معرفی
Henry Zoller is a Researcher affiliated with the Department of Hydraulic Engineering at ETH Zürich under the Professorship for Hydraulic Structures. His work focuses on developing deep learning methods for medical image analysis, particularly in Cardiac Magnetic Resonance Imaging (MRI). He specializes in automated valve motion assessment to enhance diagnostic accuracy for diastolic dysfunction.
Research Interests
- Computer Vision applications in healthcare
- Biomedical image analysis for cardiac diagnostics
- Deep learning models for medical imaging tasks
His recent work includes a novel CNN-based system for mitral valve landmark detection in 4CHV CINE MRI sequences, achieving sub-pixel accuracy. This method improves slice tracking during MRI acquisitions, reducing manual intervention and enhancing flow quantification. The system was validated using datasets from the Cardiac Atlas Project and Siemens Healthineers.
Technical Contributions
- Developed a two-stage CNN architecture combining heatmap regression and non-linear refinement
- Integrated deep learning frameworks in Python/PyTorch for medical imaging
Future work aims to extend the algorithm to 2CHV views and optimize for real-time clinical integration.
