Open Projects in Machine Learning in Medicine

In a field that is so flooded with data as medicine, learning algorithms are providing countless exiting promises. For example, the data that we already have may contain all the information needed to understand and cure numerous diseases — we only need to learn how to make those information available. Help to uncover this entirely new world with us by choosing among one of our open master’s theses listed here.

Project	Supervisor
Prototype Development of an AI-Enabled 3D-Printed Microscope	Viktor Kölzer
PEEK Shrink Compensation through Coupled FEA and AI-Based Process Optimization	Thieringer / Cattin
From Noise to Brain: Multiple Sclerosis Lesion Filling using Denoising Diffusion Probabilistic Models	Martin Styner
Graph convolutional models for White Matter Tract Profile Classification in Autism	Martin Styner
Autism Subtyping from Neuroimaging data via Tab-PRADA	Martin Styner
LLM-Based Structured Data Generation for Clinical Research Databases	Philippe Cattin
Weight-Space Learning for Clinical Multi-Task Risk Prediction	Philippe Cattin
Development of a predictive model for postoperative BMI and related diseases	Philippe Cattin
Computer vision to automate bacterial single-cell analysis	Lucas Boeck (DBM)
Developing and Evaluating Vision-Language Models for Pediatric Dermatology	Alexander Navarini
Shades of Equity: Using Generative AI to Reduce Skin Tone Bias in Dermatology Datasets	Alexander Navarini
Predicting the timing of foot events during walking	Morgan Sangeux
Transfer Learning for the collaborative segmentation platform Dafne	Francesco Santini
Segment By Example: a deep-learning approach to generic atlas-based segmentation	Francesco Santini
Deep Learning for Automatization of Video-Analysis for Malaria Vector Behavioural Studies	Philippe Cattin
Incorporation of Uncertainty into Annotations for Deep Learning-based Image Segmentation	Philippe Cattin
AI Generation of Cerebral Perfusion Maps	Philippe Cattin
Finite Element Validation of retropatellar contact pressure in AI generated trochleas	Philippe Cattin