Andreas Maier et al.
Friedrich-Alexander Universität Erlangen-Nürnberg
Lehrstuhl für Mustererkennung
In this tutorial, we perform a review of the state-of-the-art of hybrid machine learning in medical imaging. We start with a short summary of the general developments of the past in machine learning and how general and specialized approaches have been in competition in the past decades. A particular focus will be the theoretical and experimental evidence pro and contra hybrid modelling. Next, we inspect several new developments regarding hybrid machine learning with a particular focus on so-called known operator learning and how hybrid approaches gain more and more momentum across essentially all applications in medical imaging and medical image analysis. As we will point out by numerous examples, hybrid models are taking over in image reconstruction and analysis. Even domains such as physical simulation and scanner and acquisition design are being addressed using machine learning grey box modelling approaches. Towards the end of the article, we will investigate a few future directions and point out relevant areas in which hybrid modelling, meta learning, and other domains will likely be able to drive the state-of-the-art ahead.
Klaus Maier-Hein et al.
Deutsches Krebsforschungszentrum
Medical Image Computing
The remarkable rise of deep learning has led to an overwhelming number of new papers coming up by the week. This tutorial intends to filter out the research most relevant for the medical image computing (MIC) community and present it in a structured and understandable form. It will cover recent developments related to common tasks in the community (e.g. segmentation, detection), but will also discuss methods that are currently gaining traction and that are likely to become even more relevant in the future, such as multi-task learning, active learning, self-supervised learning, causal learning and transformers. Basic knowledge of neural networks and deep learning is recommended.
Mattias Heinrich et al.
Universität zu Lübeck
Institut für Medizinische Informatik
Medical image registration has been a cornerstone in the research fields of medical image computing and computer assisted intervention, responsible for many clinical applications. Conventional methods that rely on hand-crafted metrics and optimisation have matured over the last decades leading to promising examples of practical adoption, despite their high computational complexity and limited ability to incorporate expert supervision.
The recent Learn2Reg MICCAI challenge (https://arxiv.org/pdf/2112.04489) has demonstrated two main directions of research that can help improve medical image registration with the use of deep learning: 1) extraction of semantic features using segmentation networks and GPU-acceleration of conventional optimisation algorithms for pairwise registration, 2) learning the iterative optimisation steps with feed-forward networks that directly predict displacement fields based on training from large population data.
In this BVM tutorial, we start by giving a brief overview of the current research in learning-based image registration and useful datasets and benchmarks - with focus on open-source and public accessible resources. Next, we will introduce two hands-on implementation tasks that can be performed on-site and give insight to both aforementioned research directions: GPU-accelerated iterative optimisation and learning to directly predict displacement fields. Gaining new experience and expertise in this inherently multidisciplinary topic can be beneficial to many in our community, especially for the next generation of young scientists, engineers and clinicians who often have only been exposed to a subset of these methodologies and applications.
We will guide participants to understand and implement published algorithms using provided medical challenge data. We aim to provide an opportunity for the participants to bridge the gap between expertises in medical image registration and deep learning, as well as to start a forum to discuss know-hows, challenges and future opportunities in this area. The tutorial can be run in the cloud (e.g. Google colab), please bring your own laptop for active participation.