How to build the best digital twins of the human brain

The potential to create personalized digital twins of your brain and body is a hot topic neurology and medicine today. These computer models are designed to simulate how parts of your brain interact and how the brain might respond to stimulation, disease, or medication.

The unusual complexity of Art brain billions of neurons make this a very difficult task, of course, even in the age of AI and big data. so far, whole brain models tried to capture what makes each brain unique.

People’s brains are structured slightly differently, so everyone has a unique network of neural connections that is a kind of “fingerprint of the brain”.

However, most so-called brain twins are now more like distant cousins. Their performance is barely closer to the real one than if the model used a random stranger connection scheme.

This is important because digital twins are increasingly proposed as tools for testing treatments through computer simulations before applying them to real people. If these models fail to capture the fundamentals of each patient’s unique brain organization, their predictions will not be personalized, and at worst, may be misleading.

U our latest researchpublished in Neuroscience of naturewe show that realistic brain digital twins require something that many existing models overlook: competition between different brain systems.

Our findings suggest that without competition, digital twins risk being overly shared, losing what makes you ‘you’.

Too much cooperation

The human brain is never static. The ebb and flow of its activity can be mapped non-invasively using neuroimaging techniques such as functional MRI. From this, a computer model can be built, specific to that person and simulating the interaction of his brain regions. This is the idea of ​​a digital double.

The brain is often described as a highly cooperative system. Yet everyday experiences, such as focusing attention or switching between tasks, intuitively tell us that the brain’s systems compete for limited resources. Our brain cannot do everything at once, and not all regions can be active together all the time.

Despite this, the vast majority of brain simulations over the past 20 years have not taken into account these competitive interactions between regions. Rather, they “forced” neighboring regions to cooperate. This can push the simulated brain into over-synchronized states that are rarely found in real brains.

In a a large comparative study humans, macaques, and mice, our international team of researchers used noninvasive recordings of brain activity to show that the most realistic whole-brain models require not only cooperative interactions within specialized brain circuits, but also competitive interactions between different circuits.

To achieve this, we compared two types of brain models: one in which all interactions between brain regions were cooperative, and another in which regions could excite or suppress each other’s activity. In humans, monkeys, and mice, models that included competitive interactions consistently outperformed only cooperative models.

Using a large-scale analysis of more than 14,000 neuroimaging studies, we found that spontaneous activity in competitive models more accurately reflects known cognitive patterns, such as those related to attention or memory. This suggests that competition is critical for the brain to flexibly activate appropriate combinations of regions—a hallmark of intelligent behavior.

A visual summary of our research:

When whole-brain models of humans, macaques, and mice are allowed to treat interactions between certain brain regions as competitive, they consistently do so—creating patterns of activity that closely resemble those associated with real cognitive processes. Luppi et al / Nature Neuroscience, CC BY

We conclude that competitive interactions act as a stabilizing force, allowing different brain systems to take turns shaping the direction of ebb and flow without interference or distraction. This ability to avoid propulsive activity may also contribute to the remarkable energy efficiency of the mammalian brain, which is many orders of magnitude more efficient than current AI systems.

Importantly, models with competitive interactions were not only more accurate, but also more individualized. This means they are better at capturing the brain’s unique fingerprint that distinguishes one person’s brain from another’s.

No more lost in translation?

The fact is that our conclusions maintained through humans and other mammals suggests that they reflect the fundamental principles of operation of intelligent systems. In each case, we found that models with competitive interactions produced patterns of brain activity that closely resembled those associated with real cognitive processes.

This could have major implications for translational neuroscience. Animal models are commonly used to test treatments before human trials, but differences between species often limit how well these results translate. There is a cure for about 90 percent of neuropsychiatric disorders “lost in translation”, have failed in human clinical trials after promising animal trials.

Combining patient brain imaging data with whole brain simulations could radically change this. A structure that works for different species to provide a powerful bridge between basic research and clinical application.

If someone needs brain intervention, for example because of epilepsy or a tumor, their digital double can be used to study how the patient’s brain activity will change when stimulated with different levels of drugs or electrical pulses. This can greatly improve existing trial and error methods with real patients and thus provide better treatment.

Common principles of brain organization across species also offer a way to understand how to shape the next generation artificial intelligence. In the not-too-distant future, we may be able to build digital twins that more accurately reproduce the characteristics of the human brain—and potentially models of artificial intelligence that are more faithful to the human mind.

This article is reprinted from Conversation under a Creative Commons license. Read it original article.