A5018331097- Cell Image Analysis Techniques
- Digital Imaging for Blood Diseases
- Digital Holography and Microscopy
- AI in cancer detection
- Blood properties and coagulation
- Image Processing Techniques and Applications
- Cancer Genomics and Diagnostics
- Machine Learning and Data Classification
- Music and Audio Processing
- Video Analysis and Summarization
- Image and Signal Denoising Methods
- Sepsis Diagnosis and Treatment
- Brain Tumor Detection and Classification
- Privacy-Preserving Technologies in Data
- Education Methods and Technologies
- Image Enhancement Techniques
- Cancer, Hypoxia, and Metabolism
- Inflammatory Biomarkers in Disease Prognosis
- Electrical and Bioimpedance Tomography
- COVID-19 diagnosis using AI
- Explainable Artificial Intelligence (XAI)
- Venous Thromboembolism Diagnosis and Management
- Antiplatelet Therapy and Cardiovascular Diseases
- Anomaly Detection Techniques and Applications
- Seismic Imaging and Inversion Techniques
Technical University of Munich
2020-2025
The clinical spectrum of acute SARS-CoV-2 infection ranges from an asymptomatic to life-threatening disease. Considering the broad severity, reliable biomarkers are required for early risk stratification and prediction outcomes. Despite numerous efforts, no COVID-19-specific biomarker has been established guide further diagnostic or even therapeutic approaches, most likely due insufficient validation, methodical complexity, economic factors. COVID-19-associated coagulopathy is a hallmark...
Immunothrombosis is a critical aspect affecting patients in acute care settings, especially conditions involving infection, trauma, or severe inflammation. The interplay between hemostasis, innate immunity, and inflammation becomes highly relevant these scenarios can significantly impact patient outcomes. Blood cell aggregates are potential functional cellular biomarkers for prognostic predictive of immunothrombosis but require, due to the low logistical stability aggregates, point-of-care...
For several years, the determination of a differential cell count raw milk sample has been proposed as more accurate tool for monitoring udder health dairy cows compared with using absolute somatic count. However, required preparation and staining process can be labor- cost-intensive. Therefore, aim our study was to demonstrate feasibility analyzing unlabeled blood leukocytes from by means digital holographic microscopy (DHM). this, we trained three different machine learning methods, i.e.,...
The quality of datasets plays a crucial role in the successful training and deployment deep learning models. Especially medical field, where system performance may impact health patients, clean are safety requirement for reliable predictions. Therefore, outlier detection is an essential process when building autonomous clinical decision systems. In this work, we assess suitability Self-Organizing Maps specifically on dataset containing quantitative phase images white blood cells. We detect...
Label-free approaches are attractive in cytological imaging due to their flexibility and cost efficiency. They supported by machine learning methods, which, despite the lack of labeling associated lower contrast, can classify cells with high accuracy where human observer has little chance discriminate cells. In order better integrate these workflows into clinical decision making process, this work investigates calibration confidence estimation for automated classification leukocytes....
Manual blood smear analysis remains the gold standard to diagnose hematological disorders and infections of parasites. However, interpretation peripheral smears requires expert users, is time consuming, depends on inter-observer variation, not compatible with a high-throughput workflow for clinical routine diagnostics (Dunning & Safo, Biotech. Histochem. 2011, 86, 69–75; Pierre, Clin. Lab. Med., 2002, 22, 279–297). Instead, automated hematology analyzers only flag atypical results which...