Marcel Fernandez Rosas

Logo

Potsdam, Germany
| | |

I am a digital health enthusiast who enjoys connecting the dots of technology and healthcare.
My degrees in Economathematics and Digital Health have given me a strong foundation in analytical problem solving and skills to drive healthcare innovation.
Colleagues and friends appreciate me for my friendly, empathetic, creative, and determined nature. I am responsible, pragmatic, and organized, but flexible.

Projects

A partial list of projects I have worked on.


Sovereign Healthcare Data Sharing

Eclipse Data Space Components Connector for Healthcare Data Sharing

Medium

The project aims to show how the Eclipse Data Space Components (EDC) framework can facilitate the sharing of health data. The EDC framework provides a set of tools to implement data spaces. Data spaces are digital infrastructures that enable secure and sovereign data sharing between different stakeholders.


Medical Image Generation

3D Causal-StyleGAN3 for Synthesizing MR Images of Alzheimer’s Disease

Paper Medium

At a research seminar at the Hasso-Plattner-Institute, we developed a new way to generate images from underrepresented AD populations. These include younger patients and those with less advanced disease stages. The model uses a causal graph to represent the relationships between age, sex, clinical dementia rating, and brain volumes. It can intervene on these variables to generate images with specific features.


Medical Image Segmentation

A simulation framework for benchmarking active learning strategies for 3D medical image segmentation

GitHub Paper

During a research seminar at the Hasso-Plattner-Institute we developed an active learning benchmarking framework for scientists seeking comparability and reproducibility. The framework is open-source and our paper was accepted at the Adaptive Experimental Design and Active Learning in the Real World workshop at ICML 2022.


Fair Compensation

A showcase for calculating fair salary with machine learning

Backend Backend Demo

The aim of the project is to demonstrate the potential of digitalization in HR. We will show how machine learning can be used to make the salary structure fair. This may involve, for example, identifying and eliminating potential gender pay gaps.