Machine Learning in Medical Image Analysis

Machine Learning in Medical Image Analysis

Bridging the gap between AI and clinical practice

Cluster of Excellence: Machine Learning - New Perspectives for Science

University of Tübingen

Our Research

In the field of medical image analysis, the ultimate goal is to improve patient outcomes. There are a number of broad ways in which machine learning can help achieve this goal:

  • Relieving the burden on doctors and the healthcare system by accelerating and simplifying the analysis of medical images through partial or full automation of steps such as diagnosis, outcome prediction, image quantification, and image reconstruction.
  • Developing technology which enables completely novel clinical workflows which are not possible without AI support.
  • Extraction of new clinical knowledge from large image data bases, which can inform future clinical decisions, treatments and drug trials.

Even though tremendous progress has been made for all of those points in research settings, surprisingly little of this technology has leaked into medical practice. One reason for this is that the medical domain is an extremely high-stakes application field with extraordinary demands on robustness of algorithms. Another reason is that algorithmic outputs are not suitable for clinical decision making if neither the patient nor the doctor can understand the reasoning behind the prediction, and clinicians are loath to use the thus-far predominately black-box technology. Both of the above points also have important implications for the certification of AI technology.

Therefore, in order to start harnessing the massive potential of machine learning for healthcare, and to actually use it to improve real patient outcomes, the Machine Learning in Medical Image Analysis group aims to do research that helps to bridge this gap between machine learning and clinical practice. We perform this research along four broad directions:

  • Robustness, Safety and Uncertainty
  • Interpretable Machine Learning
  • Human-in-the-Loop Machine Learning Systems
  • Generative Modelling on Big Medical Datasets

These topics are described in more detail in the research areas section below.

We are part of the Cluster of Excellence: Machine Learning - New Perspectives for Science and the University of Tübingen.

The group is headed by Dr. Christian Baumgartner.

Team

Group Leader

Avatar

Christian Baumgartner

Independent Research Group Leader

PhD Students

Avatar

Jaivardhan Kapoor

PhD student

Probabilistic inference in spatio-temporal models, (starting in summer 2021, co-supervision with Prof. Jakob Macke)

Avatar

Paul Fischer

PhD student

Uncertainty quantification in medical prediction systems (starting in autumn 2021)

Master Students

Avatar

Carina Schmidt

Master student

Active and reinforcement learning for image segmentation

Avatar

Simon Gutwein

Master student

Radiation therapy dose estimation using deep neural networks, (co-supervision with Prof. Daniela Thorwarth)

Administration

Avatar

Elena Sizana

Administrative Assistant

Research Areas

Robustness, Safety and Uncertainty
How can we build robust systems that know when they don’t know?
Robustness, Safety and Uncertainty
Interpretable Machine Learning
How can we build systems that can explain themselves?
Interpretable Machine Learning
Human-in-the-Loop Machine Learning Systems
How can we integrate humans in the training and deployment of ML?
Human-in-the-Loop Machine Learning Systems
Generative Modelling on Big Medical Datasets
How can we extract new clinical knowledge from medical images?
Generative Modelling on Big Medical Datasets

Contact

  • +49 7071 29-70874
  • Maria-von-Linden-Straße 6, AI Research Building, R. 40-5/A4, Tübingen, Baden-Württemberg 72076