Arbeitsgruppe Matthias Kaschube - Forschung
Dynamik neuronaler Repräsentationen
Das Gehirn bildet bemerkenswert effiziente Repräsentationen der sensorischen Umgebung und innerer Zustände, von Erinnerungen und Zukunftserwartungen. Wie erreicht das Gehirn dies? Um diese grundlegende Frage zu beantworten, kombinieren wir dynamische Modelle der Funktion neuronaler Netzwerke mit Techniken zur Modellierung neuronaler Daten in enger Zusammenarbeit mit experimentellen Gruppen. Unsere Arbeit bildet eine Schnittstelle zwischen Informatik, Physik, Biologie und KI.
Entstehung von neuronalen Repräsentationen in der Entwicklung des Gehirns
Unsere Forschung zeigt hochstrukturierte kortikale Netzwerke vor Beginn der sensorischen Erfahrung (Smith et al., Nature Neuro 2018; Mulholland et al., Elife 2021). Ihre Repräsentationsarchitektur ist in mehreren sensorischen und höheren Assoziationskortizes sehr ähnlich (Powell et al., Cerebral Cortex 2025) und scheint durch dynamische rekurrente Interaktionen vom Typ lokale Erregung und laterale Hemmung (LELI) geformt zu werden (Mulholland et al., Nature Commun 2024). Durch erfahrungsbedingte Netzwerkreorganisation werden diese endogen strukturiereten Netzwerke dann in zuverlässige kortikale Repräsentationen umgewandelt (Trägenap et al., Nature Neuroscience 2025; Lempel et al., Neuron 2025).
Kollaborationspartner: Gordon Smith (UMN), Ben Scholl (UC Denver), David Fitzpatrick (MPFI)
Flexible Repräsentationen für Lernen, Vergessen und Kreativität
Verschiedene Töne aktivieren überlappende Netzwerke im auditorischen Kortex, was möglicherweise deren Assoziation widerspiegelt. Wir beobachten fortlaufende dynamische Veränderungen dieses Stimulus-Co-Mappings, die während des Lernens flexibel beeinflusst werden (Aschauer et al., Cell Reports 2022), während die grobe auditorische Karte weitgehend erhalten bleibt (Chambers et al., Cerebral Cortex 2022), und argumentieren, dass diese Veränderungen der spontanen Entstehung neuer Assoziationen zugrunde liegen könnten.
Kollaborationspartner: Simon Rumpel (University Mainz), Noam Ziv (Technion)
Kognitive Karten, kognitive Kontrolle und neuronale Repräsentationsräume
Wir sagen allgemeine kognitive Fähigkeiten aus wenigen spontan enstandenden Netzwerkzuständen im menschlichen Gehirn voraus (Wehrheim et al., Neuroimage 2023), untersuchen die flexible Nutzung visueller Repräsentationen bei frei agierenden Tintenfischen (Reiter et al., 2018 Nature), und betrachten die hochdimensionale Organisation von Repräsentationsräumen in tiefen neuronalen Netzwerken (Galella et al., NeurIPS 2025).
Kollaborationspartner: Christian Fiebach (GU), Gilles Laurent (MPI BR)
Analysemethoden für strukturelle und funktionelle neuronale Daten
Wir entwickeln Methoden zur Analyse von chronischen Bildgebungsdaten von dendritischen Spines (Osuna-Vargas et al., ICCV 2025). Außerdem leisteten wir Pionierarbeit bei Methoden zur Verfolgung einer großen Anzahl von Chromatophoren in frei agierenden Tintenfischen (Reiter et al., Nature 2018). Kürzlich haben wir Methoden für die automatische Charakterisierung latenter Räume in tiefen neuronalen Netzen entwickelt (Wehrheim et al., ECCV 2024).
Ausgewählte Vorträge
- Kaschube M (2023), The Emergence of Cortical Representations, van Vreeswijk Theoretical Neuroscience Seminar (online) [video]
- Mulholland H, Kaschube M, Smith GB (2022), Mechanisms underlying the self-organization of patterned activity in the developing visual cortex, Cosyne Conference, Montreal, Canada [video]
- Trägenap S, Whitney DE, Fitzpatrick D, Kaschube M (2022), Experience drives the development of novel, reliable cortical sensory representations from endogenous networks, Bernstein Conference 2022, Berlin [link] [video]
Ausgewählte Publikationen
- Eppler, J. B., Kaschube, M., & Rumpel, S. (2025). Statistical learning and representational drift: A dynamic substrate for memories. Current Opinion in Neurobiology, 94, 103107. https://doi.org/10.1016/j.conb.2025.10310
- Lempel, A. A., Trägenap, S., Tepohl, C., Kaschube, M., & Fitzpatrick, D. (2025). Development of coherent cortical responses reflects increased discriminability of feedforward inputs and their alignment with recurrent circuits. Neuron. 10.1016/j.neuron.2025.08.014 https://doi.org/10.1016/j.neuron.2025.08.014 https://www.cell.com/neuron/fulltext/S0896-6273(25)00599-
- Seiler, J. P. H., Eppler, J. B., Dan, O., Elpelt, J., Kaschube, M., & Rumpel, S. (2025). Towards circuit mechanisms of the creative process: Describing the functions, mechanisms and neural correlates of creativity. (submitted) https://osf.io/preprints/psyarxiv/mpbgy_v1
- Galella, S., Wehrheim, M., & Kaschube, M. Dimensionality Mismatch Between Brains and Artificial Neural Networks. In The Thirty-ninth Annual Conference on Neural Information Processing Systems. https://openreview.net/forum?id=fyp34w19N2&referrer=%5BAuthor%20Console%5D(%2Fgroup%3Fid%3DNeurIPS.cc%2F2025%2FConference%2FAuthors%23your-submissions)
- Noda, T., Kienle, E., Eppler, J. B., Aschauer, D. F., Kaschube, M., Loewenstein, Y., & Rumpel, S. (2025). Homeostasis of a representational map in the neocortex. Nature Neuroscience, 1-13. https://doi.org/10.1038/s41593-025-01982-7
- Osuna-Vargas, Pamela, et al. Denoising diffusion models for high-resolution microscopy image restoration. 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV). Link: https://ieeexplore.ieee.org/abstract/document/10943369
- Osuna-Vargas, Pamela, et al.
SynapFlow: A Modular Framework Towards Large-Scale Analysis of Dendritic Spines.
Proceedings of the IEEE/CVF International Conference on Computer Vision. 2025.
https://openaccess.thecvf.com/content/ICCV2025W/BIC/html/Osuna-Vargas_SynapFlow_A_Modular_Framework_Towards_Large-Scale_Analysis_of_Dendritic_Spines_ICCVW_2025_paper.html
-
Galella, S.,Osuna-Vargas, P., Wehrheim, M., Kaschube, M.
MAPS: A Dataset for Controlled Probing of Representational Topology in Vision Models. (Extended Abstract that will be presented as poster only) NeurIPS 2025 Workshop on Symmetry and Geometry in Neural Representations.
https://openreview.net/forum?id=LW3IYaduNB
- Seiler J.P.H., Elpelt J., Mahkov V., Ghobadi A., Kapoor A.,Turner D., Kaschube M., Tuescher O., Rumpel S.
A reduced perception of sensory information is linked with elevated boredom in people with and without attention-deficit hyperactivity disorder. Commun Psychol 3, 47 (2025).
https://www.nature.com/articles/s44271-025-00214-9
- Seiler J. P. H., Elpelt J., Ghobadi A., Kaschube M. and Rumpel S.
Perceptual and semantic maps in individual humans share structural features that predict creative abilities.
Commun Psychol 3, 30 (2025).
https://www.nature.com/articles/s44271-025-00214-9
- Trägenap S., Whitney D.E., Fitzpatrick D. and Kaschube M. cortical representations Nature Neuroscience. 28(2), 394–405. [https://doi.org/10.1038/s41593-024-01857-3, Share-Version]
- Powell, N.J., Hein, B., Kong, D., Elpelt, J., Mulholland, H.N., Holland, R. A., Kaschube, M. and Smith, G.B. 2025. Developmental maturation of millimeter-scale functional networks across brain areas. Cerebral Cortex [Link]
- Wehrheim, Maren H., Pamela Osuna-Vargas, and Matthias Kaschube. Linking in Style: Understanding learned features in deep learning models. European Conference on Computer Vision. Cham: Springer Nature Switzerland, 2024. Link: https://link.springer.com/chapter/10.1007/978-3-031-73013-9_10
- Mulholland, H.N., Kaschube, M. and Smith, G.B., 2024.
Self-organization of modular activity in immature cortical networks.
Nature Communications, 15(1), p.4145.
https://www.nature.com/articles/s41467-024-48341-x
- Powell, N.J., Hein, B., Kong, D., Elpelt, J., Mulholland, H.N., Kaschube, M. and Smith, G.B., 2024.
Common modular architecture across diverse cortical areas in early development. Proceedings of the National Academy of Sciences, 121(11), p.e2313743121.
https://www.pnas.org/doi/10.1073/pnas.2313743121
- Vogel, F.W., Alipek, S., Eppler, J.B., Osuna-Vargas, P., Triesch, J., Bissen, D., Acker-Palmer, A., Rumpel, S. and Kaschube, M., 2023.
Utilizing 2D-region-based CNNs for automatic dendritic spine detection in 3D live cell imaging.
Scientific Reports, 13(1), p.20497.
https://www.nature.com/articles/s41598-023-47070-3
- Wehrheim, M. H., Faskowitz, J., Sporns, O., Fiebach, C. J., Kaschube*, M., & Hilger*, K., 2023. Few temporally distributed brain connectivity states predict human cognitive abilities. NeuroImage, Vol. 277, p. 120246. * jointly directed work
- Trägenap S., Whitney D.E., Fitzpatrick D., Kaschube M., 2022. Experience drives the development of novel, reliable cortical sensory representations from endogenously structured networks. bioRxiv. 2022:2022-11 [Link]
- Chambers A.R., Aschauer D.F., Eppler JB., Kaschube M., Rumpel S., 2022. A stable sensory map emerges from a dynamic equilibrium of neurons with unstable tuning properties. Cerebral Cortex [Link]
- Aschauer DF#, Eppler JB#, Ewig L, Chambers A, Pokorny C, Kaschube M*, Rumpel S* (2022) Learning-induced biases in the ongoing dynamics of sensory representations predict stimulus generalization. Cell Reports 38(6):110340
- Mulholland HN, Hein B, Kaschube M, Smith GB (2021) Tightly coupled inhibitory and excitatory functional networks in the developing primary visual cortex. Elife. 10:e72456. [Link]
- Distributed network interactions and their emergence in developing neocortex. Nat Neurosci.; 21(11) 1600-1608 doi: 10.1038/s41593-018-0247-5. *jointly directed work. [link] Press release Link (MPFI): Movies of spontaneous activity and distributed networks in the visual cortex.
- Reiter S, Hülsdunk P, Woo T, Lauterbach MA, Eberle JS, Akay LA, Longo A, Meier-Credo J, Kretschmer F, Langer JD, Kaschube M, Laurent G. (2018) Elucidating the control and development of skin patterning in cuttlefish. Nature; 562, 361-366. doi:10.1038/s41586-018-0591-3 [link] Press release
- Kaschube M, Nelson III CA, Benasich AA, Buzsáki G, Gressens P, Hensch TK, Hübener M, Kobor MS, Singer W, Sur M. (2018) Early Childhood. Emergent Brain Dynamics: Prebirth to Adolescence, edited by A. A. Benasich and U. Ribary. Strüngmann Forum Reports, vol. 25, J. Lupp, series editor. Cambridge, MA: MIT Press
- Smith GB, Sederberg A, Elyada YM, Van Hooser SD, Kaschube M*, Fitzpatrick D. (2015). The development of cortical circuits for motion discrimination. Nat Neurosci.; 18(2), 252-261. doi: 10.1038/nn.3921. *jointly directed work. [link]
- Sederberg A, Kaschube M. (2015) Inhibition facilitates direction selectivity in a noisy cortical environment. J Comput Neurosci. doi: 10.1007/s10827-014-0538-0. [link]
- Polyakov O, He B, Swan M, Shaevitz JW, Kaschube M, Wieschaus E. (2014) Passive mechanical forces control cell-shape change during Drosophila ventral furrow formation. Biophys J.;107(4):998-1010. doi: 10.1016/j.bpj.2014.07.013. [link]
- Khan Z, Wang YC, Wieschaus EF, Kaschube M. (2014) Quantitative 4D analyses of epithelial folding during Drosophila gastrulation. Development. 141(14):2895-900. doi: 10.1242/dev.107730. [link]
- Kaschube M (2014) Neural maps versus salt-and-pepper organization in visual cortex. Curr Opin Neurobiol 24:95–102. [link]
- Reichl L, Heide D, Löwel S, Crowley JC, Kaschube M, Wolf F (2012) Coordinated Optimization of Visual Cortical Maps (II) Numerical Studies. PLoS Comput. Biol. 8(11):e1002756. [link]
- Reichl L, Heide D, Löwel S, Crowley JC, Kaschube M, Wolf F (2012) Coordinated Optimization of Visual Cortical Maps (I) Symmetry-based Analysis. PLoS Comput. Biol. 8(11):e1002466. [link]
- Gelbart MA, He B, Martin AC, Thiberge SY, Wieschaus EF, Kaschube M. (2012) Volume conservation principle involved in cell lengthening and nucleus movement during tissue morphogenesis. PNAS 109:19298–303. [link]
- Keil W, Kaschube M, Schnabel M, Kisvarday ZF, Löwel S, Coppola DM, White LE, Wolf F (2012) Response to Comment on “Universality in the Evolution of Orientation Columns in the Visual Cortex“. Science 336:413. [link]
- Nelson KS, Khan Z, Molnár I, Mihály J, Kaschube M, Beitel GJ (2012) Drosophila Src regulates anisotropic apical surface growth to control epithelial tube size. Nat Cell Biol 14(5):518–25. [link]
- Wang YC, Khan Z, Kaschube M, Wieschaus EF (2012) Differential positioning of adherens junctions is associated with initiation of epithelial folding. Nature, 484:390–393. [link]
- Macke JH, Gerwinn S, White LE, Kaschube M, Bethge M (2011) Gaussian process methods for estimating cortical maps.
- NeuroImage 56(2):570–81. [link]
- Kaschube M, Schnabel M, Löwel S, Coppola DM, White LE, Wolf F (2010) Universality in the evolution of orientation columns in the visual cortex. Science 330:1113–1116. [link]
- Keil W, Schmidt K-F, Löwel S, Kaschube M (2010) Reorganization of columnar architecture in the growing visual cortex. PNAS 107:12293–12298. [link]
- Martin AC, Gelbart M, Fernandez-Gonzalez R, Kaschube M, Wieschaus EF (2010) Integration of contractile forces during tissue invagination. JCB 188:735–749. [link]
- Doubrovinski K, Polyakov O, Kaschube M (2010) A mesoscopic description of contractile cytoskeletal meshworks. EPJE 33:105–110. [link]
- Martin AC, Kaschube M, Wieschaus EF (2009) Pulsed contractions of an actin–myosin network drive apical constriction. Nature 457:495–499. [link]
- Kaschube M, Schnabel M, Wolf F, Löwel S (2009) Interareal coordination of columnar architectures during visual cortical development. PNAS 106:17205–17210. [link]
- Macke JH, Gerwin S, White LE, Kaschube M, Bethge M (2009) Bayesian estimation of orientation preference maps. NIPS 22. [link]
- Kaschube M, Schnabel M, Wolf F (2008) Self-organization and the selection of pinwheel density in visual cortical development. NJP 10:015009. [link]
- Schnabel M, Kaschube M, Wolf F (2008) Pinwheel stability, pattern selection and the geometry of visual space. [link]
- Schnabel M, Kaschube M, Loewel S and Wolf F (2007) Random Waves in the Brain: Symmetries and Defect Generation in the Visual Cortex. EPJ ST 145:137–157. [link]
- Kaschube M, Wolf F, Puhlmann M, Rathjen S, Schmidt KF, Geisel T and Loewel S (2003) The pattern of ocular dominance columns in cat primary visual cortex: Intra- and interindividual variability of column spacing and its dependence on genetic background. EJN 18:3251–3266. [link]
- Kaschube M, Wolf F, Geisel T, Löwel S (2002) Genetic influence on quantitative features of neocortical architecture. J Neurosc. 22:7206–7217. [link]
- Kaschube M, Wolf F, Geisel T, Löwel S (2001) T he prevalence of colinear contours in the real world. Neurocomputing 38–40:1335–1339. [link]
- Kaschube M, Wolf F, Geisel T, Löwel S (2000) Quantifying the variability of patterns of orientation domains in the visual cortex of cats. Neurocomputing 32–33:415–423. [link]
Software
- Tools for analysing Calcium imaging data [Smith, Hein, Whitney et al., Nat Neurosci, 2018][link].