Imaging the brain at multiple scales: how to integrate multi...
From phrenology to connectomics: Your brain as an open book
Last year I came across an interesting newspaper article playfully titled “You are your connections” (van Hintum, 2011), referring to the immensely popular book by neurologist Dick Swaab called “You Are Your Brain” (2010). While the latter has been thoroughly criticized for its overly reductionist perspective by scientists and philosophers alike, the connectome-movement seems to provide an integrated view of the human brain, susceptible for environmental influences while adhering to the genetic bases as laid down in the organism’s genome (Seung, 2012). Connectomics focuses on connections between brain areas by investigating white matter-tracts that link various regions with each other. Looking at one voxel at the time, scientists from the Human Connectome Project try to reconstruct the neural and synaptic connections of the entire human brain, revealing a neural network that accounts for the immense complexity of the brain.
What interests me are the parallels between connectomics and the research we are currently doing at Amsterdam University College, where we adopt a multivariate-pattern analysis (MVPA) technique to predict what someone is looking at. Like connectomics, MVPA approaches the brain as an intricately interconnected machine, in which brain areas are in constant connection with each other and where anatomical modules are largely absent. Using this multivariate approach, we look at how activity of multiple brain areas together constitute a pattern that accounts for the mental activity someone experiences. As opposed to earlier techniques to interpret fMRI data which used univariate analyses, MVPA aims at decoding mental states as patterns of data generated by the interplay of multiple brain areas at the same time. The success stories of MVPA studies so far hints, in my opinion, at a paradigm shift in the neurosciences, which will emphasize the brain as a coherent machine that accounts for our infinite number mental experiences by the infinite number of combinations and permutations of the billions of neurons that our brain inhabit. What started as the ‘science’ of phrenology, which assigned specific brain areas to particular modular functions, is now becoming a science that truly recognizes the full capacity, complexity, and possibilities of the human brain.
Our research was intended to gather fMRI data from participants who watched short movie clips comprised of either of the following genres: action, comedy, drama, and non-fiction. The challenge consisted of selecting a meaningful set of brain areas to perform a classifier analysis –a computational algorithm designed to extract patterns from huge datasets– in order to predict which genre was being watched. After an intense quest of trying out countless combinations of brain areas, which was of course partly motivated by the available literature on the subject, our research group found an overall prediction rate of 69% –significantly better than the 25% chance level. In other words, merely computational analysis enables “mind reading”, albeit imperfect and assuming straightforward and discrete categories.
Groundbreaking research, right? Actually, not really. Theoretical suggestions on this matter of “mind-reading” have been proposed before. In the last decade, theory has given way for experimental studies confirming the possibilities to extract mental states from neural data (for example Naselaris et al., 2009; Poldrack et al., 2009; Kay & Gallant, 2009). Not only our private mental states are not longer impervious to the technology in the neurosciences, but also human decision-making may be predetermined, anchored in our biology, and contradicting years of philosophizing of our supposed free will (for a review see Mieras, 2012; Lamme, 2010). Although many disagree (see Gazzaniga, 2011), these experiments leave –in my opinion– little room for concepts such as free will, an immaterial psyche, and the privacy of mental states. Accompanied by the previously mentioned paradigm shift in methodology, the neurosciences may trump traditional philosophy by proving that our biology is the ultimate mediator of our behavior, personality, and decisions. Dutch cognitive neuroscientists Victor Lamme strikingly described this line of thought in his infamous book “Free Will Doesn’t Exist” (2010), proposing that ultimately every aspect of human behavior, including decision making, is inescapably subject to whatever our biology dictates us.
However, as opposed to the rather monomaniacal, reductionists views of for example Dick Swaab (2010), connectionism and Lamme’s theory on free will do not solely focus on the effect of genes or prenatal influences on development (as Swaab does); they acknowledge how experience can, literally, alter our brain and shape human behavior accordingly. Exactly this twist of the traditional reductionist perspective makes it such an elegant theory.
If this theory holds true, the future of the neurosciences is bright and exciting. If we are, indeed, our connections, we may be able to decode the brain and one’s mental experiences to perfection, flawlessly predicting what one thinks and feels. Connectionism may in the future provide the biological knowledge of neural networks that can be tested by means of multivariate pattern analyses, envisioning a thorough understanding of the human brain. Individual differences could be explained by unique experiences materialized as connections within the brain. Surely, there is a long way to go before we can even think of analyzing the brain at this scale, but methods as MVPA and connectionism pave the way for such ambitions.
It is only a matter of time before the human brain is as an open book to science.
References and further readings:
Gazzaniga, M. (2011). Who’s in Charge? Free Will and the Science of the Brain. New York: Harper
Kay, K., and Gallant, J. (2009). I can see what you see. Nature Neuroscience, 12 (3): 245-246.
Lamme, V. (2010). De vrije wil bestaat niet. Amsterdam: Uitgeverij Bert Bakker.
Mieras, M. (2012). ‘Spontaan’ besluit is voorbereid. De Volkskrant, 28 january 2012. Retrieved from
Naselaris, Th., Prenger, R., Kay, K., Oliver, M., and Gallant, J. (2009). Bayesian reconstruction of natural images from human brain activity. Neuron, 63: 902-915.
Poldrack, R., Halchenko, Y., and Hanson, S. (2009). Decoding the large-scale structure of brain function by classifying mental states across individuals. Psychological Science, 20 (11): 1364-1372.
Seugn, S. (2012). Connectome: How the brain’s wiring makes us who we are. Boston, MA: Houghton Mifflin Harcourt.
Van Hintum, M. (2011). Je bent je connecties. De Volkskrant, 7 may 2011. Retrieved from