Putting a Dream on a Screen
Abstract:
The following paper was a 2014 review of the science and engineering that investigates the phenomenon of dreaming. An addendum as of 2017 has been added at the end. An overview of the field and its prospects is described, followed by the benefits such research will be able to deliver. Current and future work is also described. The paper is oriented around the question “Will it ever be possible to record a dream and view the recording on a screen?”
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dreams have been long revered yet poorly understood throughout human history. Some of the oldest artwork found depicts the journey of a human soul into the ‘dreamworld,’ and dreams remain a popular subject of contemporary art, indicating the persistence of human intrigue in dreaming. In spite of this lasting curiosity, some pivotal questions remain: where do dreams come from and why do we dream at all? While there are still different opinions on the answers to these questions, research in neuroscience and psychology has already revealed that some of the basic assumptions about dreaming from the twentieth century were incorrect. A new kind of science is being built from the rubble of these insights. The experience of dreaming is being connected to specific parts of the brain and new experimental tools have emerged in recent years that might enable researchers to one day see the dreams of their subjects or perhaps even trigger the onset of dreams. Ultimately it will be the work of neural engineers to build the foundation for the design of these devices that record, induce, and manipulate our dreams.
The Difficulties of Dream Research
The main difficulty with dream research, also known as oneirology, is the challenge it presents to the scientific method. The scientific method holds that in order to draw reasonable conclusions from data, an experimenter must be able to control a variable in the experiment. It is hoped that this controlled variable will validate or invalidate the hypothesis proposed prior to the experiment. Currently, controlling a variable of the dreaming process isn’t possible.
In a personal correspondence, prominent dream researcher G. William Domhoff explained “[T]he problem is that you can’t make [dreams] happen, can’t see them while they are happening, and can’t get any after-the-fact info on them, except through [questionable] verbal reports.” To psychologists like Domhoff, this means that there is a lack of control that “psychologists rightly desire to have.” Aside from this methodological block to oneirology, the research in and of itself is expensive; subjects must be paid to spend a night or more in a lab, and researchers must be willing to do the same [1]. Some solutions to this problem might include automating the roles of researchers so only subjects are in the lab or just collecting dream reports and biological data from a subject sleeping at home.
There appear to be no immediate capital benefits from studying dreaming in a clinical context. This lack of foreseeable profit combined with the difficulty maintaining scientific rigor in the research itself makes for an environment in which there is very light funding and very few opportunities for academic recognition. This combination of poor conditions unfortunately scares off qualified researchers from studying this fascinating and deeply mysterious field. While these conditions make it difficult to study the workings of the dreaming mind, there are many legitimate reasons for future scientists and engineers to think deeply about this subject.
The Benefits of Oneirology
The National Academy of Engineering announced a set of problems at the turn of the millennium, which they have described as “Engineering Grand Challenges.” These challenges represent the greatest problems facing humanity and invite engineers to consider solutions of their own. Included among these challenges are “reverse engineering of the brain,” “enhancement of virtual reality,” and “personalization of medicine.” These are all efforts that would benefit from a deeper understanding of the biological basis of dreaming. In fact, oneirology is intrinsic to the endeavor of reverse engineering the brain. To fully understand cognition, consciousness needs to be explained in terms of the way the brain actually functions [2, 3], and these explanations need to be able to answer questions about the way the mind functions while it is both awake and asleep if they are to be considered complete.
The National Academy of Engineers lists virtual reality (VR) as a necessary tool of the twenty-first century, and research into dreaming is the key to the holy grail of VR: full immersion. A strong evolutionary explanation of dreaming doesn’t currently exist, yet it is a fact of life that nature has crafted a fully immersive virtual experience for us each night. With deeper investigation into the neural networks responsible for dreaming coupled with technological advances in brain-computer interfacing, it is not difficult to conceive of a system that uses the dream as the medium for constructing a virtual world.
The NAE imagines virtual reality to be an indispensable tool in the future of instruction and experimentation in fields such as medicine, warfare, and education. In fact, virtual reality is already a central tool in teaching pilots how to fly and surgeons how to operate. Engineered dreaming can be expected to enhance the utility of these tools and enable teachers to construct more realistic exercises for students to train with.
The final challenge mentioned, the personalization of medicine, will be deeply affected by dream research when a medical theory of dreaming emerges. Dreaming is one of the most personal experiences that a person may have. Presently there is no way to know for sure whether or not a dream can broaden a caregiver’s perspective on the wellness of a patient, but this is likely the case. Dreams are a result of the systemic intersection between our minds and bodies. Therefore it seems that by mining our deeply personal and (mostly) uncontrollable psychological events we can better analyze the state of our health. Imagine the value in thoroughly analyzing the content of a patient’s dreams to accurately diagnose and effectively treat mental or physical illness. This clinical approach stretches back from Freud’s psychoanalysis to the Asclepeion of ancient Greece. And while psychology has come a long way since the 1900’s, dreams remain an attractive diagnostic tool for doctors that desire a more holistic perspective on their patients.
Current State of Research
Dreams had been the object of twentieth century academic inquiry under Freud and Jung for many years. Unfortunately, the contention between the two camps and the lack of tools that produced concrete evidence about the dreaming brain led to a dwindling in research as the century marched forward. But in 1953, two graduate students studying the physiology of sleeping subjects discovered that rapid eye movements (REM) were often associated with reports of dreaming [4].
This finding greatly renewed interest in the study of dreaming and its connection to the body. This landmark work led to the discovery that sleep occurs in different stages. These different stages of sleep indicated to scientists that a subject’s propensity to dream increased as the night wore on, which happened in parallel with increasing episodes of REM, as shown by the widening gaps of REM stage sleep in Figure 1. In the 1960’s a French researcher named Michel Jouvet investigated the neurochemistry responsible for these shifting physiological states in subjects during sleep [5]. In his research with cats, he found that aminergic neurotransmitters like dopamine and serotonin were at least partially responsible for the waking portion of consciousness, known as vigilance. In sleeping animals, he found that acetylcholine acted in the body at high concentrations, and was therefore most likely to be tied into the chemistry responsible for sleep. Jouvet’s research seemed to point to dreams as a sort of ‘paradoxical’ chemical state of the body, where amines and acetylcholine fought for control of consciousness, producing a state of vigilance within an otherwise unconscious body. These findings have led some self-experimentalists to utilize drugs to induce dreams, like galantamine, calea zacatechichi and even nicotine, which apparently modulate the amine-acetylcholine balance in the body through their metabolic action [6].
Outside of a pharmacological domain, there is research being done using subjects that have the ability to willfully enter into a lucid dream. Lucid dreaming is a state of consciousness in which a person is in a dream and concurrently aware that they are in a dream. Using a technique developed in the 1980’s by Stanford researcher Stephen LaBerge, lucid dreaming subjects signal to researchers that they are conscious while in the dream state by moving their eyes in a synchronized pattern [7]. This method of synchronized eye movement convinced most researchers of the reality of lucid dreaming in the 1980’s, even though Tibetan Buddhist monks have written about the phenomenon for hundreds of years [7]. Recent research using functional Magnetic Resonance Imaging (fMRI) on lucid dreaming subjects has compared the activity of the motor cortex while awake and its activity while subjects are in a lucid dream: the subjects clenched their fists while awake, then imagined clenching their fists (again while awake) and then reported clenching their fists while in a lucid dream [8]. The researchers found significant correlation in brain activity between the three scenarios.
Figure 3 illustrates the conclusions drawn by the researchers in this study: the neural regions involved in actually clenching the fist and the neural regions associated with dreaming about clenching the fist are the same. The main difference in the brain between dreaming and waking states is the total area of neural regions involved and the magnitude with which those neurons are acting. This work in oneirology is revolutionary because dreams are often thought to be unpredictable events with unknowable contents until the subject is awoken and interviewed, and yet here the experimenters could confidently state they knew the actions of their subject while they were dreaming.
This research is unique because of its methods: the use of a subject with the ability to lucid dream at will and the use of tools that allow for high resolution images of the brain. It ultimately seems as though this will be the standard method of inquiry for the neural engineers that wish to understand how dreaming works. Lucid dreaming is a technique that a growing number of people are using to more fully engage with their dreams. As the regions of the brain responsible for lucid dreaming is mapped, and those brain regions are mathematically modeled, engineers will be able to develop technology that facilitates its occurrence.
A new field of product design and engineering is emerging as a result of overlap in public interest between lucid dreaming and wearable technology. While this market is relatively new, products related to recording dream content and documenting sleep patterns have been supported by tens of thousands through the crowd-sourcing site Kickstarter, indicating a demand in the market for more information about the way we sleep and dream. Among these products are apps like alarm clocks that help someone to remember and record their dreams and sleeping masks that flash lights onto the eyes to help achieve a lucid dream.
Collecting data on dreaming across populations and across the life of a person is an obvious and necessary first step in understanding the environmental conditions from which dreams emerge, as this information will be key in the design of engineered dreaming tools. But once these products start to bring data back to the scientists and engineers who designed them, a standard practice of dream data analysis must be constructed to the uncover the patterns that emerge.
Because engineering has its foundations in science, it must analyze data as objectively as possible. This requirement seems obvious, but it is a point of contention in the field of dream research because of the heavy influence of psychoanalysis on dream interpretation. According to modern research in cognitive psychology, dreams do not seem to be intensely symbolic or thematic, nor do they seem to be the by-product of random brain stem activity at night, as some scientists have argued [9]. Instead, evidence from the content analysis of thousands of dream reports has shown that most dreams are rather mundane reproductions of the daily life of a dreamer. All of this prior research is being considered as the current wave of dream technology products come to market. These devices will likely stream data collected from the users to a centralized source that will enable dream researches to study patterns of dreaming, much like the way social media data is analyzed for advertisement purposes. With all of this data from so many users, engineers and scientists will be equipped to drive further research and development in the creation of consumer dream technology.
The largest existing online dream report database is the website dreambank.net which was constructed by researchers at the University of California Santa Cruz. This database holds a library with over 20,000 dream reports from volunteers between the ages of 7 to 74 with some of the reports spanning several decades across the life of a subject. This is a platform anyone can browse but it is mostly used within a small academic circle of researchers. Hopefully, the data collected from the new consumer devices that are beginning to hit the market will allow for the construction of a new database, with reports from people worldwide that choose to volunteer their dreams for research. As more data is aggregated across larger groups of people over longer periods of time, anthropological studies of dreaming, especially as it relates to community and storytelling, will continue to develop, deepening the emerging field of the ‘digital humanities’. This field uses computational tools from the sciences and engineering to gain new perspectives on social phenomena. And although dreams are among our most personal experiences, they transform into stories once shared with other people. By analyzing the way people describe, respond, and reflect on their dreams, researchers will get a macroscopic view of how the process of dreaming blossoms from an isolated event in our brains into a social artifact in and of itself.
The Future of Dream Research
As neuroscientists map the regions of the brain responsible for dreaming and mathematicians construct the tools necessary to model the way these regions work, engineers have an opportunity to interface with the dreaming brain in unprecedented ways. Currently the extent of engineering projects that integrate dreaming are limited to gadgets that flash lights onto your eyelids or remind you to record your dreams when you wake up. Modern medical imaging tools and clearer understanding of the brain are enabling technology to take on a new set of engineering challenges. By using images of a brain actively dreaming and feeding this data into a working model of the brain’s visual and imaginative systems, it is theoretically possible to stream the visual content of a subject’s dream onto a screen.
Brain decoding is the science of interpreting signals within the brain in order to reconstruct the sensory experiences had by a subject [11]. This research is happening for the design and construction of effective brain-computer interfaces for patients who lack limbs or the ability to move. Moreover, understanding the encoding/decoding methods of the brain inspire engineers in the design of new robotics and AI [12]. The additional effects of this research are not difficult to imagine when considering the lab of Dr. Jack Gallant at UC Berkeley. Gallant’s team has been able to reconstruct the visual experience of their research subjects using imaging data from fMRI.
Two things are worth note: the reconstructed “movies” are very fuzzy, but they are astonishingly similar to the video clips the subjects were made to watch. And second, the data used to make this reconstruction represents blood flow through the brain [12]. This blood flow was the input, the fuzzy videos are the output. They accomplished this by building a digital model of what they think is going on in a part of the visual cortex in the brain. In effect, Gallant and his team have made a crude mind-reading device
While this work might evoke concerns about privacy that resonate with current news, it is astonishing to consider the possibilities, particularly as it relates to the field of oneirology. Another lab in the field of brain decoding doing happens at USC, in the lab of Ted Berger. Berger is part of a team that has managed to construct a cortical prosthesis that enables patients who have suffered from strokes that cause amnesia to create new long-term memories [13]. This research is significant in that it does more than just decode signals from the brain; it actually encodes signals back into it. Using microelectrodes that sense and conduct current from brain tissue, the prosthesis is a small chip with an onboard computational model of a portion of the hippocampus. While this model may not exactly represent the structure or function of its biological counterpart, it works well enough to enable former amnesiacs to form new long-term memories.
Fundamental neuroscience research has revisited old data in light of new insights, some coming from this work in decoding. Of interest to those studying dreaming, a new network involved in daydreaming and idle thinking has been uncovered. The regions involved in this network are being probed to determine their function while a subject is feeling, learning, or solving problems. This substrate has been termed the Default Mode Network (DMN) and there is evidence accruing that many parts of this network underlie dreaming itself [14]. While the DMN is mostly being studied in subjects who are awake, recent research has shown that there is extensive overlap between the DMN and the dreaming network. This overlap articulates the map of the brain and its functions that are relevant to dreams and imagination.
The future of psychology and neuroscience is truly indeterminable when you consider the gains to be made from the design of a device that can read the most private experiences of the human mind. Science, medicine, and education are just a few of the fields that will be impacted by the ongoing revolution in neuroscience, particularly as that revolution relates to dreams. And as neuroscience uncovers the means by which the brain gives rise to conscious experience, engineers will have an opportunity to interface and design for the mind in direct ways. Prosthetics that feel like limbs, remote controlled bodies that enable users to feel as though they are present in far off places, and shared dream experiences are all subjects of countless science-fiction stories, and yet when one considers the current successes of neural engineering they don’t seem as far-fetched as they used to. Billions of dollars in funding from the United States and the European Union are making their way into these kinds of projects due to recently announced initiatives to map the structures and functions of the brain. All of these conditions leave the next generation of neural engineers well equipped to tackle the problems that have faced humanity for generations, like the question of what happens when we dream.
2017 Addendum:
Since the writing of this paper a great deal has occurred. One of the greatest shake-ups in the history of oneirology was published in May of 2014, soon after I had written this paper. Ursula Voss and others published that they had been able to induce lucid dreams by stimulating the dorsolateral Prefrontal Cortex (dlPFC) with transCranial Alternating Current stimulation (tACS) set at 40 Hz [15]. This signal apparently coaxes the dlPFC into something akin to an awakened awareness while the subject is still asleep. Soon after, Martin Dresler (from the "clenched-fist" research) built off of Voss's work to map and compare the correlates of dreaming to the correlates of psychosis[16]. I believe this is the way dream researchers will begin to pull significant funding into the discipline. Following Voss's discovery, a Kickstarter project was started for the consumer-grade Lucid Dreamer device, based on her 40 Hz research. The Kickstarter was funded, but ultimately cancelled, as the researchers felt they were not getting consistent enough results. The project/company still exists however and you can follow them at http://www.luciddreamer.com/.Flash forward to 2016 - I sit down with Domhoff for the first time, after years of email correspondence. He tells me he has kept pushing on his cognitive approach to dreaming and has continued to connect his work with research in the Default Mode Network (DMN) and feels there is a trove of theory yet to be discovered lying in this neural system. Then, in March 2017, a new paper comes out of the lab of renowned consciousness/dream researcher Giulio Tononi, claiming to have not only uncovered the neural correlates of dreaming, but to connect those correlates to activity in the visual processing systems of the brain[17]. In his paper he states that he was able to roughly predict whether the dreaming subjects experienced "seeing a face" or "being inside vs. outside". He builds off of research that has compared the brain states of subjects who are 'perceiving vs. thinking'. Tononi builds off of some of Voss's work, but he doesn't use her technique of tACS to induce lucid dreaming in the subjects of his experiment.When I shared Tononi's paper with Domhoff, he was surprised there was not mention of the DMN or the Cognitive theory of dreaming. There are different schools of thought within dream research and sometimes things can be a bit political between them (i.e. Freud vs. Hobson vs. Domhoff). Perhaps this was a factor in Tononi's investigation.It is worth noting that Jack Gallant's lab has been able to reconstruct semantic thought on what a subject is seeing [18], and Ted Berger and his lab are now part of Kernal, a company in direct competition with Elon Musk's neurotechnology effort, Neuralink[19].Domhoff thinks the immediate future of dream research might be realized with functional Near-Infrared Spectroscopic imaging (fNIRS) which can be done affordably, portably, and with relatively high accuracy. Facebook recently revealed at their 2017 F8 conference that it is looking into this technology for brain-computer interface technology[20][21]. So between fNIRS, maps of the dreaming brain, and the tools used to collect dream reports upon awakening (think Siri/Alexa), we may be moving from the study of dreams (oneirology) to the engineering of dreams (oneirothetics). P.S. If you have any dreams you'd like to share, or if you'd like to contribute to dream research, please reach out to the team at http://www.dreambank.net/
Bibliography
[1] Philipp G. Sa¨mann, Renate Wehrle, David Hoehn, Victor I. Spoormaker, Henning Peters, Carolin Tully, Florian Holsboer and Michael Czisch, “Development of the Brain’s Default Mode Network from Wakefulness to Slow Wave Sleep,” Cerebral Cortex, vol. 21, no. 9, pp. 2082–2093, Sept. 2011
[2] Donna Rose Addisa, Ling Panb, Mai-Anh Vub, Noa Laiserb, Daniel L. Schacter, “Constructive episodic simulation of the future and the past: Distinct subsystems of a core brain network mediate imagining and remembering,” Neuropsychologia, vol. 47, pp. 2222–2238, Sept. 2008
[3] Donna Rose Addis, Katie Knapp, Reece P. Roberts, Daniel L. Schacter, “Routes to the past: Neural substrates of direct and generative autobiographical memory retrieval,” NeuroImage, vol. 59, pp. 2908–2922, Oct. 2011
[4] Draulio B. de Araujo, Sidarta Ribeiro, Guillermo A. Cecchi, Fabiana M. Carvalho, Tiago A. Sanchez, Joel P. Pinto, Bruno S. de Martinis, Jose A. Crippa, Jaime E.C. Hallak, Antonio C. Santos, Schacter, “Seeing With the Eyes Shut: Neural Basis of
Enhanced Imagery Following Ayahuasca Ingestion,” Human Brain Mapping, vol. 33, pp. 2550–2560, Sept. 2011
[5] Michel Jouvet, “Biogenic Amines and the States of Sleep,” Science, vol. 163, no. 3862 pp. 32–41, Jan. 3 1969
[6] E Aserinsky, N Kleitman, “Regularly occurring periods of eye motility and concomitant phenomena, during sleep”. Science vol. 118, no. 3062, pp. 273–274, Sept. 1953
[7] J. Allan Hobson, Dreaming: An Introduction to the Science of Sleep. Oxford: Oxford UP, 2002.
[8] David Foulkes, Dreaming: A Cognitive-psychological Analysis. Hillsdale, NJ: L. Erlbaum Associates, 1985.
[9] G. William Domhoff, The Scientific Study of Dreams: Neural Networks, Cognitive Development, and Content Analysis. Washington, DC: American Psychological Association, 2003.
[10] Deirdre Barrett, and Patrick McNamara. The New Science of Dreaming. Westport, CT: Praeger, 2007.
[11] Martin Dresler, Stefan P. Koch, Renate Wehrle, Victor I. Spoormaker, Florian Holsboer, Axel Steiger, Philipp G. Sämann, Hellmuth Obrig, and Michael Czisch. “Dreamed Movement Elicits Activation in the Sensorimotor Cortex.” Current Biology 21.21 (2011): 1833–837. Print.
[11] Shinji Nishimoto, An T. Vu, Thomas Naselaris, Yuval Benjamini, Bin Yu, and Jack L. Gallant. “Reconstructing Visual Experiences from Brain Activity Evoked by Natural Movies.” Current Biology 21.19 (2011): 1641–646.
[12] Doree Armstrong, and Michelle Ma. “Researcher Controls Colleague’s Motions in 1st Human Brain-to-brain Interface.” UW Today. N.p., 27 Aug. 2013. Web. 13 Feb. 2014.
[13] T. Horikawa, M. Tamaki, Y. Miyawaki, and Y. Kamitani. “Neural Decoding of Visual Imagery During Sleep.” Science 340.6132 (2013): 639–42. Print.
[14] Theodore W. Berger, Robert E. Hampson, Dong Song, Anushka Goonawardena, Vasilis Z. Marmarelis, and Sam A. Deadwyler. “A Cortical Neural Prosthesis for Restoring and Enhancing Memory.” Journal of Neural Engineering 8.4 (2011): 046017. Print.
[15]Ursula Voss, Romain Holzmann, Allan Hobson, Walter Paulus, Judith Koppehele-Gossel, Ansgar Klimke & Michael A Nitsche. “Induction of self awareness in dreams through frontal low current stimulation of gamma activity.” Nature Neuroscience 17, 810–812 (2014) doi:10.1038/nn.3719. Print.
[16]Martin Dresler, Renate Wehrle, Victor I. Spoormaker, Axel Steiger, Florian Holsboer, Michael Czisch, J. Allan Hobson. “Neural correlates of insight in dreaming and psychosis.” Sleep Medicine Reviews 20 (2015) 92e99. Print.
[17]Francesca Siclari, Benjamin Baird, Lampros Perogamvros, Giulio Bernardi, Joshua J LaRocque, Brady Riedner, Melanie Boly, Bradley R Postle, & Giulio Tononi. “The neural correlates of dreaming”. Nature Neuroscience, 10 April 2017, doi:10.1038/nn.4545. Published online.
[18]Alexander G. Huth, Tyler Lee, Shinji Nishimoto, Natalia Y. Bilenko, An T. Vu, and Jack L. Gallant. “Decoding the Semantic Content of Natural Movies from Human Brain Activity”. Front. Syst. Neurosci., 07 October 2016 | https://doi.org/10.3389/fnsys.2016.00081. Online.
[20]http://www.kurzweilai.net/what-if-you-could-type-directly-from-your-brain-at-100-words-per-minute
[21]https://developers.facebook.com/videos/f8-2017/f8-2017-keynote-day-2/