Magdalena Wysocki

Magdalena Wysocki is a researcher at Technical University of Munich's Department of Computer Science within the School of Computation, Information and Technology. She works under Prof. Nassir Navab's Chair of Computer Aided Medical Procedures at the Garching campus, focusing on medical ultrasound imaging and neural radiance fields for surgical applications. Her research integrates deep learning with medical physics to solve reconstruction challenges in ultrasound imaging.

Her primary research interests include:

• Medical Ultrasound Imaging with neural field applications
• Physics-consistent 3D reconstruction from sparse views
• Geometry-aware deep learning for medical imaging
• Real-time surgical data science applications
• Confidence mapping in neural implicit representations

Her recent publication trends show concentrated work on neural radiance fields for ultrasound (Ultra-NeRF), with expanding applications to tissue property estimation (UltRAP-Net) and surgical activity understanding (EgoExOR). Key methodological threads include physics-based constraints in deep learning and confidence quantification for clinical deployment.

She actively supervises multiple Master's students on projects including:

• 4D Ultrasound SLAM (Sebastian Herz)
• Cardiac Shape Reconstruction with Motion Correction (Abdulkader Ghandoura)
• Confidence Maps research leading to MIUA 2024 publication (Bugra Yesilkaynak)

Her GitHub repositories (ultra-nerf, ultron) demonstrate strong software engineering contributions to the medical imaging community. She contributes to the NARVIS and RobUSt research groups focusing on surgical vision systems and robotics.