Can neuroimaging reveal the roots of psychiatric disorders?

Why the answer is almost certainly: No.

How many times have we read of groundbreaking discoveries that link mental health to neurobiological aspects of the brain? For example, we have been told so many times that neuroimaging has identified the alteration of a brain area with a psychiatric disorder X, or that the neurological effects of PTSD have been uncovered, or that whatever trauma has deleterious effects on brain area X,Y,Z, or that brain structural lesion X cause psychosis Y, or that specific abnormal brain structures have been discovered and are putatively responsible for bipolar, schizophrenia, autism, ADHD, anxiety, suicidal thoughts, whatever, etc., etc. … name it (let alone that gene a,b,c links mental disorder X,Y,Z, but this is another story.) All this supposedly reveals to us the "underlying mechanisms" of mental disorders, and neuroimaging for mental health will soon allow us to treat these disorders with "precision psychiatry." I guess I have read hundreds of such stories in the last decades.

Now, who knows me knows also that I'm not a psychiatrist. And probably this is why I don’t understand what these people believe to achieve. But I know how science is and how it is supposed to work. And this doesn’t look sound like science to me. At best,  wishful thinking or, worse, wild speculations and unwarranted extrapolations. I don’t see any rational and scientific reason to believe that, once we have identified a certain brain-structure alteration correlating with a mental phenomenon this could tell us much about the mechanisms causing it (I would even doubt it is causing it in the first place, but I won’t digress…)

Imagine extraterrestrials landing on Earth and finding a laptop, but who know nothing about how computers work (let us suppose humans are long extinct and nobody can tell them…) They turn the thing on and try to understand how it functions by correlating the pictures on the monitor (they presumably don’t understand either) with the heat flow they observe coming from the chips. Let us assume they can go a bit further, such as measuring the current in a dozen of transistors (over billions), analyzing its structure, and measuring where the electricity flows, but not much more than that. Do you believe they will be able to understand how a pc works, let alone retro-engineer the software running on the pc? Well, perhaps ETs are super-intelligent, and they might be able to do so, but for us, this would be an impossible task. Even knowing the exact physical structure of a CPU, is still a far cry from knowing what it is doing. The random flickering of zeros and ones going in and out of a CPU, tells nothing about the program producing it. And if you locate some chip areas on the motherboard where the electric activity correlates with, say a picture of a brain scan on the monitor, you might be lured into believing that precisely that location produces it and furnishes information about the “underlying mechanisms.” But this is obviously not the case. It wouldn’t be wise to jump to such conclusions.

Yet, this is essentially what we are trying to do with brain scans and mental health. With neuroimaging techniques, we can see the blood flow in the brain, while microscopy has revealed the neuron’s structure, but we are light years away from being able to see in real-time the activity of each of the 80 billion neurons. And even if we would be able to measure all things in such detail together, so what? There is always an intellectual instinct making us believe that once we know the map (that, for the time being, we still know only with a very low resolution) then we will know and understand the territory.

So, it should be of no surprise that it turns out how, after more than two decades of multi-billion-dollar research by the National Institutes of Mental Health to “locate the biomarkers of brain activity that point to the biological roots of a host of mental health diseases”, there is still no clear indication of how the two are connected. We are told that neuroimaging still has not brought benefits to the diagnosis and treatment of psychiatric disorders. Of course, we are also told that it is only a matter of time, more research and more funds and that then, finally, we will get there. I’m afraid, however, that once that time will be over, we will be told again that “things turned out to be more complicated than expected” and that another twenty years are needed.

Personally, I suspect that the problem has nothing to do with the research or the complexity of the brain, or the lack of funds. The goal can’t be achieved, not even in principle. No more and no less than you can read the line script of a complicated software program running on a computer by analyzing the thermal emission of its motherboard.

What we really need is not more brain scans, more chemistry, more reductionist molecular biology. What we really need is to recognize that the explanatory power of a science that reduces everything to elementary microscopic entities is limited. Probably much more limited than what we are actually willing to admit. If we want to understand the functions (or dysfunctions) of the human mind, we will have to accept that we will never be able to encapsulate it into reductionist theory and, thereby, will have to reconsider the good old approaches that originated psychology and psychiatry in the first place. We will have to relearn to get into contact with the human dimension of the human mind, its emotionality, and its soul through our own mind and consciousness, and by first-person methods of research that machines, maps, or microscopes will never be able to apply. The multidimensional complexity of the human being can’t be reduced to a 2D flatland theory. Humans will have to return to be human if they want to understand their humanity.