Lebenswissenschaften und Ethik
Refine
Year of publication
Document Type
- Article (31)
- conference proceeding (presentation, abstract) (17)
- conference proceeding (article) (8)
- Preprint (2)
- Report (1)
Language
- English (59) (remove)
Is part of the Bibliography
- no (59)
Keywords
- Deep Learning (9)
- Artificial Intelligence (5)
- Diagnose (5)
- Künstliche Intelligenz (5)
- Maschinelles Lernen (5)
- Deep learning (4)
- Adenocarcinoma (3)
- Barrett's esophagus (3)
- Computerunterstützte Medizin (3)
- Speiseröhrenkrankheit (3)
Institute
- Regensburg Center of Health Sciences and Technology - RCHST (59) (remove)
Begutachtungsstatus
- peer-reviewed (42)
- begutachtet (5)
Objective: Artificial intelligence (AI) may reduce underdiagnosed or overlooked upper GI (UGI) neoplastic and preneoplastic conditions, due to subtle appearance and low disease prevalence. Only disease-specific AI performances have been reported, generating uncertainty on its clinical value.
Design: We searched PubMed, Embase and Scopus until July 2020, for studies on the diagnostic performance of AI in detection and characterisation of UGI lesions. Primary outcomes were pooled diagnostic accuracy, sensitivity and specificity of AI. Secondary outcomes were pooled positive (PPV) and negative (NPV) predictive values. We calculated pooled proportion rates (%), designed summary receiving operating characteristic curves with respective area under the curves (AUCs) and performed metaregression and sensitivity analysis.
Results: Overall, 19 studies on detection of oesophageal squamous cell neoplasia (ESCN) or Barrett's esophagus-related neoplasia (BERN) or gastric adenocarcinoma (GCA) were included with 218, 445, 453 patients and 7976, 2340, 13 562 images, respectively. AI-sensitivity/specificity/PPV/NPV/positive likelihood ratio/negative likelihood ratio for UGI neoplasia detection were 90% (CI 85% to 94%)/89% (CI 85% to 92%)/87% (CI 83% to 91%)/91% (CI 87% to 94%)/8.2 (CI 5.7 to 11.7)/0.111 (CI 0.071 to 0.175), respectively, with an overall AUC of 0.95 (CI 0.93 to 0.97). No difference in AI performance across ESCN, BERN and GCA was found, AUC being 0.94 (CI 0.52 to 0.99), 0.96 (CI 0.95 to 0.98), 0.93 (CI 0.83 to 0.99), respectively. Overall, study quality was low, with high risk of selection bias. No significant publication bias was found.
Conclusion: We found a high overall AI accuracy for the diagnosis of any neoplastic lesion of the UGI tract that was independent of the underlying condition. This may be expected to substantially reduce the miss rate of precancerous lesions and early cancer when implemented in clinical practice.
Exoskeletons were invented over 100 years ago but have only become popular in the last two decades, especially in the working industry as they can decrease work-related loads significantly. The most often used exoskeletons are for the lower back and shoulder since these are commonly affected body regions. All devices have in common that their purpose is to reduce internal loads of vulnerable body regions. Nevertheless, there is still little understanding on how biomechanical loading in the human body changes when exoskeletons are used. Therefore, further analyses are needed. A promising candidate for these are musculoskeletal models, which are based on an inverse dynamics approach and can calculate external parameters such as ground reaction forces or other interaction forces as well as internal parameters such as joint reaction forces or muscle activities. The various examples in the literature show that these models are increasingly used for assessing the biomechanical effects of exoskeletons on the human body. Furthermore, musculoskeletal models can calculate biomechanical loadings of humans with and without exoskeletons for all kinds of applications and allow an evaluation of their purpose.
Practical Relevance: This article highlights the possibilities of musculoskeletal models for assessing the design and efficiency of occupational exoskeletons. Several practical use cases are described along with distinct descriptions of common implications of musculoskeletal and exoskeleton modeling.
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.
Background
Older adults in care facilities face a high risk of experiencing depression. The impact that early interventions like biographical work have on the quality of life for older adults in such facilities is unknown.
Aim
To develop and evaluate a tablet-supported intervention for biographical work in long-term residential aged care to increase the quality of life for older adults.
Design
The study will be conducted in a randomized pretest–posttest control group design with follow-up testing in group and single settings. Participants will be randomized to the experimental intervention (tablet-supported biographic work) or the control intervention (planned tablet-supported game playing), each guided by senior volunteers. A total of 80 residents and 16 volunteers will be recruited. The primary outcome for the residents and volunteers will be quality of life as measured with the World Health Organization Quality of Life Assessment-for older adults. Secondary measures will be self-esteem and life satisfaction. In addition, we will examine residents’ ability to communicate and their functional independence.
Method
The first stage of the project involves developing an app. The app is developed in a user-centered, agile development process. It will use multimedia to prepare life history topics and links them to key questions. Next, a workshop is developed for the volunteers who accompany the use of the app in the institutions. During the second phase, biographic work stimulated by the app will be conducted in groups or individually with residents.
Discussion
This is the first known program tailored to older adults in care facilities and senior volunteers that aims to prevent depression by providing digitally supported biographic work.
Based on previous work by our group with manual annotation of visible Barrett oesophagus (BE) cancer images, a real-time deep learning artificial intelligence (AI) system was developed. While an expert endoscopist conducts the endoscopic assessment of BE, our AI system captures random images from the real-time camera livestream and provides a global prediction (classification), as well as a dense prediction (segmentation) differentiating accurately between normal BE and early oesophageal adenocarcinoma (EAC). The AI system showed an accuracy of 89.9% on 14 cases with neoplastic BE.
Computer-aided diagnosis using deep learning in the evaluation of early oesophageal adenocarcinoma
(2019)
Computer-aided diagnosis using deep learning (CAD-DL) may be an instrument to improve endoscopic assessment of Barrett’s oesophagus
(BE) and early oesophageal adenocarcinoma (EAC). Based on still images from two databases, the diagnosis of EAC by CAD-DL reached sensitivities/specificities of 97%/88% (Augsburg data) and 92%/100% (Medical Image Computing and Computer-Assisted Intervention [MICCAI]
data) for white light (WL) images and 94%/80% for narrow band images (NBI) (Augsburg data), respectively. Tumour margins delineated by
experts into images were detected satisfactorily with a Dice coefficient (D) of 0.72. This could be a first step towards CAD-DL for BE assessment. If developed further, it could become a useful
adjunctive tool for patient management.
Background and aims: The accurate differentiation between T1a and T1b Barrett’s cancer has both therapeutic and prognostic implications but is challenging even for experienced physicians. We trained an Artificial Intelligence (AI) system on the basis of deep artificial neural networks (deep learning) to differentiate between T1a and T1b Barrett’s cancer white-light images.
Methods: Endoscopic images from three tertiary care centres in Germany were collected retrospectively. A deep learning system was trained and tested using the principles of cross-validation. A total of 230 white-light endoscopic images (108 T1a and 122 T1b) was evaluated with the AI-system. For comparison, the images were also classified by experts specialized in endoscopic diagnosis and treatment of Barrett’s cancer.
Results: The sensitivity, specificity, F1 and accuracy of the AI-system in the differentiation between T1a and T1b cancer lesions was 0.77, 0.64, 0.73 and 0.71, respectively. There was no statistically significant difference between the performance of the AI-system and that of human experts with sensitivity, specificity, F1 and accuracy of 0.63, 0.78, 0.67 and 0.70 respectively.
Conclusion: This pilot study demonstrates the first multicenter application of an AI-based system in the prediction of submucosal invasion in endoscopic images of Barrett’s cancer. AI scored equal to international experts in the field, but more work is necessary to improve the system and apply it to video sequences and in a real-life setting. Nevertheless, the correct prediction of submucosal invasion in Barret´s cancer remains challenging for both experts and AI.
The evaluation and assessment of Barrett’s esophagus is challenging for both expert and nonexpert endoscopists. However, the early diagnosis of cancer in Barrett’s esophagus is crucial for its prognosis, and could save costs. Pre-clinical and clinical studies on the application of Artificial Intelligence (AI) in Barrett’s esophagus have shown promising results. In this review, we focus on the current challenges and future perspectives of implementing AI systems in the management of patients with Barrett’s esophagus.
The growing number of publications on the application of artificial intelligence (AI) in medicine underlines the enormous importance and potential of this emerging field of research.
In gastrointestinal endoscopy, AI has been applied to all segments of the gastrointestinal tract most importantly in the detection and characterization of colorectal polyps. However, AI research has been published also in the stomach and esophagus for both neoplastic and non-neoplastic disorders.
The various technical as well as medical aspects of AI, however, remain confusing especially for non-expert physicians.
This physician-engineer co-authored review explains the basic technical aspects of AI and provides a comprehensive overview of recent publications on AI in gastrointestinal endoscopy. Finally, a basic insight is offered into understanding publications on AI in gastrointestinal endoscopy.
Autologous lipotransfer is a promising method for tissue regeneration, because white adipose tissue contains a heterogeneous cell population, including mesenchymal stem cells, endothelial cells, immune cells, and adipocytes. In order to improve the outcome, adipose tissue can be processed before application. In this study, we investigated changes caused by mechanical processing.
Lipoaspirates were processed using sedimentation, first-time centrifugation, shear-force homogenization, and second-time centrifugation. The average adipocyte size, stromal vascular cell count, and adipocyte depot size were examined histologically at every processing step. In addition, the adipose derived stem cells (ADSCs) were isolated and differentiated osteogenically and adipogenically. While homogenization causes a disruption of adipocyte depots, the shape of the remaining adipocytes is not changed. On average, these adipocytes are smaller than the depot adipocytes, they are surrounded by the ECM, and therefore mechanically more stable. The volume loss of adipocyte depots leads to a significant enrichment of stromal vascular cells such as ADSCs. However, the mechanical processing does not change the potential of the ADSCs to differentiate adipogenically or steogenically. It thus appears that mechanically processed lipoaspirates are promising for the reparation of even mechanically stressed tissue as that found in nasolabial folds. The changes resulting from the processing correspond more to a filtration of mechanically less stable components than to a manipulation of the tissue.
To avoid dislocation of the shoulder joint after reverse total shoulder arthroplasty, it is important to achieve sufficient shoulder stability when placing the implant components during surgery. One parameter for assessing shoulder stability can be shoulder stiffness. The aim of this research was to develop a temporary reverse shoulder implant prototype that would allow intraoperative measurement of shoulder stiffness while varying the position of the implant components. Joint angle and torque measurement techniques were developed to determine shoulder stiffness. Hall sensors were used to measure the joint angles by converting the magnetic flux densities into angles. The accuracy of the joint angle measurements was tested using a test bench. Torques were determined by using thin-film pressure sensors. Various mechanical mechanisms for variable positioning of the implant components were integrated into the prototype. The results of the joint angle measurements showed measurement errors of less than 5° in a deflection range of ±15° adduction/abduction combined with ±45° flexion/extension. The proposed design provides a first approach for intra-operative assessment of shoulder stiffness. The findings can be used as a technological basis for further developments.
Older adults in long-term care homes are at high risk of experiencing
reduced quality of life (QoL) and depression. Technology-assisted biography work can have a positive impact on QoL and mood, but there is little research on its use with this target group. The purpose of this paper is to examine the effect of tablet-based biography work conducted by volunteers on the QoL of residents and volunteers. A pretest-posttest control group design with an intervention period of 3 months and a 3-month follow-up was used. Results show a significant increase in participation for volunteers and residents after
the intervention, which is stable for residents until follow-up. Volunteers also show significant improvement in mental QoL immediately after the intervention. There were no significant effects for life satisfaction, self-esteem, or depression. No significant changes were found for the control group. Digitally conducted tablet-based biography work appears to have effects on QoL-associated outcomes.
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.
Background:
People with aphasia (PWA) often suffer from reduced participation and quality of life. Nevertheless, there are currently only a few specific interventions that respond to this problem. Participation and quality of life could be increased by interacting with peers who have similar experiences. Digital social networks could stimulate an autonomous interaction. However, digital social networks need to be adapted to the specific needs of PWA. Therefore, a participatory, agile process involving the target group should be chosen to develop such a olution, i.e., an app. The research project consists of a total of three phases. In the first phase—app development—the app was developed and programmed including the target group. In the second phase—app testing—the usability and user-friendliness of the app were evaluated with four PWA. In the third phase—feasibility and preliminary effcacy—that will be described in the article, the impact of the app on PWA will be evaluated.
Aims:
The overarching aim of our study is to provide preliminary effcacy of the intervention. Digital social interaction with other PWA can lead to increased social integration. In addition to digital interaction, personal encounters between PWA should be encouraged. As a result, we expect an improvement in quality of life of PWA. Additionally, we focus on identification of the most appropriate measurements to discover changes associated with the intervention.
Methods:
The evaluation, which is described in this paper, takes place in a pre-test - post-test design with a total of n = 48 PWA. Participants will be recruited in regional clusters to facilitate face-to-face meetings. Half of the participants will be assigned to the delayed intervention group and the other half to the immediate intervention group. Participants in the delayed intervention group will go through a 3-month waiting period before using the app, while the participants of the immediate intervention group will start using the app for 3 months right away. Inclusion criteria are the presence of chronic aphasia (at least 6 months) and possession of a smartphone with internet access. Questionnaires on quality of life (SAQOL-39, GHQ-12), depression (GDS, DISCs), communicative participation (CPIB), and social support (F-SozU) will be conducted at inclusion (t0), after 3 months of app use (t1), and after another 3 months for follow-up (t2). Participants in the delayed intervention group will be assessed twice before the intervention, before the 3-month waiting period (t0a) and after the waiting period (t0b). In addition to the quantitative measures, interviews will take place with 6 to 8 selected participants after 3 months of app use. Responses will be analysed using Thematic Analysis.
Discussion:
The app will be the first social network tool that is systematically developed with PWA. Initial indications from the first phases are that the app can be used by PWA, so that the evaluation of this app version can take place in the third phase. Results of this study can provide an initial indication of whether social network support is a suitable intervention. Findings will help provide information on the feasibility of digital connectivity for PWA. Preliminary findings on its impact on the participation and quality of life of PWA could be made available.