The Bernstein Center Freiburg

Informal Seminar
Felix Hoffmann

Bernstein Center Freiburg

Structural and dynamical aspects of
neural networks with anisotropic tissue geometry

Non-random connectivity has been repeatedly reported in cortical networks, yet underlying connection principles of these patterns remain elusive. Proposing an abstract geometric network model reflecting stereotypical axonal and dendritic morphology of local cortical layer 5 networks, we investigated in how far anisotropy in connectivity can constitute such an underlying connectivity rule. Using a combination of analytical considerations and numerical analysis, we found that while standard network measures and pair connectivity remain unaffected, higher order connectivity is strongly influenced by anisotropy, in many cases reflecting patterns found in local cortical circuits. Presenting an abstract network model featuring connectivity principles beyond distance-dependency, the results shown here not only make a strong case for morphology-induced rules as underlying connection principles of non-random patterns, but may provide a step towards a network archetype improving upon the standard random model.

Wednesday, June 25, 2014

14:15 h
BCF Seminar Room "Casino"
Hansastr. 9a
The talk is open to the public. Guests are cordially invited!