Home

Pawel Herman

Researcher, PhD

Department of Computational Biology

School of Computer Science and Communication

KTH Royal Institute of Technology, Stockholm, Sweden

email: <paherman at kth dot se>,   phone: (+46) 08 790 6513

I am a researcher in Prof. Anders Lansner’s lab at the Dept. of Computational Biology at KTH and SU. Here you can find a short version of my CV.

My research interests are deeply steeped in brain and information science. My work to date has been at the crossroads of these two fields, sometimes broadly referred to as neuroinformatics. To be more precise, my research to date has revolved around scientific questions in computational and cognitive neuroscience, brain data analysis, brain-inspired computing, brain-computer interfacing (BCI) and computer-aided brain diagnostics. In the quest for conceptual understanding of brain mechanisms and general principles of neural information processing, I have extensively relied upon the breadth of computer science methodology and have made some contributions to its development.

The last few years I have mainly devoted to computational neuroscience. In Prof. Anders Lansner’s lab we have developed competence in large-scale modelling of cortical networks with the focus on bridging the gap between the brain function and the underlying mesoscopic-level biological processes. My area of special interest is centred on oscillatory and synchronisation phenomena emerging in biophysically detailed attractor network models as correlates of the simulated cortical memory, attention or perception.

I have also been involved in simulations of abstract non-spiking network models to investigate large-scale processing and coding of information in the brain. For example, we have recently built a holistic model of the mammalian olfactory system to study mechanisms of odour recognition. Abstract network modelling approach has also been used in our work to study the potential of more generic brain-inspired computing paradigms with a view to extracting salient characteristics from any data sources. In different pattern recognition tasks, the proposed aproach compares favourably to the state-of-the-art machine learning methods. Our network models are deployed on supercomputers to exploit the intrinsic parallelism and distributed nature of biological networks capable of self-organisation to adaptively handle large data sets. This inevitably opens up new opportunities for robust brain-inspired analysis of high-dimensional neuroimaging data with complex spatio-temporal dynamics.

 

For a more detailed description of my research profile, please see here.

I am also involved in teaching activities in the School of Computer Science and Communication at KTH. If you are interested, please read more here.