DORSAL COLUMN Book Review The neuronal representation of information in the human brain

The responses of single neurons provide evidence that is essential to understanding what information is encoded in a brain area, and how it is encoded, for the information conveyed by a single neuron is almost independent of that conveyed by others in the same population. The differences in the firing rate response profiles of single neurons to a set of stimuli or events encode most of the information about that stimulus or event. The representation is sparsely distributed, with neurons often having an approximately exponential firing rate distribution, with high firing rates to a few stimuli, and smaller and smaller responses to other stimuli in the set, and no response to most of the stimuli. It is in this way that information about particular stimuli is represented. For example, in the inferior temporal visual cortex, neurons encode in this manner which particular face or object is being shown (Rolls and Treves, 2011). Evidence of this type is crucial to understanding the operation of the brain, for by reading the code from different connected brain areas, it is possible to start to understand what is being performed (or computed) by each brain region. This in turn is important for understanding neurological and psychiatric disorders and their symptoms, and for suggesting treatments based on what each brain region is performing, and developing hypotheses about how each cortical area operates (Rolls, 2008, 2016). Much of the evidence on what is represented in different cortical areas comes from single neuron studies in macaques, including concepts about the roles of cortical areas that correspond to those in humans. The book Single Neuron Studies of the Human Brain Probing Cognition edited by Fried, Rutishauser, Cerf and Kreiman is therefore important, for it reviews what is now being discovered from recordings from single neurons in humans, performed as part of a programme of clinical treatment, and even a little evidence of this type may help to show how the detailed studies in the non-human primate cortex are providing evidence that is relevant to understanding the human brain, and potentially its disorders. The book has an important chapter by Adam Mamelak on the ethical and practical considerations for microelectrode recording studies in humans, pointing out that they may be useful, for example in assessing whether a brain area being considered for resection is operating pathologically or not, and for example in locating sites for possible deep brain stimulation by using the types of neuronal activity being recorded as markers. During the course of clinical investigations of this type some evidence about neuronal activity may become available, but of course there are many provisos that apply, such as that the brain region from which recordings are made may be dysfunctional, and that the time for recording from neurons is necessarily limited, so that the data obtained are likely to be limited. Given these constraints, what findings are being made from single neuron recordings in humans? Some of the most fascinating findings are about how neurons respond in medial temporal lobe regions. In Chapter 8 Mormann, Ison, Quiroga, Koch, Fried and Kreiman describe some neurons with responses that appear quite selective, with one neuron responding for example to Jennifer Aniston but much less to other individuals, and responding multimodally, for example not just to the sight of Jennifer Aniston, but also to the sound of her SINGLE NEURON STUDIES OF THE HUMAN BRAIN PROBING COGNITION Eds. I. Fried, U. Rutishauser, M. Cerf and G. Kreiman, 2014 The MIT Press: Cambridge Massachusetts and London, England. ISBN: 978-0-262-02720-5. Price: £41.95. doi:10.1093/brain/awv242 BRAIN 2015: 138; 3459–3462 | 3459