I will summarise next the initial studies of spatio-temporal patterns of activity of the nervous system in vivo in experimental animals and human subjects. These digital graphical representations as spatio-temporal maps are beginning to reveal the neural bases of different internal states.
Specific spatio-temporal patterns of neural activity have been observed in the living cortex by early investigators. For example, as early as 1934 Adrian and Matthews working on cats and rabbits, reported that neurons pulsate in small groups, and as they pulsate out of phase with each other, periodic waves of activity may spread over the whole cortex. From the 1930s to the 1950s, many more papers mentioned brain waves in humans that appeared to travel across the cortex1.
The idea of travelling brain waves goes back to Moshe Abeles, and his idea of synfire chains which describe waves of activity involved in ‘binding’ the different functional states across the cortex. More recent findings indicate that precise timing of discharges can be preserved across numerous synaptic transmission steps.
Among the first evidence that brain oscillations in vivo behave as concentric, spiral or travelling waves of activity was obtained by Prechtl et al (2000)2 in the intact turtle cortex in response to visual stimuli. The authors used optical recording of electrical activity and detected travelling waves generated by pacemaker neuron that emit periodic pulses of excitation which then propagated across the neuronal cortical tissue. Naturally these experiments could not determine if the turtle was in a conscious state.
Recording of brain waves represents a huge field in clinical medicine where they provide a window into the activity of the normal and diseased brain. Brain waves in intact brains “propagate not only locally but also through the long-range white-matter thalamo-cortical connections, and thus the ‘diffusion’ process occurs along the particular connectivity between different parts of the brain a subject of studies of the human connectome”3.
Even emotional states have been associated with particular brain wave frequencies. For example, the θ frequency band is essential to identify the nature of the emotion. The energy of the β band has better predictability for negative emotions, including fear and sadness; the α frequency band has a good prediction for happiness while high-amplitude, regular α oscillation recorded from the occipital cortex represents relaxed wakefulness (resting condition). High cognitive load is represented by prolonged α oscillations in the frontal cortex. 4
To construct a spatio-temporal map that accurately reveals patterns of brain activity, the first challenge is to record simultaneously the activity of large assemblies of neurons over large areas and across multiple levels of the internal loops with sufficient spatial and temporal resolution.
Only partial advances have been achieved in this task. Oscillating neuronal activity in the brain has been described at various spatiotemporal scales on the basis of several electrophysiological methods, including electroencephalography (EEG), electrocorticography (ECoG, a form of intracranial EEG), local field potentials (LFP), as well as multi-unit and single-unit recordings.
Recording of brain activity in conscious subjects is mostly limited to a few non-invasive methods. Non-invasive techniques that directly measure electrical activity, such as EEG and magnetoencephalography (MEG), are very good at pinpointing the timing of neuronal firing, but much worse when it comes to spatial resolution.
Conversely, recording changes in blood flow associated with neural activity (BOLD technique) is a method that allows a good spatial resolution. However, as it requires acquisition times of several seconds, it is far slower than the millisecond timescale of neuronal signals and lacks sufficient temporal resolution. Images acquired with this method (via fMRI) will often show an entire neural pathway active all at once, when really, there’s a neural signal propagating from one part of the pathway to the next. Like fMRI, functional near-infrared spectroscopy (fNIRS) tracks changes in oxygen levels in blood flow. In experimental animals, propagation of waves of neural activity has been recorded using multi-electrode arrays and voltage-sensitive dyes, within the visual, somatosensory and sensorimotor, auditory, and motor cortices as well as hippocampus5.
Recent findings suggest the possibility of a distinct infra-slow activity process that moves dynamically through the brain to establish a systems-level organisation that is captured in the resting-state BOLD signal6.
Estimates of large-scale cortical networks have been assessed on the bases of fcMRI (intrinsic functional connectivity MRI) and generally confirmed by a variety of analytic approaches, including seed-based fcMRI7, independent component analysis8, cluster analysis9, and graph theory10. The discovery that strong cross-frequency coupling exists in multiple brain areas, including the neocortex, hippocampus and basal ganglia, suggests that cross-frequency coupling reflects functional activation of these areas11.
A complex but promising strategy to identify synchronous patterns of activity in the multidimensional network of the internal neural loops has been proposed by Atasoy et al12. They “demonstrate that a ubiquitous mathematical framework, eigendecomposition of the Laplace operator, which lies at the heart of theories of heat, light, sound, electricity, magnetism, gravitation and fluid mechanics, can predict the collective dynamics of human cortical activity at the macroscopic scale”. Further, they “demonstrate a plausible biological mechanism behind the emergence of these patterns from the cortico-cortical and thalamo-cortical interactions by modelling the excitatory and inhibitory dynamics with a neural field model”.
More invasive methods include the use of two-photon microscopy to visualise neurons that express a genetically encoded calcium indicator (eg GCaMP6f)13 and using multiple beamlets that scan the sample simultaneously and leveraging advanced computational methods to interpret the data14. Using this strategy, a recent investigation showed that is possible to record accurate 3D spatio-temporal patterns of cortical activity in mice during locomotion and to correlate these patterns with the dynamic features of the actual movements15.
New methods are beginning to bridge the gap between optimising spatial or temporal discrimination of brain activity. For example, a recent NMR method can visualise highly localised neural activity at millisecond timescales.
These methodological advances will no doubt accelerate the development of methods to record suitable spatio-temporal maps in intact brains.
- For references and discussion, see JR Hughes (1995): The phenomenon of travelling waves: a review. Clinical Electroencephalography 26, 1–6. ↩︎
- JC Prechtl et al (2000): Direct evidence for local oscillatory current sources and intracortical phase gradients in turtle visual cortex. Proceedings of the National Academy of Sciences USA 97, 877–882. ↩︎
- S Atasoy et al (2016): Human brain networks function in connectome-specific harmonic waves. Nature Communications 7, 10340. ↩︎
- E Başar (2013): Brain oscillations in neuropsychiatric disease. Dialogues in Clinical Neuroscience 15, 291-300. ↩︎
- JN MacClean & NG Hatsopoulos (2019): Coding in large-scale cortical populations. In The Neocortex (eds: W Singer et al), Chapter 12, MIT Press. ↩︎
- ME Raichle et al (2019): Brain Networks – How many types are there? In The Neocortex (eds: W Singer et al), Chapter 6, MIT Press. ↩︎
- B Biswal et al (1995): Functional connectivity in the motor cortex of resting human brain using echo-planar MRI. Magnetic Resonance in Medicine 34, 537-541. ↩︎
- CF Beckmann & SM Smith (2004): Probabilistic independent component analysis for functional magnetic resonance imaging. IEEE Transactions on Medical Imaging 23, 137-152. ↩︎
- P Bellec et al (2010): Multi-level bootstrap analysis of stable clusters in resting-state fMRI. Neuroimage 51, 1126–1139. ↩︎
- NUF Dosenbach et al (2007): Distinct brain networks for adaptive and stable task control in humans. Proceedings of the National Academy of Sciences USA 104, 11073-11078. ↩︎
- RT Canolty & RT Knight (2010) The functional role of cross-frequency coupling. Trends in Cognitive Science 14, 506–515. ↩︎
- S Atasoy et al (2016): Human brain networks function in connectome-specific harmonic waves. Nature Communications 7, 10340. ↩︎
- T-W Chen at el (2013): Ultrasensitive fluorescent proteins for imaging neuronal activity. Nature 499, 295-300. ↩︎
- W Yang et al (2016) Simultaneous multi-plane imaging of neural circuits. Neuron 89, 269-284. ↩︎
- A Fedotova et al (2023): Dissociation between neuronal and astrocytic calcium activity in response to locomotion in mice. Function 4, zqad019. ↩︎