Refine
Year of publication
- 2024 (23)
- 2023 (29)
- 2022 (45)
- 2021 (53)
- 2020 (57)
- 2019 (64)
- 2018 (60)
- 2017 (61)
- 2016 (43)
- 2015 (61)
- 2014 (58)
- 2013 (54)
- 2012 (59)
- 2011 (71)
- 2010 (61)
- 2009 (73)
- 2008 (53)
- 2007 (45)
- 2006 (64)
- 2005 (40)
- 2004 (75)
- 2003 (46)
- 2002 (46)
- 2001 (48)
- 2000 (51)
- 1999 (29)
- 1998 (25)
- 1997 (25)
- 1996 (21)
- 1995 (16)
- 1994 (11)
- 1993 (16)
- 1992 (7)
- 1991 (5)
- 1990 (11)
- 1989 (11)
- 1988 (17)
- 1987 (6)
- 1986 (2)
- 1985 (3)
- 1984 (1)
- 1983 (2)
- 1982 (20)
- 1981 (13)
- 1980 (27)
- 1979 (18)
- 1978 (26)
- 1977 (13)
- 1976 (12)
- 1975 (9)
- 1974 (2)
- 1973 (1)
- 1972 (2)
- 1968 (1)
Institute
- Fachbereich Medizintechnik und Technomathematik (1692) (remove)
Language
- English (1692) (remove)
Document Type
- Article (1359)
- Conference Proceeding (212)
- Book (43)
- Part of a Book (43)
- Doctoral Thesis (18)
- Other (6)
- Patent (4)
- Preprint (3)
- Lecture (2)
- Habilitation (1)
Keywords
- Biosensor (25)
- Finite-Elemente-Methode (12)
- Einspielen <Werkstoff> (10)
- CAD (8)
- civil engineering (8)
- Bauingenieurwesen (7)
- FEM (6)
- Limit analysis (6)
- Shakedown analysis (6)
- shakedown analysis (6)
The growing body of political texts opens up new opportunities for rich insights into political dynamics and ideologies but also increases the workload for manual analysis. Automated speaker attribution, which detects who said what to whom in a speech event and is closely related to semantic role labeling, is an important processing step for computational text analysis. We study the potential of the large language model family Llama 2 to automate speaker attribution in German parliamentary debates from 2017-2021. We fine-tune Llama 2 with QLoRA, an efficient training strategy, and observe our approach to achieve competitive performance in the GermEval 2023 Shared Task On Speaker Attribution in German News Articles and Parliamentary Debates. Our results shed light on the capabilities of large language models in automating speaker attribution, revealing a promising avenue for computational analysis of political discourse and the development of semantic role labeling systems.
The growing body of political texts opens up new opportunities for rich insights into political dynamics and ideologies but also increases the workload for manual analysis. Automated speaker attribution, which detects who said what to whom in a speech event and is closely related to semantic role labeling, is an important processing step for computational text analysis. We study the potential of the large language model family Llama 2 to automate speaker attribution in German parliamentary debates from 2017-2021. We fine-tune Llama 2 with QLoRA, an efficient training strategy, and observe our approach to achieve competitive performance in the GermEval 2023 Shared Task On Speaker Attribution in German News Articles and Parliamentary Debates. Our results shed light on the capabilities of large language models in automating speaker attribution, revealing a promising avenue for computational analysis of political discourse and the development of semantic role labeling systems.
Using a cell-based gas biosensor for investigation of adverse effects of acetone vapors in vitro
(2013)
Cell-based sensors for the detection of gases have long been underrepresented, due to the cellular requirement of being cultured in a liquid environment. In this work we established a cell-based gas biosensor for the detection of toxic substances in air, by adapting a commercial sensor chip (Bionas®), previously used for the measurement of pollutants in liquids. Cells of the respiratory tract (A549, RPMI 2650, V79), which survive at a gas phase in a natural context, are used as biological receptors. The physiological cell parameters acidification, respiration and morphology are continuously monitored in parallel. Ammonia was used as a highly water-soluble model gas to test the feasibility of the sensor system. Infrared measurements confirmed the sufficiency of the medium draining method. This sensor system provides a basis for many sensor applications such as environmental monitoring, building technology and public security.
Clearance of blood components and fluid drainage play a crucial role in subarachnoid hemorrhage (SAH) and post hemorrhagic hydrocephalus (PHH). With the involvement of interstitial fluid (ISF) and cerebrospinal fluid (CSF), two pathways for the clearance of fluid and solutes in the brain are proposed. Starting at the level of capillaries, flow of ISF follows along the basement membranes in the walls of cerebral arteries out of the parenchyma to drain into the lymphatics and CSF [1]–[3]. Conversely, it is shown that CSF enters the parenchyma between glial and pial basement membranes of penetrating arteries [4]–[6]. Nevertheless, the involved structures and the contribution of either flow pathway to fluid balance between the subarachnoid space and interstitial space remains controversial. Low frequency oscillations in vascular tone are referred to as vasomotion and corresponding vasomotion waves are modeled as the driving force for flow of ISF out of the parenchyma [7]. Retinal vessel analysis (RVA) allows non-invasive measurement of retinal vessel vasomotion with respect to diameter changes [8]. Thus, the aim of the study is to investigate vasomotion in RVA signals of SAH and PHH patients.
The MYOTONES experiment is the first to monitor changes in the basic biomechanical properties (tone, elasticity and stiffness) of the resting human myofascial system due to microgravity with a oninvasive, portable device on board the ISS. The MyotonPRO device applies several brief mechanical stimuli to the surface of the skin, and the natural oscillation signals of the tissue beneath are detected and computed by the MyotonPRO. Thus, an objective, quick and easy determination of the state of the underlying tissue is possible.
Two preflight, four inflight and four post flight measurements were performed on a male astronaut using the same 10 measurement points (MP) for each session. MPs were located on the plantar fascia, Achilles tendon, M. soleus, M. gastrocnemius, M. multifidus, M. splenius capitis, M. deltoideus anterior, M. rectus femoris, infrapatellar tendon, M. tibialis anterior. Subcutaneous tissues thickness above the MPs was measured using ultrasound imaging. Magnetic resonance images (MRI) of lower limb muscles and functional tests were also performed pre- and postflight.
Our first measurements on board the ISS confirmed increased tone and stiffness of the lumbar multifidus muscle, an important trunk stabilizer, dysfunction of which is known to be associated with back pain. Furthermore, reduced tone and stiffness of Achilles tendon and plantar fascia were observed inflight vs. preflight, confirming previous findings from terrestrial analog studies and parabolic
flights. Unexpectedly, the deltoid showed negative inflight changes in tone and stiffness, and increased elasticity, suggesting a potential risk of muscle atrophy in longer spaceflight that should be addressed by adequate inflight countermeasure protocols. Most values from limb and back MPS showed deflected patterns (in either directions) from inflight shortly after the re-entry phase on the landing day and one week later. Most parameter values then normalized to baseline after 3 weeks likely due to 1G re-adaptation and possible outcome of the reconditioning protocol. No major changes in subcutaneous tissues thickness above the MPs were found inflight vs preflight, suggesting no bias (i.e., fluid shift, extreme tissue thickening or loss). Pre- and postflight MRI and functional tests showed negligible changes in calf muscle size, power and force, which is likely due to training effects from current inflight exercise protocols.
The MYOTONES experiment is currently ongoing to collect data from further crew members. The potential impact of this research is to better understand the effects of microgravity and countermeasures over the time course of an ISS mission cycle. This will enable exercise countermeasures to be tailored