Personalized brain models reveal the smarter the brain the slower the decision making
Combining magnetic resonance imaging (MRI) and mathematical modeling has allowed researchers to create personalized brain models in silico.
A new study sheds light on how brain structure affects intelligence and decision making via in silico modeling of individual brains. The research team, based at the Berlin Institute of Health (BIH) at Charité – Universitätsmedizin Berlin (Germany), developed a learning algorithm to understand how brain structure impacts decision-making capabilities. The team hopes that this technology can be used in the future to model surgical and drug interventions, among other therapeutics.
The team began by developing a general brain model using digital data derived from brain scans including MRI scans. They combined these scans with mathematical algorithms trained with theoretical knowledge of biological processes. They then utilized data from 650 participants in the Human Connectome Project, a US-based study that has been studying human brain structure for over a decade, to personalize these models. The Human Connectome Project provided information about participants’ IQ scores and performances on cognitive tests.
The multi-scale brain network models were personalized to couple a participant’s structural white-matter connectivity with a generic neural circuit for decision making and working memory. “We can reproduce the activity of individual brains very efficiently,” reported Petra Ritter, the senior author of the paper. “We found out in the process that these in silico brains behave differently from one another – and in the same way as their biological counterparts. Our virtual avatars match the intellectual performance and reaction times of their biological analogs.”
The participants and their counterpart brain models were faced with problem-solving tasks that varied in difficulty. Interestingly, they found that individuals with higher IQ scores were only faster at solving simple problems and actually took longer than lower-IQ individuals to solve more difficult tasks, however, completed tasks with fewer errors.
Sewing the seeds of diagnosis: a novel assay for abnormal α-synuclein detection
A novel assay that can detect synucleinopathies from patient sera offers a fast method for diagnosing neurodegenerative diseases.
The in silico brains reflected these real-life results, indicating their accuracy. The brain models also revealed that the brains with reduced synchrony ‘jumped to conclusions’ to arrive more quickly at solutions, which were often incorrect. Greater brain connectivity and increased synchrony resulted in slower processing to arrive at the correct answers, indicating that higher-IQ individuals require more time to process complex problems due to their synchronized brain processes.
“Synchronization, i.e., the formation of functional networks in the brain, alters the properties of working memory and thus the ability to ‘endure’ prolonged periods without a decision,” explained Michael Schirner, lead author of the study. “In more challenging tasks, you have to store previous progress in working memory while you explore other solution paths and then integrate these into each other. This gathering of evidence for a particular solution may sometimes take longer, but it also leads to better results. We were able to use the model to show how excitation-inhibition balance at the global level of the whole brain network affects decision-making and working memory at the more granular level of individual neural groups.”
Although the current study highlights how neural connectivity affects decision making, the researchers are particularly excited about the potential role that these reliable, personalized brain models might play in future therapeutic development. Due to their high biological relevance, these models could be used to help individuals with neurodegenerative diseases, such as Parkinson’s disease. “The simulation technology used in this study has made significant strides, and can be used to improve personalized in silico planning of surgical and drug interventions as well as therapeutic brain stimulation. For example, a physician can already use a computer simulation to assess which intervention or drug might work best for a particular patient and would have the fewest side effects,” concluded Ritter.
