Regensburg Center of Health Sciences and Technology - RCHST
Refine
Document Type
- conference proceeding (article) (16) (remove)
Is part of the Bibliography
- no (16)
Keywords
- Deep Learning (2)
- Aerosols (1)
- Breathing cycle (1)
- Computerunterstützte Medizin (1)
- Convolutional neural networks (1)
- Deep learning fluid mechanics (1)
- Endoskopie (1)
- Flow visualisation (1)
- Fluid Mechanics (1)
- Machine Learning (1)
Institute
- Regensburg Center of Health Sciences and Technology - RCHST (16)
- Fakultät Angewandte Sozial- und Gesundheitswissenschaften (8)
- Fakultät Maschinenbau (5)
- Labor Logopädie (LP) (5)
- Regensburg Center of Biomedical Engineering - RCBE (5)
- Fakultät Informatik und Mathematik (4)
- Labor Biofluidmechanik (4)
- Labor Physiotherapie (LPh) (4)
- Regensburg Medical Image Computing (ReMIC) (3)
- Labor Biomechanik (LBM) (1)
Begutachtungsstatus
- peer-reviewed (8)
- begutachtet (1)
Blood flow in channels of varying diameters <500μm exhibits strong non-linear effects. Multiphase finite volume approaches are feasible, but still computationally costly. Here, the feasibility of applying convolutional neural networks for blood flow prediction in artificial lungs is investigated. Training targets are precomputed using an Eulerian two-phase approach. To match with experimental data, the interphase drag and lift, as well as intraphase shear-thinning are adapted. A recursively branching regression network and convolution/deconvolution networks with plain skip connections and densely connected skips are investigated. A priori knowledge is incorporated in the loss functional to prevent the network from learning non-physical solutions. Inference from neural networks is approximately six orders of magnitude faster than the classical finite volume approach. Even if resulting in comparably coarse flow fields, the neural network predictions can be used as close to convergence initial solutions greatly accelerating classical flow computations.
Convolutional Neural Networks for Approximation of Internal Non-Newtonian Multiphase Flow Fields
(2021)
Neural networks (NNs) as an alternative method for universal approximation of differential equations have proven to be computationally efficient and still sufficiently accurate compared to established methods such as the finite volume method (FVM). Additionally, analysing weights and biases can give insights into the underlying physical laws. FVM and NNs are both based upon spacial discretisation. Since a Cartesian and equidistant grid is a raster graphics, image-to-image regression techniques can be used to predict phase velocity fields as well as particle and pressure distributions from simple mass flow boundary conditions. The impact of convolution layer depth and number of channels of a ConvolutionDeconvolution Regression Network (CDRN), on prediction performance of internal non-Newtownian multiphase flows is investigated. Parametric training data with 2055 sets is computed using FVM. To capture significant non-Newtownian effects of a particle-laden fluid (e.g. blood) flowing through small and non-straight channels, an Euler-Euler multiphase approach is used. The FVM results are normalized and mapped onto an equidistant grid as supervised learning target. The investigated NNs consist of n= {3, 5, 7} corresponding encoding/decoding blocks and different skip connections. Regardless of the convolution depth (i.e. number of blocks), the deepest spacial down-sampling via strided convolution is adjusted to result in a 1 × 1 × f · 2nfeature map, with f = {8, 16, 32}. The prediction performance expressed is as channel-averaged normalized root mean squared error (NRMSE). With a NRMSE of < 2 · 10-3, the best preforming NN has f = 32 initial feature maps, a kernel size of k = 4, n = 5 blocks and dense skip connections. Average inference time from this NN takes < 7 · 10-3s. Worst accuracy at NRMSE of approx 9 · 10-3is achieved without any skips, at k = 2, f = 16 and n = 3, but deployment takes only < 2 · 10-3s Given an adequate training, the prediction accuracy improves with convolution depth, where more features have higher impact on deeper NNs. Due to skip connections and batch normalisation, training is similarly efficient, regardless of the depth. This is further improved by blocks with dense connections, but at the price of a drastically larger model. Depending on geometrical complexity, spacial resolution is critical, as it increases the number of learnables and memory requirements massively.
Barrett's esophagus denotes a disorder in the digestive system that affects the esophagus' mucosal cells, causing reflux, and showing potential convergence to esophageal adenocarcinoma if not treated in initial stages. Thus, fast and reliable computer-aided diagnosis becomes considerably welcome. Nevertheless, such approaches usually suffer from imbalanced datasets, which can be addressed through Generative Adversarial Networks (GANs). Such techniques generate realistic images based on observed samples, even though at the cost of a proper selection of its hyperparameters. Many works employed a class of nature-inspired algorithms called metaheuristics to tackle the problem considering distinct deep learning approaches. Therefore, this paper's main contribution is to introduce metaheuristic techniques to fine-tune GANs in the context of Barrett's esophagus identification, as well as to investigate the feasibility of generating high-quality synthetic images for early-cancer assisted identification.
The utilization of virtual reality (VR) technology has shown promise in various therapeutic applications, particularly in exposure therapy
for reducing fear of certain situations objects or activities, e.g. fear of height, or negative evaluation of others in social situations. VR has been shown to yield positive outcomes in follow-up studies, and provides a safe and ecological therapeutic environment for therapists and their patients. This paper presents a collaborative
effort to develop a VR speech therapy system which simulates a virtual audience for users to practice their public speaking skills. We describe a novel web-based graphica user interface that enables
therapists to manage the therapy session using a simple timeline. Lastly, we present the results from a qualitative study with therapists and teachers with functional dysphonia, which highlight the potential of such an application to support and augment the therapists’ work and the remaining challenges regarding the design of natural interactions, agent behaviours and scenario customisation for patients.
Vorgestellt wird das Forschungsvorhaben des interdisziplinären Forschungsprojekts „DeinHaus 4.0 – Telepräsenzroboter für die Pflege und Unterstützung von Schlaganfallpatientinnen und -patienten (TePUS)“. Über geeignete Robotersysteme sollen den Proband*innen individuell abgestimmte Angebote aus den Bereichen Telenursing und Teletherapie
zur Verfügung gestellt werden. Hierbei werden einerseits Telepräsenzsitzungen durchgeführt, anderseits kommen verschiedene Apps zum Einsatz. Die Interventionen werden mittels eines Mixed-Methods-Ansatzes evaluiert. Begleitend werden Technikakzeptanz und Nutzungsbereitschaft untersucht.
Pixel-level classification is an essential part of computer vision. For learning from labeled data, many powerful deep learning models have been developed recently. In this work, we augment such supervised segmentation models by allowing them to learn from unlabeled data. Our semi-supervised approach, termed Error-Correcting Supervision, leverages a collaborative strategy. Apart from the supervised training on the labeled data, the segmentation network is judged by an additional network. The secondary correction network learns on the labeled data to optimally spot correct predictions, as well as to amend incorrect ones. As auxiliary regularization term, the corrector directly influences the supervised training of the segmentation network. On unlabeled data, the output of the correction network is essential to create a proxy for the unknown truth. The corrector’s output is combined with the segmentation network’s prediction to form the new target. We propose a loss function that incorporates both the pseudo-labels as well as the predictive certainty of the correction network. Our approach can easily be added to supervised segmentation models. We show consistent improvements over a supervised baseline on experiments on both the Pascal VOC 2012 and the Cityscapes datasets with varying amounts of labeled data.
Hintergrund/Fragestellung:
Eine Rheumatoide Arthritis (RA) verläuft chronisch schubweise und Betroffene leiden unter zunehmenden und schwankenden Symptomen. In der Physiotherapie (PT) werden individuell abgestimmte Interventionen in einem ambulanten (aPT) oder Gruppensetting (gPT) eingesetzt. In Leitlinien wird PT unspezifisch empfohlen. Ärzte verordnen spezifische Formen der PT. Wie ist eine evidenzbasierte und am Bedarf aus Sicht Betroffener orientierte Versorgung zu gestalten?
Methoden/Material:
Drei Methoden wurden kombiniert: 1) Evidenz zu PT wurde mittels eines Systematischen Reviews ermittelt und deren Qualität in Anlehnung an GRADE beurteilt. 2) Einflussfaktoren auf eine Versorgung wurden mittels Sekundäranalyse von Versorgungsdaten aus einer Längsschnittstudie bestimmt. 3) Einflussfaktoren auf den Bedarf an sowie die Verordnung von PT aus Sicht der Betroffenen wurden mittels Fragebogen erhoben. Die Daten wurden soweit möglich mittels Metaanalysen (1), binären (2) sowie multiplen linearen Regressionen und Gruppenvergleichen für abhängige Stichproben (3) analysiert. Es wurden Häufigkeiten (3) der Nennung spezifischer Interventionen sowie von Übereinstimmungen zwischen Bedarf und wahrgenommenem Verordnungsverhalten bestimmt.
Ergebnisse:
1) Eine gPT zeigte bei jungen Betroffenen mit guter Funktion Effekte, war aber bei Vorschäden der Gelenke weder effektiv noch sicher. Zu aPT liegt keine Evidenz vor. 2) Der einzige relevante Vorhersagefaktor für eine Versorgung mit PT ist eine bereits bestehende Versorgung (Exp(B)=10,81 - 171,53, p<0,001). 3) Den stärksten bedarfssteigernden Einfluss hat ein schlechtes körperliches Befinden (B=0,25, p<0,05/MWD 1,3 (NAS 1-10), p<0,001). Teilnehmer (n=305) gaben am häufigsten Bedarf an aPT an, bei schlechtem Befinden vor allem an passiven Maßnahmen. In besseren Phasen besteht zusätzlich Bedarf an gPT. Betroffene haben individuell unterschiedliche Bedarfe und eine Berücksichtigung bedarfsbestimmender Faktoren bei der Verordnung wird kaum wahrgenommen.
Schlussfolgerungen:
Angesichts der geringen Evidenzlage und individueller und schwankender Bedarfe wird ein stark individualisiertes Vorgehen empfohlen. Effektivität, Sicherheit und Passgenauigkeit sind regelmäßig gemeinsam mit den an der Versorgung Beteiligten sowie den Betroffenen zu überprüfen. Insbesondere bei schlechtem Befinden sollte über die Möglichkeit einer aPT und die notwendige Überprüfung aufgeklärt werden. Es besteht dringender Forschungsbedarf für aPT bei RA.