The connectome, the physical map of a personality and the rise of high-res cryonics
What is the essence of a person? It's a deep question for a Tuesday. While almost nothing is known about consciousness itself, science has opened some tiny doors into how the brain functions. And whether you're a biological essentialist who believes that the entirety of consciousness is embedded in the meatware of the brain, or some kind of spiritualist that sees consciousness as something separate and not of this plane that drives, observes and works with the brain, it's clear that the brain is responsible for a ton of the stuff that makes you who you are.
Memories, for example. Muscle memories, too. Thinking patterns that become more and more well-worn over time until they're a default. Connections between things. Functional representations of things, concepts, ideas, people. Whether there's a soul, or spirit, or separate consciousness involved or not, these memories and connections do seem to exist at a physical level in the brain.
And whatever exists at a physical level, transhumanists would say, can be preserved after death and re-animated at some point, or in some form. Hence the field of cryonics, in which full bodies, or just brains, are frozen after death to preserve as much of this precious thinking tissue as possible, in the hope that individuals can one day be brought back to life.
But what physical structures, exactly, are we trying to preserve, and do current cryonics techniques actually preserve it?
Meet the connectome
The connectome is, in essence, the wiring diagram of the brain. It's a physical map of all the neurons in the brain – or indeed the whole nervous system – and synaptic connections between them.
And this is something we can physically see, through certain different imaging technologies, at a range of different scales. Take a cube of brain matter, slice it finely enough and then create a 3-D model of the cube, and you begin to see the staggering physical complexity that even the simplest invertebrate's brain requires.
At the macro scale, thick neuron branches curl around and intertwine with countless others. The shapes are endlessly complex – imagine a densely packed bowl of spaghetti but with each strand varying its shape, thickness and 3-D geometry all the way down, to accommodate other strands and connect with them multiple times.
The further you zoom in, the more complex things become, as smaller and smaller neurons branch off from the larger ones, connecting to others of all sizes at specific points called synapses, where information can be exchanged in the form of neurotransmitters and electrical impulses.
It's a colossal oversimplification, but the theory is that each of these pathways, with all their touchpoints, branches and connections with other pathways, might represent a thought, memory or action. The more you do it, or think it, the larger and thicker that pathway becomes. The more you learn about it, the more physical connections it forms with other neurons and processes, further building your map of understanding.
Thus the map of a brain changes over time, constantly reinforcing some pathways as others are left to die from disuse. We might indeed be prisoners, in a sense, of our connectomes in that they determine exactly what thought patterns will come out of which situations. But by the mere act of thought itself, we have the ability to re-wire ourselves to a significant degree, to enlarge some pathways and allow others to drop off. The more you think something or do it, the more a physical part of your brain it becomes – and it's just as possible to let parts atrophy if you ignore them.
And while we have literally no idea how to separate function out from these infinitely complex forms, it appears to be a physical manifestation of how we believe things like memory, behavior, personality and understanding might work. In theory, if you could completely map the connectome of an individual, and understand which bits did what, you could build a complete model predicting what that person would think, say and do given a set of circumstances.
Where's connectome research at right now?
At the moment, it's possible through the use of different types of MRI imaging (functional and diffusal) to map some of this stuff at a macro level in living subjects.
This is already starting to help us understand certain behaviors and phenomena; researchers are trying to use diffusion MRI mapping to trace how things like depression and anxiety disorders disrupt healthy brain function. The advantages of live mapping are clear – you can get the subject to do or think things, track what brain activity is happening where, and use that to build some idea of the function of each of the areas you can distinguish.
But while MRI imaging in live subjects is more likely to be useful in the short term, the technology is far too low-res to reach the ultimate goal of a full human connectome map. MRI can highlight the brain's superhighways, but it can't get down to the level of the backroads and walking trails.
To get down to the cellular level, which is the ultimate goal, you currently need a dead patient. One whose brain you can chop into small blocks, then slice those blocks and analyze them with light and electron microscopes, stains and other tools in order to recreate a high-resolution map that allows researchers to trace, mark and depict every neural pathway and synaptic connection out in 3D.
Here's an extraordinary video by researchers at the Max Planck Institute's Department of Connectomics, showing how 2-D layer slices can be processed into a 3-D model of the connectome for a tiny piece of brain:
To call these maps complex is another colossal understatement. When researchers fully modeled the connectome of a C. elegans worm, which has only 302 neurons to trace, they ended up needing 12 terabytes of data to store it.
A map of a human brain would be more in the order of two million terabytes. That's purely to describe the physical structure of it. If you wanted to build a functioning model of a person inside some ultra-computer at some point in the future, you'd most likely also need to know what the hell each of those pathways does in a functional sense.
Let's not forget either just how much of that spaghetti tangle of neurons deals with phenomena specific to the body the brain's in. Because it connects to all kinds of physical systems and sub-systems, from the sensory inputs, to unconscious reaction circuits, to hormone secretions, motor impulses and background organ regulation. Reconstruct the connectome of the brain without all the other bits, and goodness knows what systems will be able to function when so much of the body and input mechanisms are gone.
And assuming even that this gargantuan set of infinitely complex and connected problems can be solved, and you could run a model of yourself in a computer – a model of yourself at a specific date and time, of course, because neuroplasticity ensures that "you" is a constantly evolving concept – there remains the question of whether that model could be said to be alive, or to have a consciousness. Or if it's merely able to do a really good impression of you.
But it does seem possible that this kind of technique, applied to a correctly preserved brain many years after death, could be used to extract memories and personality and behavior patterns that could give subjects some legitimate claim to immortality. If a personality is something like an idea, and an idea can be propagated forward through different substrates, perhaps a personality can too. And it's certainly cool to think about uploading your personality into some 30-foot tall rocket-powered Godzilla robot.
Preserving the brain
As we move forward, we learn more and more about how much we don't know. While brain science has progressed at a similarly astronomic rate to the other sciences, it's a subject of such immense complexity that it's likely the human brain literally isn't smart enough to understand its own workings. The known unknowns continue to pile up, and it seems we'll have to rely on machines and software that can think faster and in more dimensions than we can, in order to untangle the comically gigantic mysteries of the connectome.
But one thing seems clear: if we're going to persist with the business of cryonics – freezing brains or whole bodies in the hope of resurrecting them someday – we need to make sure we're preserving them in a way that supports the reconstruction or mapping of the connectome at the highest possible resolutions.
And the best way we know at the moment is Aldehyde-Stabilized Cryopreservation, or ASC. It's a two step process. First, a fixative chemical called gluteraldehyde is used to rapidly transform the brain from its normal, watery texture to something like a softish rubber. This freezes the neuronal pathways in place before they can begin degrading, and gives you a brain that's stored as intactly as science can currently imagine it, for up to a year or two without losing information.
The second step is vitrifixation – freezing the fixed brain for long-term storage. Using the same ethylene glycol you put in your car as antifreeze, the brain is protected from the damaging formation of ice crystals, and then it's plunged into extreme cold at -122° Celsius (-188° Fahrenheit) where it can be kept more or less indefinitely.
And it appears to work. In 2010, the Brain Preservation Foundation put up a cash prize for the first individual or team to rigorously demonstrate a technique capable of "inexpensively and completely preserving an entire human brain for long-term (>100 years) storage with such fidelity that the structure of every neuronal process and every synaptic connection remains intact and traceable using today's electron microscopic (EM) imaging techniques."
There's two stages to this prize – successfully preserving a small mammal brain, something like a mouse, and successfully preserving the brain of a larger mammal such as a pig. And in 2015, Robert McIntyre and his team nailed stage one, to the satisfaction of the judges, preserving a rabbit brain using the ASC techniques described above, and keeping the full connectome mappable in the process for the first time. Here's a video moving down through the layers of that preserved rabbit brain, that shows how you can follow the paths in this 3D structure:
Which of course raises the uncomfortable question: what about all the other brains that have already been frozen? Was the technology of the time insufficient to hit the moving target of preserving exactly we want to be able to get back out of them?
Nectome and the quest for the perfectly frozen human brain
Does McIntyre believe the technique will scale to include human brains? Yes. Indeed it's the whole point of the exercise. And he's teamed up with a fellow MIT graduate to form a company, Nectome, that he hopes will someday be able to preserve and store brains with a level of precision that will allow future technologists unfettered access to their entire connectomes, including all the memories and personality traits that are locked within.
Nectome's short term goal is to perfect the ASC preservation technique in a way that can be used on human brains – something it sees as a two-year goal. But it's also working on the complex problem of working out how to map the connectome using various color staining and imaging techniques.
The company is optimistic that within a few short years, scientists will figure out a way to extract high-level bits of memory from mouse brains, and that advances in computational technology will let them fully map a 10,000-neuron brain as early as 2022.
That's still a ways off from the ~100 billion neurons you'll find in a human brain, but at that point it's really just a matter of scale and processing power.
What to do with it once we've fully encoded the biological structure? That's another question, for another generation of scientists. Is preserving the connectome enough, or is this technology going to become sadly obsolete the next time we take a leap in understanding? Perhaps. But merely mapping out the connectome, and devising the tools to properly prepare a brain for long-term storage? That's something humanity can pat itself on the back for.