21st Century Thought Reading Technology May Soon Over-write Freud

Original article can be found here

Ever wondered what Google, Frankenstein, insulin and the theory of relativity have in common? Other than being some of our greatest creations and discoveries, they have all arisen from the swampy nether of the sleeping brain. Some of the humanity's most celebrated achievements were, literally, products of dreams. Ironically enough, the underlying mechanism of sleep and dreaming was dreamt up too, amongst a vast greatness of other, surprisingly numerous, slumbersome discoveries. Wouldn't it be wonderful if we could consciously immerse ourselves in this elusive underworld of dreams, exploring our hidden genius and treating a neurological disorder or two along the way? As frightening as this may sound, we may no longer be as distant from such a reality as we'd perhaps like to think.

And the reason for this is simple: as of late, we have gotten exceptionally nifty at invading the mind. Several neurotechnology leaps ago it was 1973 and doctors had to painfully drain blood vessels in the brain in order to X-ray the empty spaces left behind. Today, 20 minutes inside a non-invasive functional magnetic resonance imaging (fMRI) scanner is all it can take to unveil our emotional state, mood, and even basic thought. Combine this with other neuro-imaging techniques such as diffusion tensor imaging (DTI), electroencephalography (EEG), magnetoencephalography (MEG) and positron emission tomography (PET) and neuroscience can now paint a pretty picture of our brains' inner workings, detecting and treating psychological trauma and mental disease and exposing our deeply-rooted emotional responses to anything from our favorite show to our next-door neighbor.

The latest breakthroughs in fMRI imaging have been the most striking. Recently neuroscientists were able to "communicate" with patients previously believed to be long comatose by detecting activity spikes in brain areas responsible for visual processing, emotion, and so on, which are otherwise un-translated into visible signs of consciousness. A step further from this practice is actual visual "decoding" - reading and replicating vision from brain scans. In a series of visual decoding experiments, volunteers were asked to watch short clips inside an fMRI scanner. Scientists then used the external output to replicate clips of what the person was seeing. Below is a video demonstrating visual reconstruction from an fMRI scanner: 

 The incredibly discernible albeit unpolished, fMRI output attests to one frightfully simple fact: mind reading is at our doorstep. Whether we like it or not, growing data samples and meticulous calibration will continue to yield increasingly accurate brain reads until we can stream brains onto television screens.

But no need to panic just yet. The level of accuracy required to breach our darkest secrets is certainly a long way away, but this hasn't stopped the neuro-inquisitive from exploring all possible applications of our newfound mindreading ability. The subconscious is perhaps the ideal candidate for this, for the simple reason that we still seriously lack reliable psychoanalytical tools for penetrating such an elusive realm.

And what better starting point for this than dreaming. Dreams are so mysterious, liberating, frightening, life-changing and sometimes so outright odd that for centuries we have speculated on their message to no definitive avail. The famous "Interpretation of Dreams" by Sigmund Freud is a book of psychoanalysis and controversy, offering at times outlandish interpretations of shapes and figures inhabiting the dozing mind. According to Freud, "Dreams are the royal road to the unconscious", revealing glimpses of a secluded backdoor unbeknownst even to its beholder, but, on a large scale, riddled with sex symbolism and fueled by carnal instincts. Met with much skepticism, to this day "Interpretation of dreams" is lying in wait of more compelling evidence, something which our ever-expanding neurotech prowess is likely to soon afford.

If we were to trust the importance Freud attributed to dreaming, one can only imagine the untapped potential lurking in obscure cranial crevasses, offering anything from hugely ameliorated psychotherapy to unleashing hidden genius.

In the sphere of dream research, impressive progress is steadily being made. In an a series of fMRI sleep experiments Horikawa and colleagues from ATR Computational Neuroscience Laboratories, in Kyoto (Japan) caused a few seat stirs when they presented first-of-a-kind evidence of dream decoding. In what is now a peer-reviewed publication in Science, the team unveiled decoded readings of dreams achieved with an fMRI scanner much the same way as in the visual reconstruction experiments. Subjects were allowed to catch 10-minute z's inside an fMRI scanner. As soon as they entered REM sleep, they were awoken and questioned about the content of their dreams. Scientists then scoured the internet in search of images which would match the dream description. When these images were shown to the subjects inside an fMRI scanner, the brain activity spikes they induced were surprisingly similar to the spikes detected from the subjects' dreams. After appropriately calibrating the dream- and wake- signals, scientists scanned sleeping subjects once again, this time "decoding" the dreams and confirming their findings with the subjects, with a crude-but-impressive 60% accuracy on basic shapes and themes (ie -person, street, tree, etc).

For starters, the findings have unveiled a previously unknown fact: equivalent spikes generated in sleep and in waking suggest that dreams, like images, at least partially reside in the visual cortex (what happens in the case of blindness remains to be studied). Secondly, a 60% accuracy is certainly not bad for a first try, considering that the experiment was targeting dozens, if not hundreds of possible dream scenarios. Most importantly, the fact that calibration improves with every scan means that, when it comes to dream decoding, the only way is up. Perhaps dream reading will not go as far as to yield us downloadable dream databases every morning, but it will certainly give scientists the opportunity to study various dream patterns in disease, injury, and alongside a plethora of other parameters. And, with time, this research will doubtlessly begin to give meaning to the enigma which has vexed humanity for so long.

The (understandable) Confusion in Brain Mapping Terminology

The blind can see, the deaf can hear, and the paralyzed can walk—without a doubt, the Golden Age of Neuroscience is upon us. And if that weren’t enough, mind control, telepathy and in fact elaborate real-life Avatar scenarios are so nigh that scores of the neurotech-savvy have begun to pre-order their titanium thought defense helmets. So awesome have the leaps of neurotechnology been in recent years that the media has struggled to deliver this industry a deserved timely coverage, and public awareness has just barely caught up with humanity’s neuro-prowess of late. And what came about was the usual by-product of a much-too-quickly advancing technology: terminology fails. 

(view original image here)

On April 2, 2013 President Barack Obama unveiled the BRAIN (Brain Research through Advancing Innovative Neurotechnologies) initiative, whose final and uttermost objective is to visualize, map, understand and reconstruct the activity of every single neuron in the brain. This knowledge will in the future be applied to the development of new technologies and the treatment of quickly expanding array of neurological disorders.

BRAIN’s near-sighted primary goal will be to sponsor ongoing brain mapping projects, such as the Human Connectome Project, which the initiative claims will help us visualize every neuron in the brain and understand the formation of synapnses (connections) between neurons in live brain tissues. As the Human Connectome Project deals primarily with brain scans of live humans and their brains, this happens to be a feat which, for the time being, any knowledgeable neuroscientists will tell you is utter nonsense.

Here is why. 

The Human Connectome Project (HCP) has been dubbed one of the greatest neuroscience ambitions of the new millennium, much akin to the Human Genome Project a decade ago. The project showcases efforts to fathom the complex macrocosm which unites every axon bundle in the brain with the use of high-resolution fMRI or other non-invasive techniques, and symbolizes the positive level of trust which neuroscientists now assign to neurotechnology. HCP precedes and perhaps supersedes the BRAIN initiative, as already in 2010 it secured grants collectively worth over $40 million from the NIH. Initial 1200 Connectome subjects will be scanned for hours using an fMRI scanner 4 to 8 times as powerful as a standard one, utilizing the power of one nuclear submarine per scan, in hopes to yield insight into the connectivity of nerves within a brain, and how this connectivity varies throughout species and in disease. 

With novel neuro-terminology threatening to become its own language, much confusion and mis-representation surrounds HCP, and naturally the media hype it triggered has seemingly gotten lost in translation –or, rather, in literary terms. In bold truth, the techniques which the HCP is set to exploit are entirely non-invasive brain scans whose resolution is equivalent to that of a rusty security camera. Although general brain regions of humans will be clearly visible, local inter-region synapses, which form 80% of connections in the cerebral cortex, are entirely inaccessible to supracranial scans, as even a light microscope of the highest magnification is insufficient to visualize a single neuron. 

In other words, an fMRI scanner does not, and probably never will, visualize single cells and local synapses in the brain. However, HCP may yield great insights into the "connective visuality" of the brain, as vibrant HCP images (such as the one below) show thousands of connections between particular brain regions via axonal bundles (which are cables composed of many nerve cells, a bit like your landline telephone wire, not single brain cells). Scientists say that it is this inter-region connectivity which is responsible for our superior intelligence to other species, as preliminary scans clearly show a much more "connected" human brain in comparison with that of a chimpanzee, for instance. Scientists also believe that comparing such deep brain scan images between healthy and ailed individuals can help us understand neurological disease in the future.

Does this imply that we shall never be able to dwell into the microscopic ecosystem of grey matter? Not at all, although it may take some time until we are able to do so in humans. Non-human species, however can now be studied with a dozen of other, much more invasive techniques you probably wouldn't want performed on yourself. These techniques' findings collectively contribute to the general brain Connectome (note the unnecessary similarity to the Human Connectome Project) across all species-which is, in essence, a brain equivalent of the genome. To make matters more ponderous, Connectomics is the study of the connectome with all the research techniques available today, comprising MRI, electron microscopy, fluorescent cell staining, rendering brains see-through and a fantastical endeavor which goes by the name of optogenetics. 

One rather straight-forward technique used to visualize individual neurons and synapses in the cortical column is, literally, brain carving. A team of researchers at Harvard, led by prominent brain explorer Dr. Lichtman, have been working towards creating the most efficient contraption for slicing off pastrami-thin slices of mouse brain and observing them under a powerful electron microscope. Theoretically, this process may yield a complete 3D cellular reconstruction of a brain, but it would require creating and aligning thousands, if not millions, of brain slices. 

Amongst other, less brute-force neuronal visualization technique is the Clarity Project, which helps visualize neurons in brains of live mice. The Clarity project revolves around scientists' ability to chemically produce completely transparent brains (or other organs) in mice--a technique which was recently developed by Prof Karl Deisseroth and his team at Stanford. The chemical treatment strips away lipids which normally block the passage of light using the detergent SDS.

(original image here)

To help them through the treacherous cranial terrain scientists at Dr Lichtman's Dr Diesseroth's labs utilize fluorescent staining of neurons, which helps isolate single cells spanning a number of brain areas. The staining produces images such as this, termed BrainBows: 

(original here)

But perhaps the most compelling technique developed at Deisseroth's lab is optogenetics - a research tool which revolves around a gene discovered in 2002, encoding light-gated ion channels, or Channelrhodopsins, found in unicellular green algae. Upon activation with a specific-frequency light, the channels open, causing an ion influx into the cell much akin to the mechanism which causes our neurons to fire. Using a different light frequency can also activate a different set of channels—the Halorhodopsins which will cause a negative ion influx into the cell and thus inhibition of the neuron. For algae, this mechanism’s function is to orient the cells towards or away from light. But this mechanism can also be applied to mammalian cells.

(original image here)

In recent years scientists succeeded in engineering rhodopsins from algae into nerve cells of mice, either grown externally or, in fact, in perfectly alive animals. In essence, genes encoding these channels can be “injected” into very specific areas of the brain, which will render those areas light-sensitive. By shining a light with the use of optic cables into those areas, scientists can trigger very specific neurons, and examine the animal’s behavior, thereby demonstrating the role of the specific neurons they are studying.

Click here for a video of an optogenetically-challenged mouse commanded by scientists to go in perpetual circles. 

(original image here)

Ultimately, all the techniques which have armed neuroscience in recent years work towards one ambitious goal: to comptuationally simulate the workings of a brain, which would give us the ability to re-create, and perform "virtual healing", on neurological ailments. A breakthrough milestone was achieved in this sphere of connectomics in 2012: scientists were able to unveil key principles which guide the formation of neuronal interceptions, and applied these principles to a computational structural prediction model. In this way they were able to demonstrate that brain neurons grow independently of each other, and form connections once they intercept each other’s paths in the brain. 

Some might argue that virtually recreating a self-sufficient human brain capable of conscious thought like you and I might not be a great idea, but that's for later debates (we are not there yet).

Bioassociate will soon publish an in-depth report on the Global Neuroscience Revolution, where many recent fascinating developments, techniques and initiatives will be described. If you wish to receive a copy of the report please subscribe to our mailing list on our home page: www.bioassociate.com