Researchers from the Human Brain project have created the first large-scale model of how Parkinson’s brain responds to deep brain stimulation. Deep brain stimulation is a common method of treating Parkinson’s symptoms, but currently the results of invasive treatment are difficult to predict. Modeling and simulating electrical stimulation at multiple levels of the brain network can help clinicians “review” their effects and plan therapies accordingly.
Deep brain stimulation is a common method of treating Parkinson’s symptoms, but currently the results of invasive treatment are difficult to predict. Modeling and simulating electrical stimulation at multiple levels of the brain network can help clinicians “review” their effects and plan therapies accordingly.
The study marks the first published case of a large-scale collaborative simulation of the human brain applied for clinical use, and the methodology is now openly available at THE BRAIN platform, the digital research infrastructure developed by the Human Brain project.
“Among the goals of HBP is the introduction of brain models on a large scale, able to show high detail of neural activity, inserted in a broader context,” explains Petra Ritter from the Berlin Institute of Health in Charité, senior author of the study. “We described the methodology for this earlier and are now able to implement it in EBRAINS and have identified the first case of use: deep brain stimulation in Parkinson’s disease.”
Observing what happens during Parkinson’s disease on various scales is very useful for understanding the patient’s brain dynamics and applying therapeutic deep brain stimulation. However, multiscale models are very rare, and neurologists usually focus on either a small area with high detail or larger models that look at the average field activity of the entire brain. “In the case of Parkinson’s disease, you have small subcortical nuclei, where you can look at the activity of individual neurons not only in space but also in time, observing the exact time of the peaks. You can also model the effects of plasticity that occur during stimulation of these areas, ”says Ritter. “But you’ll miss valuable information about what’s happening across the brain, which we can instead capture in Virtual Brain, an open source platform for constructing and simulating customized network models across the brain.”
The team uses chip models of the basal ganglia and thalamic areas. This network has already been set up and optimized for small-scale work by a team led by Fred Hamker from the Department of Computer Science at the Chemnitz University of Technology in Germany, who co-authored the study. “We included the entire network in TVB, first making sure it had the same behavior as the original,” Ritter said. “We then constructed two multiscale models, one for a Parkinson’s patient, the other for health control, based on measured brain data for both models.” The researchers then compared the first differences in resting behavior in Parkinson’s healthy and brain models. They then began programming and simulating various targeted deep brain stimulations, often performed in vivo, in a clinical setting.
Interestingly, the researchers found that the rate of firing of subcortical regions, and in particular parts of the thalamus, has a reduced rate in the Parkinson’s patient model at rest, an observation that has also been reported in the true brains of patients with Parkinson’s. This decrease in firing rate can be normalized by in-silico deep brain stimulation. “However, this also has an effect on the cortical scale, leading to differentiated motor cortex activity depending on the stimulus. “Some effects are potentially less beneficial to the patient.” “All this brain effect would not have been noticeable just by looking at the microscale, but it became apparent when looking at multiple scales at once.
With the new approach, it is possible both to observe some structures with high detail (not only spatially, but also in time), and to experience the global effect of the simulation, Ritter concludes. “This has the potential to translate into the clinical environment, improving prediction and personalization when performing deep brain stimulation. The next step is to introduce rules for plasticity in the system and see what other phenomena we are able to capture on a multiscale.
Help: Meier, JM; Perdikis, D .; Blikensdörfer, A .; Stefanovski, L .; Liu, Q .; Matt, O .; Dinkelbach, H. Yu .; Baladron, J .; Hamker, FH; Ritter, P. Virtual deep brain stimulation: a large-scale joint simulation of a basal ganglia model with peak velocity and a midbrain midfield model with the virtual brain. Experimental neurology 2022; 354. https://doi.org/10.1016/j.expneurol.2022.114111.
This article has been republished from the following materials. Note: The material can be edited for length and content. For more information, please contact the cited source.