Prof. Dr. Andreas M. Kist

Juniorprofessur für Artificial Intelligence in Communication Disorders

My group mines artificial intelligence methodologies for efficiency and robustness to allow eventual clinical applicability.
We focus on the analysis and integration of biomedical imaging and multisensory data to compute quantitative clinically relevant features.

Research Projects

  • Diagnostik und Therapieverlauf durch den Einsatz von KI-Methoden
  • KI-basierte multimodale Datenerhebung, insbesondere von Audio- und Bildmaterial
  • Optimierung und Generalisierung von neuronalen Netzen für die erfolgreiche Anwendung in der Klinik
  • Erstellung von Benchmark-Datensätzen für umfassende Reproduzierbarkeit
  • Neuronale Architekturoptimierung für die Anwendung auf Mikrocontrollern
  • Hardware- und Softwareseitige Implementierung von Hardwarebeschleunigern für die effiziente Datenverarbeitung
  • Architekturoptimierung für den Einsatz in der Medizin
  • Maximierung von Robustheit und Generalität neuronaler Netze
  • Balancierte neuronale Netze durch ganzheitliche Ressourcenoptimierung
  • Entwicklung von Software für effizientes überwachtes Lernen

  • Do stability of hippocampal neuronal excitatory synaptic connectivity and representations support spatial learning?

    (Third Party Funds Single)

    Project leader:
    Term: 1. August 2025 - 31. July 2028
    Funding source: DFG-Einzelförderung / Sachbeihilfe (EIN-SBH)

    Hippocampal pyramidal neurons encode spatial locations through localized firing patterns, or place cells. Studies in humans and animals have shown a critical role for the hippocampus in spatial and episodic memory. Several theories have sought to explain how the hippocampus supports these functions and have considered place cells as basic substrates of long-term spatial memory. In many such frameworks, a central tenet is that long-term spatial memory arises from ensembles of cells that retain their spatial coding properties over time periods relevant to long-term memory. However, hippocampal spatial representations exhibit a surprising level of instability, or drift, which seems counterintuitive for a brain region important for memory formation. A few hypotheses exist on the function of this phenomenon, but the cellular and circuit mechanisms underlying hippocampal representational drift representations are completely unknown. By using two-photon time-lapse imaging to track simultaneously excitatory structural plasticity and activity patterns of the same pyramidal neurons over several days in the hippocampus of live mice, we want to shed light on the link between the stability connectivity and the drift of hippocampal representations. Moreover, we will study the relationship between structural plasticity, representation turnover, and learning by using a behavioral learning task, concomitant to imaging, and by perturbing the activity patterns of hippocampal dorsal CA1 pyramidal neurons by increasing local inhibition. In parallel, - and to increase the efficiency and reliability of our data analysis – we will develop a method based on machine and deep learning to consistently and automatically track dendritic spines in two-photon image time lapses and to match these to the activity of single pyramidal neurons in hippocampal dorsal CA1.

2026

  • Neubig, L., Larsen, D., Ikuma, T., Kopp, M., Kunduk, M., & Kist, A. (2026). Markerless Tracking-Based Registration for Medical Image Motion Correction. In Lina Felsner, Thomas Küstner, Andreas Maier, Chen Qin, Seyed-Ahmad Ahmadi, Anees Kazi, Xiaoling Hu (Eds.), Lecture Notes in Computer Science (pp. 34-43). Daejeon, KOR: Springer Science and Business Media Deutschland GmbH.

2025

2024

2023

2022

2021

Related Research Fields

Contact: