A5067635650- Research Data Management Practices
- Scientific Computing and Data Management
- Semantic Web and Ontologies
- Biomedical Text Mining and Ontologies
- Data Quality and Management
- Privacy-Preserving Technologies in Data
- Bioinformatics and Genomic Networks
- AI in cancer detection
- Machine Learning in Healthcare
- Electronic Health Records Systems
- Artificial Intelligence in Healthcare and Education
- Explainable Artificial Intelligence (XAI)
- Gene expression and cancer classification
- Radiomics and Machine Learning in Medical Imaging
- Ethics and Social Impacts of AI
- Artificial Intelligence in Healthcare
- Adversarial Robustness in Machine Learning
- Genetic Associations and Epidemiology
- Blockchain Technology Applications and Security
- Cancer Genomics and Diagnostics
- Generative Adversarial Networks and Image Synthesis
- Chronic Disease Management Strategies
- Health, Environment, Cognitive Aging
- COVID-19 diagnosis using AI
- Computational Drug Discovery Methods
University Hospital Cologne
2021-2025
University of Cologne
2021-2025
Fraunhofer Institute for Applied Information Technology
2017-2024
RWTH Aachen University
2016-2023
Institut für Medizinische Informatik, Biometrie und Epidemiologie
2023
University Hospital Leipzig
2022
Fraunhofer Society
2020
Inform (Germany)
2019
Ollscoil na Gaillimhe – University of Galway
2013-2016
Middle East Technical University
2010
In this paper <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup> , we proposed an explainable deep neural networks (DNN)-based method for automatic detection of COVID-19 symptoms from chest radiography (CXR) images, which call 'DeepCOVIDExplainer'. We used 15,959 CXR images 15,854 patients, covering normal, pneumonia, and cases. are first comprehensively preprocessed augmented before classifying with a ensemble method, followed by...
Summary Introduction: This article is part of the Focus Theme Methods Information in Medicine on German Medical Informatics Initiative. “Smart Technology for Healthcare (SMITH)” one four consortia funded by Initiative (MI-I) to create an alliance universities, university hospitals, research institutions and IT companies. SMITH’s goals are establish Data Integration Centers (DICs) at each SMITH partner hospital implement use cases which demonstrate usefulness approach. Objectives: To give...
Interference between pharmacological substances can cause serious medical injuries. Correctly predicting so-called drug-drug interactions (DDI) does not only reduce these cases but also result in a reduction of drug development cost. Presently, most drug-related knowledge is the clinical evaluations and post-marketing surveillance; resulting limited amount information. Existing data-driven prediction approaches for DDIs typically rely on single source information, while using information...
In recent years, as newer technologies have evolved around the healthcare ecosystem, more and data been generated. Advanced analytics could power collected from numerous sources, both institutions, or generated by individuals themselves via apps devices, lead to innovations in treatment diagnosis of diseases; improve care given patient; empower citizens participate decision-making process regarding their own health well-being. However, sensitive nature prohibits organizations sharing data....
Abstract Although rare diseases (RDs) affect over 260 million individuals worldwide, low data quality and scarcity challenge effective care research. This work aims to harmonise the Common Data Set by European Rare Disease Registry Infrastructure, Health Level 7 Fast Healthcare Interoperability Base Resources, Global Alliance for Genomics Phenopacket Schema into a novel disease common model (RD-CDM), laying foundation developing international RD-CDMs aligned with these standards. We...
In recent years, data-driven medicine has gained increasing importance in terms of diagnosis, treatment, and research due to the exponential growth health care data. However, data protection regulations prohibit centralisation for analysis purposes because potential privacy risks like accidental disclosure third parties. Therefore, alternative usage policies, which comply with present guidelines, are particular interest.We aim enable analyses on sensitive patient by simultaneously complying...
The explosion of data-driven applications using Artificial Intelligence (AI) in recent years has given rise to a variety ethical issues regarding data collection, annotation, and processing mostly opaque algorithms, as well the interpretation employment results AI pipeline. ubiquity negatively impacts sensitive areas, ranging from discrimination against vulnerable populations privacy invasion environmental cost that these algorithms entail, puts into focus on ever present domain ethics. In...
Medical real-world data stored in clinical systems represents a valuable knowledge source for medical research, but its usage is still challenged by various technical and cultural aspects. Analyzing these challenges suggesting measures future improvement are crucial to improve the situation. This comment paper such an analysis from perspective of research.
Abstract Digitization of medicine requires systematic handling the increasing amount health data to improve medical diagnosis. In this context, integration versatile diagnostic information, e.g., from anamnesis, imaging, histopathology, and clinical chemistry, its comprehensive analysis by artificial intelligence (AI)–based tools is expected precision therapeutic conduct. However, complex environment poses a major obstacle translation integrated diagnostics into research routine. There high...
Amid the coronavirus disease(COVID-19) pandemic, humanity experiences a rapid increase in infection numbers across world. Challenge hospitals are faced with, fight against virus, is effective screening of incoming patients. One methodology assessment chest radiography(CXR) images, which usually requires expert radiologist's knowledge. In this paper, we propose an explainable deep neural networks(DNN)-based method for automatic detection COVID-19 symptoms from CXR call DeepCOVIDExplainer. We...
Osteoarthritis (OA) is a degenerative joint disease, which significantly affects middle-aged and elderly people. Although primarily identified via hyaline cartilage change based on medical images, technical bottlenecks like noise, artifacts, modality impose an enormous challenge high-precision, objective, efficient early quantification of OA. Owing to recent advancements, approaches neural networks (DNNs) have shown outstanding success in this application domain. However, due nested...
Gradient inversion attacks can reconstruct the victim's private data once they have access to model and gradient. However, existing research is still immature, many are conducted in ideal conditions. It unclear how damaging such really be effectively defended. In this paper, we first summarize current relevant researches their limitations. Then design a general gradient attack framework, which both FedSGD FedAVG. We propose approaches enhance label inference image restoration, respectively....
ABSTRACT Blockchain technology has the potential to extend beyond its traditional use in cryptocurrency and make significant strides critical sectors like healthcare. Clinical research, which plays a pivotal role enhancing healthcare quality by guiding activities, determining equipment usage, recommending preferred medications, stands benefit greatly from blockchain integration. The unique technical capabilities of offer promising solutions across various phases clinical study design patient...
In the digital age, data has emerged as one of most valuable assets across various sectors, including academia, industry, and healthcare. Effective preservation involves management to ensure its long-term accessibility usability. Given importance sensitivity data, need for effective is a crucial necessity. One big recent proposed approaches FAIR Digital Objects (FDOs) which revolutionize field preservation. Central this revolution alignment FDOs with principles (Findable, Accessible,...
The dielectric properties of biological tissues characterise the interaction human with electromagnetic (EM) fields. Accurate knowledge are vital in EM-based therapeutic and diagnostic techniques, for assessing safety wireless devices. Despite importance these properties, field has suffered from inconsistencies reported data. measurement process is known to be affected by both confounders clinical confounders; however, adequate metadata often lacking literature. For this reason, work...
The Oxford Classification for IgA nephropathy is the most successful example of an evidence-based nephropathology classification system. aim our study was to replicate glomerular components scoring with end-to-end deep learning pipeline that involves automatic segmentation followed by mesangial hypercellularity (M), endocapillary (E), segmental sclerosis (S) and active crescents (C).A total number 1056 periodic acid-Schiff (PAS) whole slide images (WSIs), coming from 386 kidney biopsies,...
Abstract An accurate diagnosis and prognosis for cancer are specific to patients with particular types molecular traits, which needs address carefully. The discovery of important biomarkers is becoming an step toward understanding the mechanisms carcinogenesis in genomics data clinical outcomes need be analyzed before making any decision. Copy number variations (CNVs) found associated risk individual cancers hence can used reveal genetic predispositions develops. In this paper, we collect...
Cancer is one of the deadliest diseases caused by abnormal behaviors genes that control cell division and growth. Genomics data clinical outcomes from multiplatform heterogeneous sources are used to make decisions for cancer patients, where both multimodality heterogeneity impose significant challenges bioinformatics tools algorithms. Numerous works have been proposed overcome these using sophisticated machine learning algorithms as either primary or supporting tools. In this paper, we...
Unlike traditional central training, federated learning (FL) improves the performance of global model by sharing and aggregating local models rather than data to protect users' privacy. Although this training approach appears secure, some research has demonstrated that an attacker can still recover private based on shared gradient information. This on-the-fly reconstruction attack deserves be studied in depth because it occur at any stage whether beginning or end training; no relevant...
Currently, the medical field is witnessing an increase in use of machine learning techniques. Supervised methods adopted classification, prediction, and segmentation tasks for images always experience decreased performance when training testing datasets do not follow independent identically distributed assumption. These distribution shift situations seriously influence applications’ robustness, fairness, trustworthiness domain. Hence, this article, we adopt CycleGAN (generative adversarial...