Search

Your search keyword '"Mehnert A"' showing total 4,310 results
Start Over Author "Mehnert A" Search Limiters Full Text
4,310 results on '"Mehnert A"'

1. Squeeze-and-Remember Block

Author

Cakaj, Rinor, Mehnert, Jens, and Yang, Bin

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

Convolutional Neural Networks (CNNs) are important for many machine learning tasks. They are built with different types of layers: convolutional layers that detect features, dropout layers that help to avoid over-reliance on any single neuron, and residual layers that allow the reuse of features. However, CNNs lack a dynamic feature retention mechanism similar to the human brain's memory, limiting their ability to use learned information in new contexts. To bridge this gap, we introduce the "Squeeze-and-Remember" (SR) block, a novel architectural unit that gives CNNs dynamic memory-like functionalities. The SR block selectively memorizes important features during training, and then adaptively re-applies these features during inference. This improves the network's ability to make contextually informed predictions. Empirical results on ImageNet and Cityscapes datasets demonstrate the SR block's efficacy: integration into ResNet50 improved top-1 validation accuracy on ImageNet by 0.52% over dropout2d alone, and its application in DeepLab v3 increased mean Intersection over Union in Cityscapes by 0.20%. These improvements are achieved with minimal computational overhead. This show the SR block's potential to enhance the capabilities of CNNs in image processing tasks., Comment: Accepted by The International Conference on Machine Learning and Applications (ICMLA) 2024

Published
2024

2. Spectral Wavelet Dropout: Regularization in the Wavelet Domain

Author

Cakaj, Rinor, Mehnert, Jens, and Yang, Bin

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

Regularization techniques help prevent overfitting and therefore improve the ability of convolutional neural networks (CNNs) to generalize. One reason for overfitting is the complex co-adaptations among different parts of the network, which make the CNN dependent on their joint response rather than encouraging each part to learn a useful feature representation independently. Frequency domain manipulation is a powerful strategy for modifying data that has temporal and spatial coherence by utilizing frequency decomposition. This work introduces Spectral Wavelet Dropout (SWD), a novel regularization method that includes two variants: 1D-SWD and 2D-SWD. These variants improve CNN generalization by randomly dropping detailed frequency bands in the discrete wavelet decomposition of feature maps. Our approach distinguishes itself from the pre-existing Spectral "Fourier" Dropout (2D-SFD), which eliminates coefficients in the Fourier domain. Notably, SWD requires only a single hyperparameter, unlike the two required by SFD. We also extend the literature by implementing a one-dimensional version of Spectral "Fourier" Dropout (1D-SFD), setting the stage for a comprehensive comparison. Our evaluation shows that both 1D and 2D SWD variants have competitive performance on CIFAR-10/100 benchmarks relative to both 1D-SFD and 2D-SFD. Specifically, 1D-SWD has a significantly lower computational complexity compared to 1D/2D-SFD. In the Pascal VOC Object Detection benchmark, SWD variants surpass 1D-SFD and 2D-SFD in performance and demonstrate lower computational complexity during training., Comment: Accepted by The International Conference on Machine Learning and Applications (ICMLA) 2024

Published
2024

3. CNN Mixture-of-Depths

Author

Cakaj, Rinor, Mehnert, Jens, and Yang, Bin

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

We introduce Mixture-of-Depths (MoD) for Convolutional Neural Networks (CNNs), a novel approach that enhances the computational efficiency of CNNs by selectively processing channels based on their relevance to the current prediction. This method optimizes computational resources by dynamically selecting key channels in feature maps for focused processing within the convolutional blocks (Conv-Blocks), while skipping less relevant channels. Unlike conditional computation methods that require dynamic computation graphs, CNN MoD uses a static computation graph with fixed tensor sizes which improve hardware efficiency. It speeds up the training and inference processes without the need for customized CUDA kernels, unique loss functions, or finetuning. CNN MoD either matches the performance of traditional CNNs with reduced inference times, GMACs, and parameters, or exceeds their performance while maintaining similar inference times, GMACs, and parameters. For example, on ImageNet, ResNet86-MoD exceeds the performance of the standard ResNet50 by 0.45% with a 6% speedup on CPU and 5% on GPU. Moreover, ResNet75-MoD achieves the same performance as ResNet50 with a 25% speedup on CPU and 15% on GPU., Comment: Conference Paper of the Asian Conference on Computer Vision (ACCV) 2024

Published
2024

4. Weather effects on the lifecycle of U.S. Department of Defense equipment replacement (WELDER)

Author

Larsen, Peter, Grussing, Michael, Bercos-Hickey, Emily, Bidner, Christine, LaCommare, Kristina, Landers, Kirsten, Mehnert, Brenda, Patricola, Christina, Powell, Austin, Spears, Michael, and Wehner, Michael

Subjects
Built Environment and Design, Building, Climate Action, Extreme weather, Climate change, Facilities, Lifecycle costs, Sustainment management, Environmental Science and Management, Architecture, Building & Construction, Built environment and design, Engineering
Abstract

Extreme weather has a direct and significant impact on buildings and infrastructure, resulting in billions of dollars of damage each year. This problem continues to grow as climate patterns change and buildings are exposed to new and different hazards than what they were designed to withstand. In order to better plan for the long-range sustainment, restoration, modernization, and eventual recapitalization of these buildings, organizations with large building portfolios, such as the U.S. Department of Defense (DoD), must have an awareness of the risks that these extreme weather events present. This research aimed to develop an approach to estimate condition loss and reduction in service life for the components of a building due to extreme weather hazards, to understand the risks that may be present in certain buildings and building systems. To achieve this objective, a damage association matrix was developed that categorizes climate hazards, the damage modes that they produce, and the individual component types impacted. This damage matrix formally links state-of-the-art climate model output, which provides projections of the probability of various climate hazards with a damage effects model that quantifies the consequence on component-level condition and service life. This method is applied to an actual portfolio of buildings in a particular geographic location and with a pre-defined component inventory that comprises the building. This approach can be aggregated to the system-, facility-, and site-level thus helping support billions of dollars in recapitalization decisions related to restoration/modernization of facilities.

Published
2024

5. Scalable Distributed String Sorting

Author

Kurpicz, Florian, Mehnert, Pascal, Sanders, Peter, and Schimek, Matthias

Subjects
Computer Science - Data Structures and Algorithms
Abstract

String sorting is an important part of tasks such as building index data structures. Unfortunately, current string sorting algorithms do not scale to massively parallel distributed-memory machines since they either have latency (at least) proportional to the number of processors $p$ or communicate the data a large number of times (at least logarithmic). We present practical and efficient algorithms for distributed-memory string sorting that scale to large $p$. Similar to state-of-the-art sorters for atomic objects, the algorithms have latency of about $p^{1/k}$ when allowing the data to be communicated $k$ times. Experiments indicate good scaling behavior on a wide range of inputs on up to 49152 cores. Overall, we achieve speedups of up to 5 over the current state-of-the-art distributed string sorting algorithms.

Published
2024

6. Instant Complexity Reduction in CNNs using Locality-Sensitive Hashing

Author

Meiner, Lukas, Mehnert, Jens, and Condurache, Alexandru Paul

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

To reduce the computational cost of convolutional neural networks (CNNs) for usage on resource-constrained devices, structured pruning approaches have shown promising results, drastically reducing floating-point operations (FLOPs) without substantial drops in accuracy. However, most recent methods require fine-tuning or specific training procedures to achieve a reasonable trade-off between retained accuracy and reduction in FLOPs. This introduces additional cost in the form of computational overhead and requires training data to be available. To this end, we propose HASTE (Hashing for Tractable Efficiency), a parameter-free and data-free module that acts as a plug-and-play replacement for any regular convolution module. It instantly reduces the network's test-time inference cost without requiring any training or fine-tuning. We are able to drastically compress latent feature maps without sacrificing much accuracy by using locality-sensitive hashing (LSH) to detect redundancies in the channel dimension. Similar channels are aggregated to reduce the input and filter depth simultaneously, allowing for cheaper convolutions. We demonstrate our approach on the popular vision benchmarks CIFAR-10 and ImageNet. In particular, we are able to instantly drop 46.72% of FLOPs while only losing 1.25% accuracy by just swapping the convolution modules in a ResNet34 on CIFAR-10 for our HASTE module.

Published
2023

7. Spectral Batch Normalization: Normalization in the Frequency Domain

Author

Cakaj, Rinor, Mehnert, Jens, and Yang, Bin

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

Regularization is a set of techniques that are used to improve the generalization ability of deep neural networks. In this paper, we introduce spectral batch normalization (SBN), a novel effective method to improve generalization by normalizing feature maps in the frequency (spectral) domain. The activations of residual networks without batch normalization (BN) tend to explode exponentially in the depth of the network at initialization. This leads to extremely large feature map norms even though the parameters are relatively small. These explosive dynamics can be very detrimental to learning. BN makes weight decay regularization on the scaling factors $\gamma, \beta$ approximately equivalent to an additive penalty on the norm of the feature maps, which prevents extremely large feature map norms to a certain degree. However, we show experimentally that, despite the approximate additive penalty of BN, feature maps in deep neural networks (DNNs) tend to explode at the beginning of the network and that feature maps of DNNs contain large values during the whole training. This phenomenon also occurs in a weakened form in non-residual networks. SBN addresses large feature maps by normalizing them in the frequency domain. In our experiments, we empirically show that SBN prevents exploding feature maps at initialization and large feature map values during the training. Moreover, the normalization of feature maps in the frequency domain leads to more uniform distributed frequency components. This discourages the DNNs to rely on single frequency components of feature maps. These, together with other effects of SBN, have a regularizing effect on the training of residual and non-residual networks. We show experimentally that using SBN in addition to standard regularization methods improves the performance of DNNs by a relevant margin, e.g. ResNet50 on ImageNet by 0.71%., Comment: Accepted by The International Joint Conference on Neural Network (IJCNN) 2023

Published
2023

8. Weight Compander: A Simple Weight Reparameterization for Regularization

Author

Cakaj, Rinor, Mehnert, Jens, and Yang, Bin

Subjects
Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition
Abstract

Regularization is a set of techniques that are used to improve the generalization ability of deep neural networks. In this paper, we introduce weight compander (WC), a novel effective method to improve generalization by reparameterizing each weight in deep neural networks using a nonlinear function. It is a general, intuitive, cheap and easy to implement method, which can be combined with various other regularization techniques. Large weights in deep neural networks are a sign of a more complex network that is overfitted to the training data. Moreover, regularized networks tend to have a greater range of weights around zero with fewer weights centered at zero. We introduce a weight reparameterization function which is applied to each weight and implicitly reduces overfitting by restricting the magnitude of the weights while forcing them away from zero at the same time. This leads to a more democratic decision-making in the network. Firstly, individual weights cannot have too much influence in the prediction process due to the restriction of their magnitude. Secondly, more weights are used in the prediction process, since they are forced away from zero during the training. This promotes the extraction of more features from the input data and increases the level of weight redundancy, which makes the network less sensitive to statistical differences between training and test data. We extend our method to learn the hyperparameters of the introduced weight reparameterization function. This avoids hyperparameter search and gives the network the opportunity to align the weight reparameterization with the training progress. We show experimentally that using weight compander in addition to standard regularization methods improves the performance of neural networks., Comment: Accepted by The International Joint Conference on Neural Network (IJCNN) 2023

Published
2023

11. Peer2Me – evaluation of a peer supported program for adolescent and young adult (AYA) cancer patients: study protocol of a randomised trial using a comprehensive cohort design

Author

Hannah Brock, Sarah Dwinger, Corinna Bergelt, Annekathrin Sender, Kristina Geue, Anja Mehnert-Theuerkauf, and Diana Richter

Subjects
AYA, Cancer, Peer support, Mentoring, Emotional support, Young adults, Neoplasms. Tumors. Oncology. Including cancer and carcinogens, RC254-282
Abstract

Abstract Background Developing cancer in young adulthood is a non-normative life event and associated with adverse physical, social and psychological consequences. High psychological distress is common in AYA cancer patients including anxiety, depression or fear of recurrence. At the same time, it is well known that AYA often report unmet needs for support, particularly in terms of informational exchange and emotional support from peers in order to benefit from shared experiences and enhance self-efficacy. Especially in the AYA group, interactions with other same-aged cancer patients may represent an essential resource in terms of coping with the disease, as family members and friends are often overwhelmed and struggling with helplessness. Currently, there is a lack of professional support services using peer support (e.g. psycho-oncological support, aftercare consultations, social legal counselling) or evaluated peer support interventions in Germany. Our aim is to assess the effectiveness of the Peer2Me intervention for AYAs, in which acute patients (mentees) are accompanied by an AYA survivor (mentor) over a period of three months. Methods A prospective Comprehensive Cohort Design with repeated measures will be used to evaluate the effectiveness of Peer2Me for AYA. A sample of 180 patients in active cancer treatment aged 18 to 39 years will be enrolled and randomized to the intervention or control condition (a single AYA-specific consultation). Following mentor training, mentees and mentors are matched by diagnosis, age, and gender. The primary outcome is self-efficacy; secondary outcomes include measures of anxiety, depression, health literacy, life satisfaction and social support life. Outcomes will be measured at baseline before the intervention (t1), immediately after completion of the three-month intervention (t2) and three months after completion the intervention (t3). For the final analyses, we will use an intention-to-treat approach (ITT) and compare patients in the assigned treatment groups. Discussion Peer2Me might be an important addition to existing professional psychosocial support services for young cancer patients. At the end of the study, a psycho-oncological intervention for young cancer patients undergoing acute treatment should be available, from which both mentors and mentees could benefit. The long-term continuity of Peer2Me should be ensured through collaboration with different partners. Trial Registration The study was retrospectively registered on February 4, 2022 at clinicaltrials.gov (NCT05336318).

Published
2024
Full Text
View/download PDF

12. Identifying central dimensions of quality of life including life-related values, preferences and functional health in older patients with cancer: a scoping review protocol

Author

Franziska Springer, Ayumu Matsuoka, Kyoko Obama, Anja Mehnert-Theuerkauf, Yosuke Uchitomi, and Maiko Fujimori

Subjects
cancer, elderly cancer patients, quality of life, geriatric, scoping review, Psychology, BF1-990
Abstract

BackgroundOlder patients with cancer already represent the largest proportion of cancer survivors which will further increase in the upcoming years. However, older patients are highly underrepresented in clinical research, leading to a detrimental knowledge gap. Research on important aspects of quality of life (QoL) and associated factors for older patients with cancer is insufficient to date.AimThe objective of this scoping review therefore is to investigate the dimensions of QoL including functional health, life-relevant values and preferences in older adults with cancer across all tumor entities and health care settings. It will further identify medical, sociodemographic, psychosocial and geriatric aspects associated with QoL in the elderly and compare these with younger cancer patients and older non-cancer cases.MethodsPublished articles investigating QoL dimensions and associated factors in older patients with cancer, i.e., exclusively patients ≥65 years or mean/median age ≥ 70 years for age-mixed samples, or that compare results of older with younger cancer patients or with older non-cancer cases will be considered for this scoping review. Older patients with cancer across all tumor entities, disease stages and health care setting will be included. PubMed and PsychINFO databases will be searched for relevant articles. Abstracts and titles will be screened for basic inclusion, and two independent reviewers will conduct a full text screening to evaluate the age criteria and decide on the final inclusion of the study. Data on study and participant characteristics, QoL dimensions and geriatric factors will be extracted using a data extraction sheet. Results will be summarized descriptively to address the objectives of this review.DiscussionThe findings of this scoping review will provide valuable insights into central dimensions of QoL, including values, preferences and functional health in older adults with cancer, and help to improve targeted interventions and healthcare planning.

Published
2024
Full Text
View/download PDF

13. Phosphorylation‐linked complex profiling identifies assemblies required for Hippo signal integration

Author

Uliana, Federico, Ciuffa, Rodolfo, Mishra, Ranjan, Fossati, Andrea, Frommelt, Fabian, Keller, Sabrina, Mehnert, Martin, Birkeland, Eivind Salmorin, Drogen, Frank, Srejic, Nevena, Peter, Matthias, Tapon, Nicolas, Aebersold, Ruedi, and Gstaiger, Matthias

Subjects
Biochemistry and Cell Biology, Biological Sciences, Underpinning research, 1.1 Normal biological development and functioning, Generic health relevance, Humans, Phosphorylation, Signal Transduction, Adaptor Proteins, Signal Transducing, YAP-Signaling Proteins, Protein Serine-Threonine Kinases, Transcription Factors, Protein Tyrosine Phosphatases, Non-Receptor, Interaction proteomics, Protein phosphorylation, Protein-protein interactions, YAP1, Other Biological Sciences, Bioinformatics, Biochemistry and cell biology
Abstract

While several computational methods have been developed to predict the functional relevance of phosphorylation sites, experimental analysis of the interdependency between protein phosphorylation and Protein-Protein Interactions (PPIs) remains challenging. Here, we describe an experimental strategy to establish interdependencies between protein phosphorylation and complex formation. This strategy is based on three main steps: (i) systematically charting the phosphorylation landscape of a target protein; (ii) assigning distinct proteoforms of the target protein to different protein complexes by native complex separation (AP-BNPAGE) and protein correlation profiling; and (iii) analyzing proteoforms and complexes in cells lacking regulators of the target protein. We applied this strategy to YAP1, a transcriptional co-activator for the control of organ size and tissue homeostasis that is highly phosphorylated and among the most connected proteins in human cells. We identified multiple YAP1 phosphosites associated with distinct complexes and inferred how both are controlled by Hippo pathway members. We detected a PTPN14/LATS1/YAP1 complex and suggest a model how PTPN14 inhibits YAP1 via augmenting WW domain-dependent complex formation and phosphorylation by LATS1/2.

Published
2023

14. Effective Constitutive Relations for Simulating CO2 Capillary Trapping in Heterogeneous Reservoirs with Fluvial Sedimentary Architecture

Author

Gershenzon, Naum I., Ritzi Jr., Robert W., Dominic, David F., and Mehnert, Edward

Subjects
Physics - Geophysics
Abstract

Carbon dioxide (CO2) storage reservoirs commonly exhibit sedimentary architecture that reflects fluvial deposition. The heterogeneity in petrophysical properties arising from this architecture influences the dynamics of injected CO2. We previously used a geocellular modeling approach to represent this heterogeneity, including heterogeneity in constitutive saturation relationships. The dynamics of CO2 plumes in fluvial reservoirs was investigated during and after injection. It was shown that small-scale (centimeter to meter) features play a critical role in capillary trapping processes and have a primary effect on physical- and dissolution-trapping of CO2, and on the ultimate distribution of CO2 in the reservoir. Heterogeneity in saturation functions at that small scale enhances capillary trapping (snap off), creates capillary pinning, and increases the surface area of the plume. The understanding of these small-scale trapping processes from previous work is here used to develop effective saturation relationships that represent, at a larger scale, the integral effect of these processes. While it is generally not computationally feasible to represent the small-scale heterogeneity directly in a typical reservoir simulation, the effective saturation relationships for capillary pressure and relative permeability presented here, along with an effective intrinsic permeability, allow better representation of the total physical trapping at the scale of larger model grid cells, as typically used in reservoir simulations, and thus the approach diminishes limits on cell size and decreases simulation time in reservoir simulations., Comment: 31 pages, 18 figures

Published
2022

15. Expanding access to early phase trials: the CATCH-UP.2020 experience

Author

Baranda, Joaquina C, Diaz, Francisco J, Rubinstein, Larry, Shields, Anthony F, Dayyani, Farshid, Mehta, Amitkumar, Mehnert, Janice M, Trent, Jonathan, Mabaera, Rodwell, Mooney, Margaret, Moscow, Jeffrey A, Doroshow, James, Waters, Brittany, Ivy, Percy, Gore, Steven D, and Thomas, Alexandra

Subjects
Biomedical and Clinical Sciences, Clinical Sciences, Clinical Research, Patient Safety, Rural Health, Cancer, Clinical Trials and Supportive Activities, Humans, COVID-19, Ethnicity, Minority Groups, Neoplasms, Pandemics, Clinical Trials as Topic, Oncology and carcinogenesis
Abstract

BackgroundDisparities in cancer outcomes persist for underserved populations; one important aspect of this is limited access to promising early phase clinical trials. To address this, the National Cancer Institute-funded Create Access to Targeted Cancer Therapy for Underserved Populations (CATCH-UP.2020) was created. We report the tools developed and accrual metrics of the initial year of CATCH-UP.2020 with a focus on racial, ethnic, geographic, and socioeconomically underserved populations.MethodsCATCH-UP.2020 is a P30 supplement awarded to 8 National Cancer Institute-designated cancer centers with existing resources to rapidly open and accrue to Experimental Therapeutics Clinical Trials Network (ETCTN) trials with emphasis on engaging patients from underserved populations. Sites used patient-based, community-based, investigator-based, and program-based tools to meet specific program goals.ResultsFrom September 2020 to August 2021, CATCH-UP.2020 sites opened 45 ETCTN trials. Weighted average trial activation time for the 7 sites reporting this was 107 days. In the initial year, sites enrolled 145 patients in CATCH-UP.2020 with 68 (46.9%) representing racial, ethnic, rural, and socioeconomically underserved populations using the broader definition of underserved encompassed in the grant charge. During the initial year of CATCH-UP.2020, a time impacted by the COVID-19 pandemic, 15.8% (66 of 417) and 21.4% (31 of 145) of patients enrolled to ETCTN trials at network and at CATCH-UP sites, respectively, were from racial and ethnic minority groups, a more limited definition of underserved for which comparable data are available.ConclusionTargeted funding accelerated activation and accrual of early phase trials and expanded access to this therapeutic option for underserved populations.

Published
2023

17. Temporal stability of optimism and pessimism (LOT-R) over 6 years in the general population

Author

Andreas Hinz, Michael Friedrich, Heide Glaesmer, Barbara Brendel, Yuriy Nesterko, Jochen Ernst, Matthias L. Schroeter, and Anja Mehnert-Theuerkauf

Subjects
optimism, pessimism, longitudinal study, temporal stability, reliability, Psychology, BF1-990
Abstract

ObjectiveThe aims of this study were to examine changes in habitual optimism over a six-year period and to analyze the relationship between changes in optimism and changes in other quality of life-related variables.MethodA randomly selected community sample of the German adult general population (N = 4,965) was surveyed twice, with a time interval of 6.04 years.ResultsDuring the course of the 6 years, the mean score of the LOT-R total scale improved (effect size d = 0.11). The temporal stability in terms of the test–retest correlation was r = 0.61 for the total sample. There were only marginal gender differences in this temporal stability, however, the stability in the oldest age group ≥70 years (r = 0.50) was lower than the stability of the other age groups. The cross-sectional correlations showed clear relationships between optimism on the one hand and quality of life, life satisfaction, social support, and low levels of anxiety and physical complaints on the other. The corresponding longitudinal correlations between changes in optimism and changes in the other variables were less pronounced, but in the same direction.ConclusionThe study confirmed the applicability of the LOT-R in longitudinal studies. In samples with participants of 70 years and above, the limited stability in the optimism assessments needs to be considered in clinical practice and epidemiologic research.

Published
2024
Full Text
View/download PDF

18. Dimensionality Reduced Training by Pruning and Freezing Parts of a Deep Neural Network, a Survey

Author

Wimmer, Paul, Mehnert, Jens, and Condurache, Alexandru Paul

Subjects
Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition, Statistics - Machine Learning
Abstract

State-of-the-art deep learning models have a parameter count that reaches into the billions. Training, storing and transferring such models is energy and time consuming, thus costly. A big part of these costs is caused by training the network. Model compression lowers storage and transfer costs, and can further make training more efficient by decreasing the number of computations in the forward and/or backward pass. Thus, compressing networks also at training time while maintaining a high performance is an important research topic. This work is a survey on methods which reduce the number of trained weights in deep learning models throughout the training. Most of the introduced methods set network parameters to zero which is called pruning. The presented pruning approaches are categorized into pruning at initialization, lottery tickets and dynamic sparse training. Moreover, we discuss methods that freeze parts of a network at its random initialization. By freezing weights, the number of trainable parameters is shrunken which reduces gradient computations and the dimensionality of the model's optimization space. In this survey we first propose dimensionality reduced training as an underlying mathematical model that covers pruning and freezing during training. Afterwards, we present and discuss different dimensionality reduced training methods., Comment: This preprint has not undergone peer review or any post-submission improvements or corrections. The Version of Record of this article is published in Artificial Intelligence Review (2023), and is available online at https://doi.org/10.1007/s10462-023-10489-1

Published
2022
Full Text
View/download PDF

19. Interspace Pruning: Using Adaptive Filter Representations to Improve Training of Sparse CNNs

Author

Wimmer, Paul, Mehnert, Jens, and Condurache, Alexandru Paul

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

Unstructured pruning is well suited to reduce the memory footprint of convolutional neural networks (CNNs), both at training and inference time. CNNs contain parameters arranged in $K \times K$ filters. Standard unstructured pruning (SP) reduces the memory footprint of CNNs by setting filter elements to zero, thereby specifying a fixed subspace that constrains the filter. Especially if pruning is applied before or during training, this induces a strong bias. To overcome this, we introduce interspace pruning (IP), a general tool to improve existing pruning methods. It uses filters represented in a dynamic interspace by linear combinations of an underlying adaptive filter basis (FB). For IP, FB coefficients are set to zero while un-pruned coefficients and FBs are trained jointly. In this work, we provide mathematical evidence for IP's superior performance and demonstrate that IP outperforms SP on all tested state-of-the-art unstructured pruning methods. Especially in challenging situations, like pruning for ImageNet or pruning to high sparsity, IP greatly exceeds SP with equal runtime and parameter costs. Finally, we show that advances of IP are due to improved trainability and superior generalization ability., Comment: Accepted as conference paper for CVPR 2022

Published
2022

20. Retinal capillary perfusion heterogeneity in diabetic retinopathy detected by optical coherence tomography angiography

Author

Po Hsiang (Shawn) Yuan, Arman Athwal, Mena Shalaby, Andrew Mehnert, Dao-Yi Yu, Rony C. Preti, Marinko Sarunic, and Eduardo V. Navajas

Subjects
Diabetic retinopathy, Optical coherence tomography angiography (OCTA), Perfusion heterogeneity, Retinal imaging, Retinal perfusion, Segmentation, Ophthalmology, RE1-994
Abstract

Abstract Background Diabetic retinopathy (DR) is a leading cause of blindness and involves retinal capillary damage, microaneurysms, and altered blood flow regulation. Optical coherence tomography angiography (OCTA) is a non-invasive way of visualizing retinal vasculature but has not been used extensively to study blood flow heterogeneity. The purpose of this study is to detect and quantify blood flow heterogeneity utilizing en-face swept source OCTA in patients with DR. Methods This is a prospective clinical study which examined patients with either type 1 or 2 diabetes mellitus. Each included eye was graded clinically as no DR, mild DR, or moderate-severe DR. Ten consecutive en face 6 × 6 mm foveal SS-OCTA images were obtained from each eye using a PLEX Elite 9000 (Zeiss Meditec, Dublin, CA). Built-in fixation-tracking, follow-up functions were utilized to reduce motion artifacts and ensure same location imaging in sequential frames. Images of the superficial and deep vascular complexes (SVC and DVC) were arranged in temporal stacks of 10 and registered to a reference frame for segmentation using a deep neural network. The vessel segmentation was then masked onto each stack to calculate the pixel intensity coefficient of variance (PICoV) and map the spatiotemporal perfusion heterogeneity of each stack. Results Twenty-nine eyes were included: 7 controls, 7 diabetics with no DR, 8 mild DR, and 7 moderate-severe DR. The PICoV correlated significantly and positively with DR severity. In patients with DR, the perfusion heterogeneity was higher in the temporal half of the macula, particularly in areas of capillary dropout. PICoV also correlates as expected with the established OCTA metrics of perfusion density and vessel density. Conclusion PICoV is a novel way to analyze OCTA imaging and quantify perfusion heterogeneity. Retinal capillary perfusion heterogeneity in both the SVC and DVC increased with DR severity. This may be related to the loss of retinal capillary perfusion autoregulation in diabetic retinopathy.

Published
2024
Full Text
View/download PDF

31. High genetic connectivity of two pollinating flies despite urban disturbance

Author

Julian Wittische, Stéphanie Lippert, Amanda Luttringer, Hinatea Ariey, António Cruz, Balint Andrasi, Dylan Thissen, Anna Schleimer, Frank Drygala, Joerg Mehnert, Ximo Mengual, Lisette Cantú‐Salazar, and Alain C. Frantz

Subjects
Diptera, flower flies, landscape genetics, spatial ecology, urbanization, Ecology, QH540-549.5
Abstract

Abstract Hoverflies (Syrphidae) are essential pollinators, and their severe decline jeopardizes their invaluable contribution to plant diversity and agricultural production. However, we know little about the dispersal abilities of hoverflies in urbanized landscapes, limiting our understanding of the spatiotemporal dynamics of plant–pollinator systems and reducing our ability to preserve biodiversity in the context of global change. Previous work has not addressed how urbanization affects the functional connectivity of hoverflies and whether dispersal is a limiting factor in their population dynamics. In this study, we investigated the spatial genetic structure of two nonmigratory species of hoverflies in two urban areas. We collected more than a thousand specimens of each Syritta pipiens and Myathropa florea, each, by hand netting in two western European urbanized study areas of 490 and 460 km2 in 2021, and we genotyped them at 14 and 24 microsatellite loci, respectively. Based on spatial and nonspatial Bayesian clustering methods, we failed to reject the null hypothesis of panmixia, suggesting that both species exhibited high genetic connectivity despite urbanization. The distribution of allele frequencies was not correlated with geographic distance, implying that isolation‐by‐distance was negligible at the investigated spatial scale in both species. Although anthropogenic land cover changes generally have dramatic consequences on biodiversity, these hoverfly species retain high connectivity, which suggests that dispersal is not a strong limiting factor in their metapopulational dynamics. However, high effective population size and its confounding effect on signals of genetic drift may limit our ability to conclude confidently about landscape effects on gene flow in those two species. Provided we maintain or restore habitat, recolonization may be prompt even in urban areas.

Published
2024
Full Text
View/download PDF

32. Digital Therapeutic (Mika) Targeting Distress in Patients With Cancer: Results From a Nationwide Waitlist Randomized Controlled Trial

Author

Franziska Springer, Ayline Maier, Michael Friedrich, Jan Simon Raue, Gandolf Finke, Florian Lordick, Guy Montgomery, Peter Esser, Hannah Brock, and Anja Mehnert-Theuerkauf

Subjects
Computer applications to medicine. Medical informatics, R858-859.7, Public aspects of medicine, RA1-1270
Abstract

BackgroundDistress is highly prevalent among patients with cancer, but supportive care needs often go unmet. Digital therapeutics hold the potential to overcome barriers in cancer care and improve health outcomes. ObjectiveThis study conducted a randomized controlled trial to investigate the efficacy of Mika, an app-based digital therapeutic designed to reduce distress across the cancer trajectory. MethodsThis nationwide waitlist randomized controlled trial in Germany enrolled patients with cancer across all tumor entities diagnosed within the last 5 years. Participants were randomized into the intervention (Mika plus usual care) and control (usual care alone) groups. The participants completed web-based assessments at baseline and at 2, 6, and 12 weeks. The primary outcome was the change in distress from baseline to week 12, as measured by the National Comprehensive Cancer Network Distress Thermometer. Secondary outcomes included depression, anxiety (Hospital Anxiety and Depression Scale), fatigue (Functional Assessment of Chronic Illness Therapy-Fatigue), and quality of life (Clinical Global Impression-Improvement Scale). Intention-to-treat and per-protocol analyses were performed. Analyses of covariance were used to test for outcome changes over time between the groups, controlling for baseline. ResultsA total of 218 patients (intervention: n=99 and control: n=119) were included in the intention-to-treat analysis. Compared with the control group, the intervention group reported greater reductions in distress (P=.03; ηp²=0.02), depression (P

Published
2024
Full Text
View/download PDF

33. COPS: Controlled Pruning Before Training Starts

Author

Wimmer, Paul, Mehnert, Jens, and Condurache, Alexandru

Subjects
Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition
Abstract

State-of-the-art deep neural network (DNN) pruning techniques, applied one-shot before training starts, evaluate sparse architectures with the help of a single criterion -- called pruning score. Pruning weights based on a solitary score works well for some architectures and pruning rates but may also fail for other ones. As a common baseline for pruning scores, we introduce the notion of a generalized synaptic score (GSS). In this work we do not concentrate on a single pruning criterion, but provide a framework for combining arbitrary GSSs to create more powerful pruning strategies. These COmbined Pruning Scores (COPS) are obtained by solving a constrained optimization problem. Optimizing for more than one score prevents the sparse network to overly specialize on an individual task, thus COntrols Pruning before training Starts. The combinatorial optimization problem given by COPS is relaxed on a linear program (LP). This LP is solved analytically and determines a solution for COPS. Furthermore, an algorithm to compute it for two scores numerically is proposed and evaluated. Solving COPS in such a way has lower complexity than the best general LP solver. In our experiments we compared pruning with COPS against state-of-the-art methods for different network architectures and image classification tasks and obtained improved results., Comment: Accepted by The International Joint Conference on Neural Network (IJCNN) 2021

Published
2021
Full Text
View/download PDF

34. Quantitative study of spatial and temporal variation in retinal capillary network perfusion in rat eye by in vivo confocal imaging

Author

Paula Kun Yu, Andrew Mehnert, Jayden Brendan Dickson, Hassanain Qambari, Chandrakumar Balaratnasingam, Stephen Cringle, Dean Darcey, and Dao-Yi Yu

Subjects
Medicine, Science
Abstract

Abstract Microvascular dysfunction is the underlying pathological process in many systemic diseases. However, investigation into its pathogenesis is impeded by the accessibility and complexity of the microvasculature within different organs, particularly for the central nervous system. The retina as an extension of the cerebrum provides a glimpse into the brain through which the microvasculature can be observed. Two major questions remain unanswered: How do the microvessels regulate spatial and temporal delivery to satisfy the varying cellular demands, and how can we quantify blood perfusion in the 3D capillary network? Here, quantitative measurements of red blood cell (RBC) speed in each vessel in the field were made in the in vivo rat retinal capillary network using an ultrafast confocal technique with fluorescently labelled RBCs. Retinal RBC speed and number were found to vary remarkably between microvessels ranging from 215 to 6641 microns per second with significant variations spatially and temporally. Overall, the RBC speed was significantly faster in the microvessels in the superficial retina than in the deep retina (estimated marginal means of 2405 ± 238.2 µm/s, 1641 ± 173.0 µm/s respectively). These observations point to a highly dynamic nature of microvasculature that is specific to its immediate cellular environment and is constantly changing.

Published
2023
Full Text
View/download PDF

35. A phase 1b single-arm trial of intratumoral oncolytic virus V937 in combination with pembrolizumab in patients with advanced melanoma: results from the CAPRA study

Author

Silk, Ann W., O’Day, Steven J., Kaufman, Howard L., Bryan, Jennifer, Norrell, Jacqueline T., Imbergamo, Casey, Portal, Daniella, Zambrano-Acosta, Edwin, Palmeri, Marisa, Fein, Seymour, Wu, Cai, Guerreiro, Leslie, Medina, Daniel, Bommareddy, Praveen K., Zloza, Andrew, Fox, Bernard A., Ballesteros-Merino, Carmen, Ren, Yixin, Shafren, Darren, Grose, Mark, Vieth, Joshua A., and Mehnert, Janice M.

Published
2023
Full Text
View/download PDF

37. Indwelling catheter vs intermittent catheterization: is there a difference in UTI susceptibility?

Author

Vera Neumeier, Fabian P. Stangl, Joëlle Borer, Collene E. Anderson, Veronika Birkhäuser, Oksana Chemych, Oliver Gross, Miriam Koschorke, Jonas Marschall, Shawna McCallin, Ulrich Mehnert, Helen Sadri, Lara Stächele, Thomas M. Kessler, and Lorenz Leitner

Subjects
Asymptomatic bacteriuria, Intermittent catheterization, Indwelling catheter, Neurogenic lower urinary tract dysfunction, Urinary tract infection, Infectious and parasitic diseases, RC109-216
Abstract

Abstract Background Patients with neurogenic lower urinary tract dysfunction (NLUTD) often rely on some type of catheterization for bladder emptying. Intermittent catheterization (IC) is considered the gold standard and is preferred over continuous catheterization, since it is considered to cause fewer urinary tract infections (UTIs) than indwelling catheterization. The main objective of our study was to describe UTI prevalence (at visit) and incidence (within the last 12 months) and urine culture characteristics between patients using an indwelling catheter versus (vs) those performing IC. Methods In this cross-sectional study, we prospectively evaluated from 02/2020 to 01/2021 patients with NLUTD undergoing urine cultures for prophylactic reasons or due to UTI symptoms. At visit, all patients underwent a standardized interview on current UTI symptoms as well as UTI history and antibiotic consumption within the past year. Patients using an indwelling catheter (n = 206) or IC (n = 299) were included in the analysis. The main outcome was between-group differences regarding UTI characteristics. Results Patients using an indwelling catheter were older (indwelling catheter vs IC: median 66 (Q1-Q3: 55—77) vs 55 (42—67) years of age) and showed a higher Charlson comorbidity index (indwelling catheter vs IC: median 4 (Q1-Q3: 2–6) vs 2 (1–4) (both p

Published
2023
Full Text
View/download PDF

38. E-mental health in Germany — what is the current use and what are experiences of different types of health care providers for patients with mental illnesses?

Author

Elena Caroline Weitzel, Maria Schwenke, Georg Schomerus, Peter Schönknecht, Markus Bleckwenn, Anja Mehnert-Theuerkauf, Steffi G. Riedel-Heller, and Margrit Löbner

Subjects
E-mental-health, Health care apps, Implementation, Routine care, Public aspects of medicine, RA1-1270
Abstract

Abstract Background As a new and effective support option, e-mental health interventions can be useful in complementing treatment in mental health care. To date, little is known about how health care providers use these programs to treat patients with mental illnesses in Germany. The present study aims to examine the use of and experiences with e-mental health interventions from the point of view of different types of health care providers for patients with mental illnesses. Methods Data from a cross-sectional survey of routine care health care providers in Germany in 2021 were analysed. In this survey, data were collected from n = 107 general practitioners (GPs), n = 114 specialist doctors, n = 102 psychotherapists, and n = 102 inpatient clinicians. Assessments included professional use of digital media, as well as knowledge, use and experiences regarding e-mental health interventions in care of people with mental illness. Results In the total sample of n = 425, 65.6% (n = 279) were female. The study participants had an average age of 47.7 years (SD = 11.0) and their average work experience was 20.0 years (SD = 11.1). Overall, the majority (83.8%, n = 353) had heard of e-mental health interventions, but few felt well informed. Only 28.5% (n = 121) had already used e-mental health interventions for treatment support. The most commonly recommended e-mental health interventions in the sample were deprexis (39.7%, n = 48), moodgym (24.8%, n = 30), and iFightDepression (22.3%, n = 27). The use was predominantly considered to be helpful and satisfactory. Insufficient knowledge about e-mental health interventions and lack of informational materials for patients were reported as relevant barriers to the use of e-mental health interventions. Conclusions E-mental health interventions can be a useful support option, but they are rarely used in the treatment of patients with mental illnesses. There is a need to disseminate information specific to the various types of health care providers. Tailored implementation strategies need to be developed in order to capitalize on the potential of effective e-mental health interventions and to improve health care for patients with mental illnesses.

Published
2023
Full Text
View/download PDF

39. Benign tumours of foot and ankle

Author

Hagen Fritzsche, Anne Weidlich, Klaus-Dieter Schaser, Elisabeth Mehnert, Doreen Winkler, and Stefan Rammelt

Subjects
foot and ankle, benign tumours, soft tissue tumours, bone tumours, treatment modalities, Orthopedic surgery, RD701-811
Abstract

Musculoskeletal tumours of foot or ankle make up about 4–5% of all musculoskeletal tumours. Fortunately, about 80% of them are benign. However, due to the rarity and low prevalence of each single tumour entity, diagnosis is often difficult and delayed. Ultrasonography is an important diagnostic tool to safely recognize ganglion cysts as a frequently encountered ‘bump’ in the foot. In suspicious lesions, malignancy must be excluded histologically in a tumour center by biopsy after imaging procedures using x-ray, computed tomography (CT) and magnetic resonance imaging (MRI). Most of the benign tumours do not require any further surgical therapy. Resection should be performed in the case of locally aggressive tumour growth or local symptoms of discomfort. In contrast to malignant tumours, the primary purpose in the resection is the least possible loss of function.

Published
2023
Full Text
View/download PDF

40. Identifying patients at risk for myasthenic crisis with hemogram and inflammation-related laboratory parameters – a pilot study

Author

Anne Mehnert, Sivan Bershan, Jil Kollmus-Heege, Lea Gerischer, Meret Luise Herdick, Sarah Hoffmann, Sophie Lehnerer, Franziska Scheibe, Frauke Stascheit, Maike Stein, Alastair M. Buchan, Andreas Meisel, Annette Aigner, and Philipp Mergenthaler

Subjects
myasthenia gravis, myasthenia, myasthenic crisis, hemogram, laboratory parameters, inflammation, Neurology. Diseases of the nervous system, RC346-429
Abstract

BackgroundMyasthenia gravis (MG) is a rare autoimmune disease characterized by fatigable weakness of the voluntary muscles and can exacerbate to life-threatening myasthenic crisis (MC), requiring intensive care treatment. Routine laboratory parameters are a cost-effective and widely available method for estimating the clinical outcomes of several diseases, but so far, such parameters have not been established to detect disease progression in MG.MethodsWe conducted a retrospective analysis of selected laboratory parameters related to inflammation and hemogram for MG patients with MC compared to MG patients without MC. To identify potential risk factors for MC, we applied time-varying Cox regression for time to MC and, as a sensitivity analysis, generalized estimating equations logistic regression for the occurrence of MC at the next patient visit.Results15 of the 58 examined MG patients suffered at least one MC. There was no notable difference in the occurrence of MC by antibody status or sex. Both regression models showed that higher counts of basophils (per 0.01 unit increase: HR = 1.32, 95% CI = 1.02–1.70), neutrophils (per 1 unit increase: HR = 1.40, 95% CI = 1.14–1.72), potentially leukocytes (per 1 unit increase: HR = 1.15, 95% CI = 0.99–1.34), and platelets (per 100 units increase: HR = 1.54, 95% CI = 0.99–2.38) may indicate increased risk for a myasthenic crisis.ConclusionThis pilot study provides proof of the concept that increased counts of basophils, neutrophils, leukocytes, and platelets may be associated with a higher risk of developing MC in patients with MG.

Published
2024
Full Text
View/download PDF

41. Return to work, work productivity loss and activity impairment in Chinese breast cancer survivors 12-month post-surgery: a longitudinal study

Author

Danielle Wing Lam Ng, Serana Chun Yee So, Richard Fielding, Anja Mehnert-Theuerkauf, Ava Kwong, Dacita Suen, Ling Wong, Sara Wai Wun Fung, Oi Kwan Chun, Daniel Y. T. Fong, Sharon Chan, Alex Molasiotis, Winnie K. W. So, and Wendy Wing Tak Lam

Subjects
return to work, employment, work productivity, activity impairment, breast cancer, work condition, Public aspects of medicine, RA1-1270
Abstract

IntroductionExisting evidence of returning-to-work (RTW) after cancer comes predominately from Western settings, with none prospectively examined since the initial diagnostic phase. This study prospectively documents RTW-rate, time-to-RTW, work productivity loss, and activity impairment, within the first-year post-surgery among Chinese women with breast cancer (BCW) and identify potential causal co-variants.MethodsThis observational longitudinal study followed 371 Chinese BCW who were employed/self-employed at the time of diagnosis at 4-week post-surgery (baseline). RTW-status and time-to-RTW were assessed at baseline (T1), 4-month (T2), 6-month (T3), and 12-month (T4) post-baseline. WPAI work productivity loss and activity impairment were assessed at T4. Baseline covariates included demographics, medical-related factors, work satisfaction, perceived work demand, work condition, RTW self-efficacy, B-IPQ illness perception, COST financial well-being, EORTC QLQ-C30 and QLQ-BR23 physical and psychosocial functioning, and HADS psychological distress.ResultsA 68.2% RTW-rate (at 12-month post-surgery), prolonged delay in RTW (median = 183 days), and significant proportions of T4 work productivity loss (20%), and activity impairment (26%), were seen. BCW who were blue-collar workers with lower household income, poorer financial well-being, lower RTW self-efficacy, poorer job satisfaction, poorer illness perception, greater physical symptom distress, impaired physical functioning, and unfavorable work conditions were more likely to experience undesired work-related outcomes.DiscussionUsing a multifactorial approach, effective RTW interventions should focus on not only symptom management, but also to address psychosocial and work-environmental concerns. An organizational or policy level intervention involving a multidisciplinary team comprising nurses, psychologists, occupational health professionals, and relevant stakeholders in the workplace might be helpful in developing a tailored organizational policy promoting work-related outcomes in BCW.

Published
2024
Full Text
View/download PDF

42. The validity of the distress thermometer in patients with musculoskeletal tumors

Author

Finn Marie Uhlenbruch, Nikolas Schopow, Elisabeth Roschke, Christian Lycke, Christoph-Eckhard Heyde, Anja Mehnert-Theuerkauf, and Georg Osterhoff

Subjects
Cancer, Tumor, Musculoskeletal, Distress, Anxiety, Depression, Diseases of the musculoskeletal system, RC925-935, Neoplasms. Tumors. Oncology. Including cancer and carcinogens, RC254-282
Abstract

Background: Visits to an outpatient cancer clinic represent a challenging situation for patients, which can trigger anxiety and helplessness in those affected. It is important to identify patients with high psychological distress as early as possible in order to provide them with supportive psychological interventions. The aim of this study was to validate the Distress Thermometer (DT), a widely used screening for distress, in a cohort of patients with musculoskeletal tumors and to explore associations between distress, treatment satisfaction and health literacy. Methods: All patients presenting to a University outpatient clinic for musculoskeletal cancers were asked to complete a set of questionnaires including the DT), the Hospital Anxiety and Depression Scale (HADS) as a comparison scale, the Patient Satisfaction with Comprehensive Cancer Care (SCCC) and European Health Literacy Survey Questionnaire (HLS-EU-Q16).To assess the sensitivity and specificity of the DT in a cohort of patients with musculoskeletal tumors, we compare the performance of the DT in relation to an established screener for anxiety and depression using receiver operating characteristics (ROC) analyses. Results: A total of 120 patients (age 58 ± 18, 51% female) were analyzed. Patients reported a mean DT of 5.0 (SD 2.3, range, 0 to 10). Eighty-six patients (71.7 %) had a DT score ≥ 5 indicating moderate or severe psychological distress.The mean total HADS score (scale 0 to42 points) was 11.7 (SD 7.6, range, 0 to 32) with a HADS score of ≥ 15 in 29.2% of patients. The DT correlated moderately with anxiety and depression (HADS total r = 0.48, p

Published
2024
Full Text
View/download PDF

43. FreezeNet: Full Performance by Reduced Storage Costs

Author

Wimmer, Paul, Mehnert, Jens, and Condurache, Alexandru

Subjects
Computer Science - Machine Learning, Computer Science - Computer Vision and Pattern Recognition
Abstract

Pruning generates sparse networks by setting parameters to zero. In this work we improve one-shot pruning methods, applied before training, without adding any additional storage costs while preserving the sparse gradient computations. The main difference to pruning is that we do not sparsify the network's weights but learn just a few key parameters and keep the other ones fixed at their random initialized value. This mechanism is called freezing the parameters. Those frozen weights can be stored efficiently with a single 32bit random seed number. The parameters to be frozen are determined one-shot by a single for- and backward pass applied before training starts. We call the introduced method FreezeNet. In our experiments we show that FreezeNets achieve good results, especially for extreme freezing rates. Freezing weights preserves the gradient flow throughout the network and consequently, FreezeNets train better and have an increased capacity compared to their pruned counterparts. On the classification tasks MNIST and CIFAR-10/100 we outperform SNIP, in this setting the best reported one-shot pruning method, applied before training. On MNIST, FreezeNet achieves 99.2% performance of the baseline LeNet-5-Caffe architecture, while compressing the number of trained and stored parameters by a factor of x 157., Comment: Conference Paper of the Asian Conference on Computer Vision (ACCV) 2020

Published
2020
Full Text
View/download PDF

44. A Geometrically Exact Continuum Framework for Light-Matter Interaction in Photo-Active Polymers I. Variational Setting

Author

Mehnert, M, Oates, W, and Steinmann, P

Subjects
Computer Science - Computational Engineering, Finance, and Science
Abstract

Molecular photo-switches as, e.g., azobenzene molecules allow, when embedded into a polymeric matrix, for photo-active polymer compounds responding mechanically when exposed to light of certain wavelength. Photo-mechanics, i.e. light-matter interaction in photo-active polymers holds great promise for, e.g., remote and contact-free activation of photo-driven actuators. In a series of earlier contributions, Oates et al. developed a successful continuum formulation for the coupled electric, electronic and mechanical problem capturing azobenzene polymer compounds, thereby mainly focussing on geometrically linearized kinematics. Building on that formulation, we here explore the variational setting of a geometrically exact continuum framework based on Dirichlet's and Hamilton's principle as well as, noteworthy, Hamilton's equations. Thereby, when treating the dissipative case, we resort to incremental versions of the various variational problems via suited incorporation of a dissipation potential. In particular, the Hamiltonian setting of geometrically exact photo-mechanics is up to now largely under-explored even for the energetic case, arguably since the corresponding Lagrangian is degenerate in Dirac's sense. Moreover, in general, the Hamiltonian setting of dissipative dynamical systems is a matter of ongoing debate per se. In this contribution, by advocating a novel incremental version of the Hamiltonian setting exemplified for the dissipative case of photo-mechanics, we aim to also unify the variational approach to dissipative dynamical systems. Taken together, the variational setting of a geometrically exact continuum framework of photo-mechanics paves the way for forthcoming theoretical and numerical analyses.

Published
2020

45. Targeting colorectal cancer cells using AND-gated adaptor RevCAR T-cells

Author

Karla E. G. Soto, Liliana R. Loureiro, Tabea Bartsch, Claudia Arndt, Alexandra Kegler, Nicola Mitwasi, Laura Drewitz, Lydia Hoffmann, Haidy A. Saleh, Eugenia Crespo, Maria Mehnert, Cansu Daglar, Hinrich Abken, Frank Momburg, Michael Bachmann, and Anja Feldmann

Subjects
colorectal cancer, CAR T-cells, AND-gate targeting, CEA, EpCAM, Immunologic diseases. Allergy, RC581-607
Abstract

Despite the success of chimeric antigen receptor (CAR) T-cells especially for treating hematological malignancies, critical drawbacks, such as “on-target, off-tumor” toxicities, need to be addressed to improve safety in translating to clinical application. This is especially true, when targeting tumor-associated antigens (TAAs) that are not exclusively expressed by solid tumors but also on hea9lthy tissues. To improve the safety profile, we developed switchable adaptor CAR systems including the RevCAR system. RevCAR T-cells are activated by cross-linking of bifunctional adaptor molecules termed target modules (RevTM). In a further development, we established a Dual-RevCAR system for an AND-gated combinatorial targeting by splitting the stimulatory and co-stimulatory signals of the RevCAR T-cells on two individual CARs. Examples of common markers for colorectal cancer (CRC) are the carcinoembryonic antigen (CEA) and the epithelial cell adhesion molecule (EpCAM), while these antigens are also expressed by healthy cells. Here we describe four novel structurally different RevTMs for targeting of CEA and EpCAM. All anti-CEA and anti-EpCAM RevTMs were validated and the simultaneous targeting of CEA+ and EpCAM+ cancer cells redirected specific in vitro and in vivo killing by Dual-RevCAR T-cells. In summary, we describe the development of CEA and EpCAM specific adaptor RevTMs for monospecific and AND-gated targeting of CRC cells via the RevCAR platform as an improved approach to increase tumor specificity and safety of CAR T-cell therapies.

Published
2023
Full Text
View/download PDF

46. Prevalence of mental disorders, psychosocial distress and need for psychosocial support in cancer patients – study protocol of an epidemiological multi-center study

Author

Mehnert Anja, Koch Uwe, Schulz Holger, Wegscheider Karl, Weis Joachim, Faller Hermann, Keller Monika, Brähler Elmar, and Härter Martin

Subjects
Psychiatry, RC435-571
Abstract

Abstract Background Empirical studies investigating the prevalence of mental disorders and psychological distress in cancer patients have gained increasing importance during recent years, particularly with the objective to develop and implement psychosocial interventions within the cancer care system. Primary purpose of this epidemiological cross-sectional multi-center study is to detect the 4-week-, 12-month-, and lifetime prevalence rates of comorbid mental disorders and to further assess psychological distress and psychosocial support needs in cancer patients across all major tumor entities within the in- and outpatient oncological health care and rehabilitation settings in Germany. Methods/Design In this multicenter, epidemiological cross-sectional study, cancer patients across all major tumor entities will be enrolled from acute care hospitals, outpatient cancer care facilities, and rehabilitation centers in five major study centers in Germany: Freiburg, Hamburg, Heidelberg, Leipzig and Würzburg. A proportional stratified random sample based on the nationwide incidence of all cancer diagnoses in Germany is used. Patients are consecutively recruited in all centers. On the basis of a depression screener (PHQ-9) 50% of the participants that score below the cutoff point of 9 and all patients scoring above are assessed using the Composite International Diagnostic Interview for Oncology (CIDI-O). In addition, all patients complete validated questionnaires measuring emotional distress, information and psychosocial support needs as well as quality of life. Discussion Epidemiological data on the prevalence of mental disorders and distress provide detailed and valid information for the estimation of the demands for the type and extent of psychosocial support interventions. The data will provide information about specific demographic, functional, cancer- and treatment-related risk factors for mental comorbidity and psychosocial distress, specific supportive care needs and use of psychosocial support offers.

Published
2012
Full Text
View/download PDF

47. An assessment of microvascular hemodynamics in human macula

Author

Dao-Yi Yu, Andrew Mehnert, Chandrakumar Balaratnasingam, Paula K. Yu, Martin Hein, Dong An, and Stephen J. Cringle

Subjects
Medicine, Science
Abstract

Abstract An adequate blood supply to meet the energy demands is essential for any tissue, particularly for high energy demand tissues such as the retina. A critical question is: How is the dynamic match between neuronal demands and blood supply achieved? We present a quantitative assessment of temporal and spatial variations in perfusion in the macular capillary network in 10 healthy human subjects using a non-invasive and label-free imaging technique. The assessment is based on the calculation of the coefficient of variation (CoV) of the perfusion signal from arterioles, venules and capillaries from a sequence of optical coherence tomography angiography images centred on the fovea. Significant heterogeneity of the spatial and temporal variation was found within arterioles, venules and capillary networks. The CoV values of the capillaries and smallest vessels were significantly higher than that in the larger vessels. Our results demonstrate the presence of significant heterogeneity of spatial and temporal variation within each element of the macular microvasculature, particularly in the capillaries and finer vessels. Our findings suggest that the dynamic match between neuronal demands and blood supply is achieved by frequent alteration of local blood flow evidenced by capillary perfusion variations both spatially and temporally in the macular region.

Published
2023
Full Text
View/download PDF

48. The 'Great Debate' at Melanoma Bridge 2022, Naples, December 1st–3rd, 2022

Author

Paolo A. Ascierto, Christian Blank, Alexander M. Eggermont, Claus Garbe, Jeffrey E. Gershenwald, Omid Hamid, Axel Hauschild, Jason J. Luke, Janice M. Mehnert, Jeffrey A. Sosman, Hussein A. Tawbi, Mario Mandalà, Alessandro Testori, Corrado Caracò, Iman Osman, and Igor Puzanov

Subjects
Melanoma, Immunotherapy, Anti-PD-1, Targeted therapy, Adjuvant, Neoadjuvant, Medicine
Abstract

Abstract The Great Debate session at the 2022 Melanoma Bridge congress (December 1–3) featured counterpoint views from leading experts on five contemporary topics of debate in the management of melanoma. The debates considered the choice of anti-lymphocyte-activation gene (LAG)-3 therapy or ipilimumab in combination with anti-programmed death (PD)-1 therapy, whether anti-PD-1 monotherapy is still acceptable as a comparator arm in clinical trials, whether adjuvant treatment of melanoma is still a useful treatment option, the role of adjuvant therapy in stage II melanoma, what role surgery will continue to have in the treatment of melanoma. As is customary in the Melanoma Bridge Great Debates, the speakers are invited by the meeting Chairs to express one side of the assigned debate and the opinions given may not fully reflect personal views. Audiences voted in favour of either side of the argument both before and after each debate.

Published
2023
Full Text
View/download PDF

49. Empirical retinal venous pulse wave velocity using modified photoplethysmography

Author

Anmar Abdul-Rahman, William Morgan, Aleksandar Vukmirovic, Andrew Mehnert, Danail Obreschow, and Dao-Yi Yu

Subjects
Ophthalmodynamometry, Retinal venous pulse, Modified photoplethysmography, Medicine, Biology (General), QH301-705.5, Science (General), Q1-390
Abstract

Abstract Objective Using the novel imaging method of high-speed modified photoplethysmography we measured the retinal venous pulse wave velocity in a single case. Results A healthy 30-year-old subject underwent high-speed modified photoplethysmography (120 frames per second) with simultaneous ophthalmodynamometry at 26 Meditron units. A video of the optic nerve was analyzed using custom software. A harmonic regression model was fitted to each pixel in the time series and used to quantify the retinal vascular pulse wave parameters. Retinal venous pulsation at the optic disc was observed as a complex dynamic wall motion, whereas contraction commenced at a point in the vein at the center of the optic disc, and progressed centrifugally. The empirically estimated retinal venous pulse wave velocity at this segment was approximately 22.24694 mm/s. This measurement provides an estimate for future studies in the field.

Published
2023
Full Text
View/download PDF

50. Phosphorylation‐linked complex profiling identifies assemblies required for Hippo signal integration

Author

Federico Uliana, Rodolfo Ciuffa, Ranjan Mishra, Andrea Fossati, Fabian Frommelt, Sabrina Keller, Martin Mehnert, Eivind Salmorin Birkeland, Frank van Drogen, Nevena Srejic, Matthias Peter, Nicolas Tapon, Ruedi Aebersold, and Matthias Gstaiger

Subjects
Interaction proteomics, Protein phosphorylation, Protein–protein interactions, YAP1, Biology (General), QH301-705.5, Medicine (General), R5-920
Abstract

Abstract While several computational methods have been developed to predict the functional relevance of phosphorylation sites, experimental analysis of the interdependency between protein phosphorylation and Protein–Protein Interactions (PPIs) remains challenging. Here, we describe an experimental strategy to establish interdependencies between protein phosphorylation and complex formation. This strategy is based on three main steps: (i) systematically charting the phosphorylation landscape of a target protein; (ii) assigning distinct proteoforms of the target protein to different protein complexes by native complex separation (AP‐BNPAGE) and protein correlation profiling; and (iii) analyzing proteoforms and complexes in cells lacking regulators of the target protein. We applied this strategy to YAP1, a transcriptional co‐activator for the control of organ size and tissue homeostasis that is highly phosphorylated and among the most connected proteins in human cells. We identified multiple YAP1 phosphosites associated with distinct complexes and inferred how both are controlled by Hippo pathway members. We detected a PTPN14/LATS1/YAP1 complex and suggest a model how PTPN14 inhibits YAP1 via augmenting WW domain‐dependent complex formation and phosphorylation by LATS1/2.

Published
2023
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results