The Race to Map the Brain So We Can Upload It Into a Computer

Neuroscientists talk about mind uploading a lot more than you'd expect.

It's 2015, and we can't even map a mouse brain, let alone a human one.

"Nobody really knows how the brain is wired," neuroscientist Shawn Mikula, who recently discovered a breakthrough method of preserving brains for imaging, said in a Skype call with Motherboard. "Mammalian brains in general has been a huge black box."

There are several reasons we'd like to map the connections in the human brain—or at this point, any brain. With an understanding of how brains work at a neural level, scientists hope they can find the cause of brain disorders like Alzheimer's or schizophrenia.

Researchers at Washington University and Harvard have pursued producing a complete diagram of the human brain—known as the connectome—through MRI scans and an online game played by volunteers. Both efforts fail to provide precise microscopic synaptic transactions between neurons, however, which is needed to understand how the brain works on a fundamental basis.

The only existing technology available for imaging the circuitry of the neurons is electronic microscopy, which involves cutting the brain into ultrathin slices, imaging them under an electron microscope, then stitching the images together. Despite the highly detailed images, the thin slices sever larger neural circuits that circle the brain that cannot be captured in one slice, distorting an accurate image of the larger neurons. By focusing on the smaller, individual neurons, connectomics loses sight of the whole picture.

Proponents say the mind is nothing more than the relationships and behaviors of neurons, which can be translated into algorithms

Back in May, however, Mikula published research describing a method to prepare a brain with a new chemical solution so that it can be imaged in great detail without requiring any cuts to the brain.

In electronic microscopic experiments, the brain samples were stained with heavy metals, like osmium, so they would react with the electron beams of the powerful microscope. But because the heavy metal solution could not penetrate deep into the brain membranes, forcing scientists to cut the brain into samples as thin as 20 nanometers. Through trial and error, Mikula found that mixing formamide into the solution could stain an entire mouse brain deep enough.

Hailed as a milestone in brain preservation, Mikula's method—known as brain-wide reduced-osmium staining with pyrogallol-mediated amplification, or BROPA—has cleared a barrier in electronic microscopy preparation and may open doors into imaging larger mammalian brains, like those of a pig or a human.

"Being able to look at any corner of the exact same mouse brain with nanometer resolution is the only way to unambiguously map all neural connections," said Albert Cardona, a neuroscience researcher at the Janelia Research Campus in Maryland, in an email. "Until now, researchers could only image tiny bits with electron microscopy, due to limitations in both sample preparation and imaging technology. Mikula's work overcomes the sample preparation road block."

In the grand scale of brain mapping, BROPA is a tiny addition as it doesn't directly deal with figuring out how the brain is organized. But Mikula believes that BROPA is a necessary tool to further progress within the field still heavily shrouded in mystery. With BROPA, Mikula hopes the mouse brain will be mapped out in entirety in the near future.

Neurons mapped after slicing a brain into thin pieces, imaging them one by one, and combining them into a composite. Image: Shawn Mikula

According to Mikula, the memory size of the first edition atlas of a mouse brain will be around 40 to 60 petabytes, or 40 million to 60 million gigabytes. The memory size of a human brain atlas is expected to be several thousand times larger than that of a mouse brain.

Mikula hoped that the success with a mouse brain will lead to a whole brain imaging of a small primate like a marmoset within 10 years. He doesn't have a timeline in mind for human brains, but he's generally optimistic.

"I know (human brains) will happen because the momentum is building," Mikula said. "Pretty soon governments will put the money forward and assemble a team of people to do a human brain at the electronic microscope level."

With BROPA paving new ways for brain preservation for future research, neuroscientists is now pondering what other technologies are feasible when the connectome is completed.

Beyond the basic ways having a map of the brain could advance medicine and science, neuroscientists talk about how an accurate map could be used to reproduce the brain in a computer. We could then fashion an artificial intelligence, or even more weirdly, upload the memories and personalities of an individual into an artificial operating system to create their digital clone.

This weird process, called "mind uploading," is controversial in the scientific community—although it's not as farfetched as you might think.

A stained mouse brain embedded in epoxy and viewed from the top. Image: Shawn Mikula

Proponents of mind uploading, including Hayworth, argue the individual mind is nothing more than the relationships and behaviors of neurons, and that these neural relationships can be translated into algorithms inside a computing system.

Skeptics like Mikula, however, believe it is impossible because of the physiological complexities in the brain that go beyond neurons, which he thinks cannot be replicated on a computer.

Mikula hopes that his discovery—and future progress in connectomics—will help us not to preserve the human mind via mind uploading, but instead help us examine what it means to be a human.

"It will be an awakening and help us define ourselves more and understand what we are," Mikula said. "Maybe some people prefer introspection or doing other thing to understand who we are, but this is my thing: to see what we are at the neural level."

Jacked In is a series about brains and technology. Follow along here.