in which scenarios the use of SMC is particularly of advantage.
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
Extrem kompakte Antriebssysteme für elektrische Fahrzeuge
20.01.2020 von 17:00 bis 18:00
Aquarium-D-036
Fahrzeuge mit elektrischen Antriebsstrang werden zukünftig unseren Straßenverkehr prägen. Die Reichweite und der Preis solcher Fahrzeuge beeinflussen maßgeblich die Nutzerakzeptanz. Die Kosten und die Reichweite werden üblicherweise von der Batteriegröße und dem Fahrzeuggewicht beeinflusst, daher werden für solche Fahrzeuge extrem kompakte elektrische Antriebsstränge mit einem sehr hohen Wirkungsgrad benötig.
Die Fachhochschule Kiel arbeitet im Rahmen verschiedener Forschungsprojekte, u.a. mit Volkswagen, an solchen Antriebssystemen der nächsten bzw. übernächsten Generation. Daraus sind Inverter mit einer Leistungsdichte von mehr als 100kW/l entstanden.
Im Rahmen dieses Vortrages sollen beispielhaft Lösungsansetze für verschiedene Herausforderungen anhand von Forschungsprojekten dargestellt werden.
Zu den Herausforderungen gehören u.a.: das Systemdesgin bei hohen thermischen Anforderungen; die Erhöhung des Wirkungsgrades durch Minimierung der Streu-Induktivitäten im Kommutierungskreis; innovative Ansätze zur Strommessung in Invertern auf möglichst kleinem Bauraum; Treiberentwicklung für einen minimalen Bauraum; Zustandsüberwachung etc.
Im Rahmen des Vortrages sollen die Lösungswegen zu den unterschiedlichen Problemen dargestellt werden, die vielfach aus einem theoretischen Ansatz, einer Schaltungs- oder FEM-Simulation und praktischer Verifizierung bestehen.
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
On Privacy Notions in Anonymous Communication
The Resurrection of Models: Generative Software Engineering in Research and Real-World Projects
Optimale Planung und den Betrieb von effizienten „Net-Zero Energy Factories“
MIMO - Wireless Communications meets Radar
Intelligente Systeme - Vom autonomen maschinellen Lernen zur technischen Anwendung
31.01.2020 von 14:00 bis 16:00
Audimax, Hörsaal A
Immer öfter übernehmen intelligente Systeme Aufgaben aus dem täglichen Leben und der Arbeitswelt autonom und zuverlässig. Tätigkeiten laufen automatisiert im Hintergrund ab und schaffen damit Freiräume für den Nutzer.
Neben der Entlastung des Menschen von einfachen Aufgaben besteht ein weiterer Vorteil darin, dass intelligente Systeme lernfähig sind. Die zunehmende Nutzung dieser Systeme führt zu robusteren, effizienteren und flexibleren Lösungen, da sie sich kontinuierlich selbst verbessern und kontext-abhängig agieren.
Beispielsweise kann eine Kaffeemaschine ihren Nutzer erkennen und die bevorzugte Zubereitungsform wählen, eine Zahnbürste informiert, wenn der Zahlbelag noch nicht vollständig entfernt wurde und eine Ampelanlage schaltet auf die jeweilige Verkehrssituation um.
Intelligente Systeme basieren auf Mechanismen zur kontinuierlichen Überwachung und Anpassung an sich verändernde Parameter. Wesentlich ist dabei die Herausforderung, auf unbekannte Situationen adäquat zu reagieren oder auch bei stark gestörten Bedingungen einsetzbar zu bleiben: Herausforderungen, für die bislang Menschen ihre Kreativität einsetzen. Deswegen besteht das Ziel bei der Entwicklung intelligenter System vorrangig darin, Verfahren und Algorithmen zu entwickeln, die menschliche Kreativität nachbilden können.
In meinem Vortrag werde ich in das Gebiet intelligente Systeme einführen, einen Überblick über die Arbeit meiner Gruppe geben sowie einen Ausblick auf Herausforderungen und künftige Entwicklungen dieses Forschungsbereiches geben.
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
Nano Self-assembly and Electrical Function
10.02.2020 von 16:00 bis 17:00
Aquarium-D-036
Self-assembled structures are of large importance in chemistry and biology. Their construction principles may become significant for the future bottom-up construction of three-dimensional electronic architectures. I will review our recent investigations on the electronic charge transport in various nanoscale self-assembled structures, ranging from molecular mono- and bilayers on solid surfaces to regular arrays of metallic nanocubes, formed by directed self-assembly.
Diesen Termin meinem iCal-Kalender hinzufügen
zurück
