A5006735571- AI in cancer detection
- Cell Image Analysis Techniques
- Colorectal Cancer Screening and Detection
- Radiomics and Machine Learning in Medical Imaging
- Digital Imaging for Blood Diseases
- Advanced Image and Video Retrieval Techniques
- Medical Image Segmentation Techniques
- Image Retrieval and Classification Techniques
- Image Processing Techniques and Applications
- Voice and Speech Disorders
- Digital Radiography and Breast Imaging
- Robotics and Sensor-Based Localization
- Advanced Vision and Imaging
- Esophageal Cancer Research and Treatment
- Surgical Simulation and Training
- Gastric Cancer Management and Outcomes
- 3D Surveying and Cultural Heritage
- Augmented Reality Applications
- Advanced X-ray and CT Imaging
- Anatomy and Medical Technology
- Image and Signal Denoising Methods
- Cardiac Valve Diseases and Treatments
- Colorectal Cancer Surgical Treatments
- Biomedical Text Mining and Ontologies
- Speech and Audio Processing
Fraunhofer Institute for Integrated Circuits
2016-2025
Friedrich-Alexander-Universität Erlangen-Nürnberg
2007-2024
Forschungsinstitut für Anwendungsorientierte Wissensverarbeitung
2021
Fraunhofer Society
2006-2013
Universitätsklinikum Aachen
2013
Fraunhofer Institute for Integrated Systems and Device Technology
2012
Universitätsklinikum Erlangen
1999-2000
University of Oregon
1971
Mammography is the most effective method for breast cancer screening available today. However, low positive predictive value of biopsy resulting from mammogram interpretation leads to approximately 70% unnecessary biopsies with benign outcomes. To reduce high number biopsies, several computer‐aided diagnosis (CAD) systems have been proposed in last years. These help physicians their decision perform a on suspicious lesion seen or short term follow‐up examination instead. We present two novel...
Although mammographic density is an established risk factor for breast cancer, its use limited in clinical practice because of a lack automated and standardized measurement methods. The aims this study were to evaluate variety texture features mammograms as factors cancer compare them with the percentage (PMD) by using case-control design. A including 864 cases 418 controls was analyzed automatically. Four hundred seventy explored possible cancer. These included statistical features,...
Background . Computed X-ray tomography (CTX) is a high-end nondestructive approach for the visual assessment of root architecture in soil. Nevertheless, order to evaluate high-resolution CTX data architectures, manual segmentation depicted systems from large-scale volume currently necessary, which both time consuming and error prone. The duration such importance, especially time-resolved growth analysis, where several instances plant need be segmented evaluated. Specifically, our...
A digital high-speed camera system for the endoscopic examination of larynx allows recording speeds up to 5,600 frames/s. Recordings 1 s duration can be stored and used further evaluation. Combined with an image processing program is able render x-t diagrams vocal cord movement. Data acquired from different locations each plotted separately. All known objective parameters voice derived highspeed glottograms.
Mammographic percent density (MD) is recognized as one of the strongest risk factors associated with breast cancer. This matched case-control study investigated whether MD represents an independent factor. Mammograms were obtained from 1025 cancer patients and 520 healthy controls. was measured using a quantitative computer-based threshold method (0-100%). Breast had higher than controls (38 vs. 32%, P<0.01). significantly in association such age over 60 years, body mass index (BMI) 25-30...
For many breast cancer (BC) risk factors, there is growing evidence concerning molecular subtypes for which the factor specific. With regard to mammographic density (MD), are inconsistent data its association with estrogen receptor (ER) and progesterone (PR) expression. The aim of our study was analyze between ER PR expression MD. In case-only study, on BC hormone MD were available 2,410 patients incident BC. assessed as percent (PMD) using a semiautomated method by two readers every...
Modern techniques for medical diagnostics and therapy in keyhole surgery scenarios as well technical inspection make use of flexible endoscopes. Their characteristic bendable image conductor consists a very limited number coated fibers, which leads to so-called comb structure. This effect has negative impact on further processing steps such feature tracking because these overlaid structures are wrongly detected features. With respect tasks, we propose an automatic approach generate optimal...
Differential blood count is a standard method in hematological laboratory diagnosis. In the course of developing computer-assisted microscopy system for generation differential counts, detection and segmentation white red cells forms an essential step its exactness fundamental prerequisite effectiveness subsequent classification step. We propose exact leukocytes erythrocytes simultaneous cooperative way. combine pixel-wise with template matching to locate use level-set approach order get...
Oral presentations Session 1: Entry & uncoating O1 Host cell polo-like kinases (PLKs) promote early prototype foamy virus (PFV) replication Irena Zurnic, Sylvia Hütter, Ute Lehmann, Nicole Stanke, Juliane Reh, Tobias Kern, Fabian Lindel, Gesche Gerresheim, Martin Hamann, Erik Müllers, Paul Lesbats, Peter Cherepanov, Serrao, Alan Engelman, Dirk Lindemann O2 A novel entry/uncoating assay reveals the presence of at least two species viral capsids during synchronized HIV-1 infection Claire Da...
The delta-retrovirus Human T-cell leukemia virus type 1 (HTLV-1) preferentially infects CD4+ T-cells via cell-to-cell transmission. Viruses are transmitted by polarized budding and transfer of viral biofilms at the virological synapse (VS). Formation VS requires Tax protein polarization host cytoskeleton, however, molecular mechanisms HTLV-1 transmission remain incompletely understood. Recently, we could show Tax-dependent upregulation actin-bundling Fascin (FSCN-1) in HTLV-1-infected...
Abstract Early detection of polyps is one central goal colonoscopic screening programs. To support gastroenterologists during this examination process, deep convolutional neural network can be applied for computer-assisted neoplastic lesions. In work, a Mask R-CNN architecture was applied. For training and testing, three independent colonoscopy data sets were used, including 2484 HD labelled images with from our clinic, as well two public image the MICCAI 2015 polyp challenge, consisting 612...
<title>Abstract</title> <bold>Background</bold> Agricultural systems are under extreme pressure to meet the global food demand, hence necessitating faster crop improvement. Rapid evaluation of crops using novel imaging technologies coupled with robust image analysis could accelerate these improvements. This study assesses potential X-ray for non-destructive rice grain traits. By analyzing 2D images paddy grains, we aim approximate their key physical traits important in breeding: (1)...
This article shows some examples of applications digital high-speed imaging in phoniatrics. Advantages and disadvantages as compared to the current world-wide standard, videostroboscopy, are discussed well possible future developments this area. Parts paper were presented by authors at Pan European Voice Conference Stockholm 2001 (Pevoc IV).
The pUL97 protein kinase encoded by human cytomegalovirus is a multifunctional determinant of the efficiency viral replication and phosphorylates as well cellular substrate proteins. Here, we report that expressed in two isoforms with molecular masses approximately 90 100 kDa. ORF UL97 comprises an unusual coding strategy five in-frame ATG start codons are contained within N-terminal 157 aa. Site-directed mutagenesis, transient expression point deletion mutants proteomic analyses accumulated...
The ability to adapt quickly a new task or data distribution based on only few examples is challenge in AI and highly relevant for various domains. In digital pathology, slight variations the scanning staining process can lead shift that provokes significant performance degradation of classical neural networks tasks like tissue cartography where reliable classification essential. To overcome this problem, we propose few-shot learning technique, specifically k-means extension Prototypical...