COmputing BRAin signals (COBRA): Biophysical computations of electrical and magnetic brain signals

Most of what we know about the dynamics of the brain has been learned from measurements of electrical and magnetic brain signals recorded inside and outside the brain. Despite their long history and widespread use, the proper interpretation of these brain signals in terms of the biophysical activity in underlying neurons (nerve cells) and neuronal networks is still lacking. Present-day analysis of these signals is predominantly statistical, and results do not have a clear biophysical interpretation. New biophysics-based analysis methods are thus needed to take full advantage of these brain-imaging techniques. The primary goal of the transdisciplinary project COBRA was to address this challenge by developing physics-based computational schemes, based on biologically detailed neuron models and validated against in-house experiments, for calculating the contributions cortical neurons to these electric and magnetic brain signals. This COBRA scheme was then used to, and made available for, comparison of model results with experimental data in mice and men in collaboration with various collaborators, including prominent international projects such as EUs Human Brain Project and Project MindScope at the Allen Brain Institute. The software developed in the project was made into a Python software package (LFPy) for use in large-scale brain-network simulations and analysis tools. In the first years of the project a the new simulation package (LFPy2.0) was developed and presented at a series of scientific conferences and research schools. A paper on the tool was published in the journal Frontiers in Neuroinformatics in 2018. Further, experimental studies in the mouse visual cortex were done ongoing, and an article on using "deep neural networks" to interpret electrical brain signals was published in 2020. Other machine learning methods for doing data analysis was also studied and new articles are on their way. The project was done at CINPLA (cinpla.org), a multidisciplinary neuroscience research centre located in the Faculty of Mathematics and Natural Science at the University of Oslo

Prosjektet har gitt oss ny kunnskap og en ny plattform for samarbeid med ledende internasjonale forskningsmiljøer som Allen Brain Institute for Brain Science i Seattle. Det har dessuten ført til rekruttering av en eksellent tysk forsker til Norge.

Most of what we know about the dynamics of the brain has been learned from measurements of electrical brain signals such as local field potentials (LFP), i.e., electrical potentials recordings inside the brain,, electroencephalography (EEG), i.e., recordings of electrical potentials at the scalp, electrocorticography (ECoG), i.e., potentials recorded on the cortical surface, and magnetoencephalography (MEG), i.e., recordings of magnetic fields outside the head. Despite their long history and widespread use, the proper interpretation of these brain signals in terms of the biophysical activity in underlying neurons (nerve cells) and neuronal networks is still lacking. Present-day analysis is predominantly statistical and limited to identification of phenomenological signal generators without a clear biophysical interpretation. New biophysics-based analysis methods are thus needed to take full advantage of these brain-imaging techniques. The primary goal of the transdisciplinary project COBRA is to address this challenge by developing physics-based computational schemes, based on biologically detailed neuron models and validated against in-house experiments, for calculating the contributions from populations of cortical neurons to electric (LFP, EEG, ECoG) and magnetic (MEG) brain signals, i.e., do 'virtual brain measurements'. This COBRA scheme will then in collaboration with various collaborators, including prominent international projects such as EUs Human Brain Project and Project MindScope at the Allen Brain Institute, be (i) used to explore how the various brain signals depend on the properties and state of the cortical networks of the cortical neurons, (ii) compared with various types of experimental data from mice and humans, (iii) developed into a Python software package (LFPy) for use in large-scale brain-network simuations and analysis tools.