Understanding how the brain accomplishes its numerous functions, and using this understanding to develop artificial intelligence algorithms, is the final frontier in science. Full understanding of brain/mental processes requires multi-level phenomics. We would like to create maps between brain states to mental states [1]. Such modeling attempts have been made in the past at high levels of abstraction [2]. Attractor network models are a step closer to neurobiological reality. Using neural models that incorporate some neurobiological mechanisms gives an insight into putative brain processes. Trajectories of brain activations may be linked to mental events. Transitions between brain states depend on the biophysical properties of neurons. Graphs that show such transitions illustrate associative thinking processes, creation of conceptual networks, and help to understand mental disorders. Dynamics of such networks allows to formulate novel  hypotheses, taking into account processes contributing to activation of neurons on many levels. As an example, I will present a simulation showing how autism or ADHD can result from properties of individual neurons changed by genetic mutations [3], and a sketch of learning mechanisms leading to the formation of conspiracy theories [4].

Progress in understanding mechanisms of spatial navigation in hippocampus formation (Nobel 2014) shows the way to create abstract cognitive maps. Analysis of real brain signals to discover fingerprints of brain cognitive activity is still a challenge, but has a great potential to create better brain-computer interfaces, neurofeedback and therapeutic procedures. Using functional magnetic resonance imaging (fMRI) and electroencephalographic (EEG) data, it is possible to assess changes in the activation of large-scale brain networks and to develop biomarkers useful for the diagnosis of mental diseases. These networks change as a result of cognitive load and working memory training, modifying the network of potentially accessible brain states [5].  The challenge is to recognize brain fingerprints using EEG and develop practical methods to extract relevant information from the EEG signals in real time, and use it to parameterize neuromodulation methods (DCS, TMS) that act directly on the structure of brain connections, repairing or optimizing their performance. We develop our own methods of signal source localization and reconstruction by solving the inverse problem in EEG [6], new spectral methods for analysis of EEG / MEG signals [7], and try to understand oscillatory dynamics of the brain using recurrence analysis [8].

• [1] Duch W (1996) Computational physics of the mind. Computer Physics Communication 97: 136-153
• [2] Duch W and Diercksen GHF (1995) Feature Space Mapping as a universal adaptive system. Computer Physics Communications 87: 341-371
• [3] Duch W. (2019), Autism Spectrum Disorder and deep attractors in neurodynamics.  Ch. 13, Springer Handbook of Multi-Scale Models of Brain Disorders, pp.135-148.
• [4] Duch W. (2021). Memetics and Neural Models of Conspiracy Theories. Patterns. Cell Press
• [5] Finc, K, Bonna, K, He, X, Lydon-Staley, D.M, Kühn, S, Duch, W, & Bassett, D. S. (2020). Dynamic reconfiguration of functional brain networks during working memory training. Nature Communications 11, 2435 (IF 11.8)
• [6] Rykaczewski, K, Nikadon, J, Duch, W, Piotrowski, T. (2021). SupFunSim: spatial filtering toolbox for EEG. Neuroinformatics 19, 107–125
• [7] M.K. Komorowski, K. Rykaczewski, T. Piotrowski, K. Jurewicz, J. Wojciechowski, A. Keitel, J. Dreszer, W. Duch (2021). ToFFi – Toolbox for Frequency-based Fingerprinting of Brain Signals. Neurocomputing (rev. 5/2022).
• [8] Duch W, Furman Ł, Tołpa K, Minati L. Short-Time Fourier Transform and Embedding Method for Recurrence Quantification Analysis of EEG Time Series. The European Physical Journal Special Topics (sub. 4/2022)