Upload Complete: Transferring your brain to a digital format

Posted by Susan Lubejko on July 19, 2018 in Neuroscience, Philosophy | Leave a comment

Imagine a future in which your brain–conscious mind, memories, and emotions included–could be uploaded to a computer. Once running, the complex computer code would be able to perfectly reflect your mind: reacting as you would, showing your major personality traits, holding your unique knowledge from your life as a human. Maybe your personal code could be transferred to a robotic form, allowing “you” to carry out your physical demands as well. Would you still be you?

Though it sounds as though this idea would be confined to a familiar episode of Black Mirror, the technique at the intersection of biology, computational neuroscience, and philosophy at play here is called whole brain emulation (WBE). WBE entails taking whole human brains and imaging them in such a way as to understand the fine detailed structure of each of their 86 billion neurons in order to model each in a large software network. Once running on the right hardware, the software should act exactly as the physically connected neurons would–a principle authors from the Oxford University Future of Humanity Institute refer to as creating a “1-to-1 model,” since every single relevant property of the physical neurons should be accounted for in the computational code [1].

Computational and Mathematical Modeling of Brains

The idea to use mathematical equations and computer algorithms to simulate the activity of single neurons and networks of interconnected cells is not new to computational neuroscientists. Math equations have been derived to explain how electrical activity moves down the length of a specific neuron’s dendrites or axons and which types of ion channels are present in that cell’s membrane. Others detail the form and function of the synapses, the tiny gaps between neurons that allow for the transmission of information from one to the other, that cell creates with other neurons. Additionally, other types of models simulate how multiple neurons in a network communicate with each other [2].

And these models often do a very good job of recreating and helping to explain physiological data that neuroscientists can record from real neurons. A caveat in neuronal modeling, however, is that the neuron or network is represented in a simplified fashion. A single set of equations may be used to represent a neuron with a lot of dendrites that differ in diameter and membrane type, but this simplified fashion of modeling is insufficient for a “1-to-1 model” for WBE. Instead, computational neuroscientists would have to painstakingly determine a path to represent each and every neural property in every neuron in the brain in order to upload a brain with complete fidelity.

Purkinje cells of the cerebellum are known for their complex dendritic structure. Imagine representing these complex cells in computer code! Source: https://en.wikipedia.org/wiki/Purkinje_cell

Where do we stand on the achievement of WBE?

With 86 billion neurons and hundreds of trillions of synapses in the average human brain, this is an overwhelming task to take on in the eyes of many. However, a growing number of researchers, theorists, foundations, and companies are up to the challenge. The nonprofit organization Carboncopies Foundation aims to support collaboration and the development of technology for WBE efforts in hopes to create “substrate-independent minds” that can aid our fallible and fleeting human memories by giving us a technological place to store them, eventually eliminating the need for just one, permanent physical form. The foundation plans to update the original WBE guidelines set out by those Oxford University professors with newer technology that has become available.

As it currently stands, WBE is not yet in reach. Most researchers working on WBE accept that the best way to extract enough structural information from a brain for reliable computational encoding is to preserve full brains, slice them into super thin sections, and look for synapses to reconstruct. Synapses are key–the theory of WBE relies on the full reconstruction of the connectome, the map of all neural connections in the brain. Synaptic structure, such as the size of the axon terminal, the width of the space between the cells, and the amount of signalling molecules that are ready to be released, is especially important for transfer of existing memories, as research suggests that long-term memory exists as a strengthening of synapses between specific neurons that are active together [3]. Essentially, synapses may be what makes you, you.

A major technological roadblock to mind uploading lies in data acquisition and processing. Synaptic connections between neurons can only be fully elucidated using the high magnification and small fields of view of a super high-resolution method of imaging called electron microscopy. A research group at the University of Utah using a much lower magnification to take microscope images of a small subset of the rhesus macaque’s 6 billion neurons suggests that their dataset will include 600 terabytes of image data per animal [4]. Now imagine the computing power needed to image the approximately 86 billion neurons of the human brain, where cerebral cortex neurons have an estimated 7,000 synapses each [5], at much finer magnification to reconstruct and model individual synapses.

Electron microscopy image of a synapse. The presynaptic neuron contains small spheres filled with neurotransmitter signalling molecules that are released onto the receptors on a postsynaptic neuron. Source: Hines DJ, Haydon PG (2014) Phil. Trans. R. Soc. B 369:20130594.

Additional biological and philosophical questions remain as well. Since our brain is constantly bombarded with sensory information from our environment, would a brain in computer isolation function in the same way? Is the physical structure of synapses sufficient to understand how the synapse functions? Are the genetics of individual neurons or the presence of different molecules in and around the neurons important for how we store memories and personality as well? What role might important support cells in the brain, called glial cells, play? Finally, a hallmark of human existence is consciousness. As the neuroscience field continues to search for the seat of consciousness in the human brain, the question remains as to whether our uploaded selves would continue to show this trait.

Brain preservation & strides toward WBE

The biggest strides in WBE research have come in the form of improved brain preservation techniques. The process of preserving full brains for long term storage, until our technology has advanced enough to meaningfully use them, is notoriously difficult. Brain tissue degrades quickly with a lack of oxygenated blood after death and often does not stand up well to freezing and thawing procedures. Companies such as Alcor that offer cryonic preservation, the process of preserving full bodies by keeping them frozen in hopes of future reanimation, try to use combinations of embalming with preservative chemicals and freezing in liquid nitrogen whenever possible. They also monitor for fracturing events, where supercooled tissue loses its elasticity and cracks under the cold temperatures.

The most significant new addition to the field of brain preservation occurred in 2015. Scientists from 21 Century Medicine introduced a new method of vitrification–transforming a solid into a glass-like substance for preservation–in rabbit and pig brain trials. The team uses the brain’s own circulatory system to introduce a fixative (similar to embalming) solution, followed by the chemical ethylene glycol, a common component of antifreeze in cars, and eventually deep freezes the brain [6]. Through intense judging by experts in diagnostic electron microscopy, the Brain Preservation Foundation, a foundation dedicated to the improvement of brain preservation specifically for WBE advancement, awarded its Technology Prize to the 21 Century Medicine scientists in early 2018 for the technique’s near perfect preservation of the structure of individual synapses throughout the entire brain of a large mammal.

21 Century Medicine’s submission for the Brain Preservation Foundation Technology Prize. The vitrification technique near perfectly preserves the entire structure of the brain of a large mammal. Source: Brain Preservation Foundation, brainpreservation.org

One of the authors from 21 Century Medicine, Dr. Robert McIntyre, has since moved on to co-found Nectome, a startup company committed to continuing research into how to keep memories intact in preserved brains. A 2018 article in The MIT Technology Review suggests that Nectome has already used its preservation technique on a recently deceased human brain, and may have already collected refundable deposits from consumers interested in preserving their own brains with the technique. As Nectome’s founders remind us, however, the process is fatal–brains must be preserved before or at the moment of natural death for physical connections to remain intact. Controversy surrounding these statements has lead to a break in a collaborative relationship between Nectome and prolific MIT neuroscientist Dr. Ed Boyden, and Nectome’s website now includes a statement that emphasizes that the team has no plans to use the preservation technique on humans for long-term brain storage in the foreseeable future.

 
Are you in?

Whole brain emulation could give us a shot at immortality, but many questions remain about the feasibility of making a digital copy of one’s brain. The answers will undoubtedly require huge technological advancements while perfectly-preserved brains remain in storage. Are you ready to sign up?

References

1. Sandberg A, Bostrom N (2008) Whole brain emulation: a roadmap, Technical Report #2008-3, Future of Humanity Institute, Oxford University.
2. Koch C, Segev I (Eds.) (1999) Methods in Neuronal Modeling: From Ions to Networks, 2ndedition. Cambridge, Massachusetts: The MIT Press.
3. Mayford M, Siegelbaum SA, Kandel ER (2012) Synapses and Memory Storage. Cold Spring Harbor Perspectives in Biology. 4:a005751
4. Landhuis E (2017) Big brain, big data. Nature, 541:559-561
5. Pakkenberg B, Pelvig D, Marner L, Bundgaard MJ, Gundersen HJG, Nyengaard JR, Regeur L (2003) Aging and the human neocortex. Experimental Gerontology, 38:95-99
6. McIntyre RL, Fahy GM (2015) Aldehyde-stabilized cryopreservation. Cryobiology, 71:448-458

Cover Image Source: Brain Preservation Foundation, brainpreservation.org