Prof. Dr. Katharina Breininger

Department Artificial Intelligence in Biomedical Engineering (AIBE)

In my group, we are working at different topics concerning machine learning, pattern recognition, and artificial intelligence applied to medical imaging. Specifically, we are interested in how we can work in settings with limited data and/or sparse labels for a number of clinical challenges, both supporting clinical workflows and allowing for new scientific insights.

Research projects

  • Human in the loop labeling
  • Image Analysis for a Digital Twin for Rheumatic Diseases

Current Projects

  • Exploring Brain Mechanics (EBM): Understanding, engineering and exploiting mechanical properties and signals in central nervous system development, physiology and pathology

    (Third Party Funds Group – Overall project)

    Term: 1. January 2023 - 31. December 2026
    Funding source: DFG / Sonderforschungsbereich / Transregio (SFB / TRR)

    Thecentral nervous system (CNS) is our most complex organ system. Despite tremendousprogress in our understanding of the biochemical, electrical, and geneticregulation of CNS functioning and malfunctioning, many fundamental processesand diseases are still not fully understood. For example, axon growth patterns inthe developing brain can currently not be well-predicted based solely on thechemical landscape that neurons encounter, several CNS-related diseases cannotbe precisely diagnosed in living patients, and neuronal regeneration can stillnot be promoted after spinal cord injuries.

    Duringmany developmental and pathological processes, neurons and glial cells aremotile. Fundamentally, motion is drivenby forces. Hence, CNS cells mechanicallyinteract with their surrounding tissue. They adhere to neighbouring cells and extracellular matrix using celladhesion molecules, which provide friction, and generate forces usingcytoskeletal proteins.  These forces aretransmitted to the outside world not only to locomote but also to probe themechanical properties of the environment, which has a long overseen huge impacton cell function.

    Onlyrecently, groups of several project leaders in this consortium, and a few other groupsworldwide, have discovered an important contribution of mechanical signalsto regulating CNS cell function. For example, they showed that brain tissuemechanics instructs axon growth and pathfinding in vivo, that mechanicalforces play an important role for cortical folding in the developing humanbrain, that the lack of remyelination in the aged brain is due to an increasein brain stiffness in vivo, and that many neurodegenerative diseases areaccompanied by changes in brain and spinal cord mechanics. These first insights strongly suggest thatmechanics contributes to many other aspects of CNS functioning, and it islikely that chemical and mechanical signals intensely interact at the cellularand tissue levels to regulate many diverse cellular processes.

    The CRC 1540 EBM synergises the expertise of engineers, physicists,biologists, medical researchers, and clinicians in Erlangen to explore mechanicsas an important yet missing puzzle stone in our understanding of CNSdevelopment, homeostasis, and pathology. Our strongly multidisciplinary teamwith unique expertise in CNS mechanics integrates advanced invivo, in vitro, and in silico techniques across time(development, ageing, injury/disease) and length (cell, tissue, organ) scalesto uncover how mechanical forces and mechanical cell and tissue properties,such as stiffness and viscosity, affect CNS function. We especially focus on(A) cerebral, (B) spinal, and (C) cellular mechanics. Invivo and in vitro studies provide a basic understanding ofmechanics-regulated biological and biomedical processes in different regions ofthe CNS. In addition, they help identify key mechano-chemical factors forinclusion in in silico models and provide data for model calibration andvalidation. In silico models, in turn, allow us to test hypotheses without the need of excessive or even inaccessibleexperiments. In addition, they enable the transfer and comparison of mechanics data and findingsacross species and scales. They also empower us to optimise processparameters for the development of in vitro brain tissue-like matricesand in vivo manipulation of mechanical signals, and, eventually, pavethe way for personalised clinical predictions.

    Insummary, we exploit mechanics-based approaches to advance ourunderstanding of CNS function and to provide the foundation for futureimprovement of diagnosis and treatment of neurological disorders.

  • Maschinelles Lernen und Datenanalyse für heterogene, artübergreifende Daten (X02)

    (Third Party Funds Group – Sub project)

    Overall project: SFB 1540: Erforschung der Mechanik des Gehirns (EBM): Verständnis, Engineering und Nutzung mechanischer Eigenschaften und Signale in der Entwicklung, Physiologie und Pathologie des zentralen Nervensystems
    Term: 1. January 2023 - 31. December 2026
    Funding source: DFG / Sonderforschungsbereich (SFB)

    X02 nutzt die in EBM erzeugten Bilddaten und mechanischen Messungen, um Deep Learning-Methoden zu entwickeln, die Wissen über Spezies hinweg transferieren. In silico und in vitro Analysen werden deutlich spezifischere Daten liefern als in vivo Experimente, insbesondere für menschliches Gewebe. Um hier Erkenntnisse aus datenreichen Experimenten zu nutzen, werden wir Transfer Learning-Algorithmen für heterogene Daten entwickeln. So kann maschinelles Lernen auch unter stark datenlimitierten Bedingungen nutzbar gemacht werden. Ziel ist es, ein holistisches Verständnis von Bilddaten und mechanischen Messungen über Artgrenzen hinweg zu ermöglichen.

Recent publications







Related Research Fields