1. Neuroscience
Download icon

Functional MRI: The spinal cord is never at rest

  1. Falk Eippert  Is a corresponding author
  2. Irene Tracey  Is a corresponding author
  1. University of Oxford, United Kingdom
  • Cited 7
  • Views 1,302
  • Annotations
Cite this article as: eLife 2014;3:e03811 doi: 10.7554/eLife.03811


Even when we are at rest, our spinal cords show spontaneous, yet well organised, fluctuations of activity that might reflect sensory and motor networks.

Main text

Even when we are resting, our brain is far from inactive and instead displays an enormous amount of spontaneous activity. This activity is not just noise, but is organised in both space and time, with different brain networks showing distinct fluctuations. Moreover, these ‘resting-state signals’ account for most of the energy consumed by the brain.

Resting-state signals are organised into several distinct networks that can be modified by experience and can also change in response to numerous diseases, but it is still unclear what causes these signals (Fox and Raichle, 2007; Buckner et al., 2013). Now, in eLife, Robert Barry of Vanderbilt University and co-workers—including Seth Smith, Adrienne Dula and John Gore—report that such resting-state signals can also be seen in the human spinal cord (Barry et al., 2014). Thus, these signals seem to be a hallmark of the entire central nervous system.

The spinal cord is the brain's main interface with the rest of the body. The centre of the spinal cord is made up of four ‘horns’. The two dorsal horns at the back of the spinal cord contain the neurons that receive sensory information from the body; and the two ventral horns at the front contain the neurons that send signals to the muscles. The Vanderbilt team recorded activity from the human spinal cord using a technique called fMRI (short for functional magnetic resonance imaging). fMRI can measure neuronal activity indirectly by detecting changes in blood flow: when an area of the brain is in use, blood flow to that region increases (Logothetis, 2008). It is also non-invasive as it only involves positioning a person inside an MRI scanner.

Performing fMRI of the spinal cord, however, is technically challenging for a number of reasons. In particular, the spinal cord is very small (around 12 mm in diameter; Figure 1A), and there are also various sources of noise that degrade the quality of the signal to a much greater degree than happens in brain imaging (Summers et al., 2014). Correspondingly, the first reports of spinal fMRI only occurred in the late 1990s (Yoshizawa et al., 1996; Stroman et al., 1999) and relatively few groups are currently undertaking such studies.

Resting-state signals in the human spinal cord.

(A) Horizontal section of a brain (top) and a spinal cord (middle, bottom); the small size of the spinal cord makes it difficult to image neuronal activity. The spinal cord contains two ventral horns (one outlined in red) that are involved in motor function, and two dorsal horns (one outlined in green) that are involved in sensory function. (B) Barry et al. measured the correlation between spontaneous fluctuations in the fMRI signal in the ventral horns (red traces; top) and the dorsal horns (green traces; bottom). This revealed that the ventral horns show a positive correlation with each other, as do the dorsal horns. However, there is no significant correlation between ventral and dorsal horns. This suggests that at rest, the spinal cord is intrinsically organised into two separate networks, corresponding to motor and sensory functions. (C) Possible mechanisms that could explain the spontaneous activity in the spinal cord include input from the peripheral nervous system (top), locally generated rhythms from the interneurons within spinal networks (middle), and ongoing communication between the brain and spinal cord (bottom).

Barry et al. took the level of technical difficulty one step further by imaging the spinal cord at a very high magnetic field strength. Although this exacerbates many of the problems in spinal fMRI (problems that were overcome by developing tailored data acquisition and analysis techniques), it also provides a unique opportunity to look at the spinal cord in very fine detail and with a high signal-to-noise ratio. Barry et al. acquired fMRI data from the spinal cords of a group of healthy volunteers, who were simply resting inside the MRI scanner. Next, they extracted the fMRI signal over time from both ventral horns and both dorsal horns, and investigated the correlation between the activity in each of these four regions over time (Figure 1B).

Barry et al. found highly significant positive correlations between the left and right ventral (‘motor’) horns, and also between the left and right dorsal (‘sensory’) horns, suggesting that these structures form two distinct networks. This is obviously in good agreement with the functional organisation of the spinal cord. It also, to some degree, reflects the patterns of spontaneous fluctuations in the brain, where distinct resting-state networks (e.g., visual, auditory, etc) are observed (Fox and Raichle, 2007).

Barry et al. did not observe significant correlations in activity between the ventral and dorsal horns, suggesting that although the connections between these structures are essential for reflex responses, they might otherwise be dormant. However, it remains to be tested if the correlations in the spontaneous fluctuations will continue when an individual is performing a task or whether different patterns will emerge (Cole et al., 2014).

It is natural to assume that the observed resting-state correlations of fMRI signals reflect communication or connections between neurons. However, a variety of other mechanisms that are not related to neuronal activity are known to contribute to resting-state fMRI signals, including breathing-related changes and pulsations that are caused by each heartbeat (Birn, 2012). Barry et al. used several different techniques to address these issues, which strengthens the idea that their results do indeed reflect the activity of the spinal cord at rest.

So, what is the mechanism behind the observed spontaneous fluctuations in the spinal cord? There are several possible answers to this question, which are not mutually exclusive (Figure 1C). First, these fluctuations could be driven by ongoing input from the peripheral nervous system relaying information to the central nervous system, for example, about the relative position of body parts. Second, they could reflect signals that are generated locally within the spinal cord, including those created by the central pattern generators that control activities such as walking (Grillner and Jessell, 2009). Third, the fluctuations could reflect ongoing sensory and motor signals travelling up and down between the brain and spinal cord. Most likely, a mixture of these (and possibly other) mechanisms is involved, and it will require carefully designed experiments to disentangle these mechanisms.

In any case, the possibility to non-invasively investigate if the spinal cord's sensory and motor circuits are functioning properly is of enormous clinical interest. Many neurological disorders involve the spinal cord and the most obvious example is spinal cord injury. In these instances, the method developed by the Vanderbilt team could complement approaches that are focussed on structural changes (Freund et al., 2013). Yet, for this to happen we first need to answer several other questions: what is the organisation of connectivity across the different segments of the spine? Are the reported networks stable over time or do they change? Also, how do the different networks (sensory and motor) interact with each other? Finally, it would be interesting to record such resting-state signals from the brain and spinal cord at the same time (Finsterbusch et al., 2013) in order to obtain an integrative picture of the function of the central nervous system.


  1. Book
    1. Summers PE
    2. Brooks JCW
    3. Cohen-Adad J
    Spinal cord fMRI
    In: Cohen-Adad J, Wheeler-Kingshott CA, editors. Quantitative MRI of the spinal cord. Academic Press. pp. 221–239.

Article and author information

Author details

  1. Falk Eippert

    Oxford Centre for Functional Magnetic Resonance Imaging of the Brain, Nuffield Department of Clinical Neurosciences, University of Oxford, Oxford, United Kingdom
    For correspondence
    Competing interests
    The authors declare that no competing interests exist.
  2. Irene Tracey

    Oxford Centre for Functional Magnetic Resonance Imaging of the Brain, Nuffield Department of Clinical Neurosciences and Nuffield Division of Anaesthetics, University of Oxford, Oxford, United Kingdom
    For correspondence
    Competing interests
    The authors declare that no competing interests exist.

Publication history

  1. Version of Record published: August 5, 2014 (version 1)


© 2014, Eippert and Tracey

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.


  • 1,302
    Page views
  • 73
  • 7

Article citation count generated by polling the highest count across the following sources: Crossref, Scopus, PubMed Central.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Cell Biology
    2. Neuroscience
    Domenica Ippolito et al.
    Research Article Updated

    Sensory and behavioral plasticity are essential for animals to thrive in changing environments. As key effectors of intracellular calcium signaling, Ca2+/calmodulin-dependent protein kinases (CaMKs) can bridge neural activation with the many regulatory processes needed to orchestrate sensory adaptation, including by relaying signals to the nucleus. Here, we elucidate the molecular mechanism controlling the cell activation-dependent nuclear translocation of CMK-1, the Caenorhabditis elegans ortholog of mammalian CaMKI/IV, in thermosensory neurons in vivo. We show that an intracellular Ca2+ concentration elevation is necessary and sufficient to favor CMK-1 nuclear import. The binding of Ca2+/CaM to CMK-1 increases its affinity for IMA-3 importin, causing a redistribution with a relatively slow kinetics, matching the timescale of sensory adaptation. Furthermore, we show that this mechanism enables the encoding of opposite nuclear signals in neuron types with opposite calcium-responses and that it is essential for experience-dependent behavioral plasticity and gene transcription control in vivo. Since CaMKI/IV are conserved regulators of adaptable behaviors, similar mechanisms could exist in other organisms and for other sensory modalities.

    1. Evolutionary Biology
    2. Neuroscience
    Jan Clemens et al.
    Research Article Updated

    How neural networks evolved to generate the diversity of species-specific communication signals is unknown. For receivers of the signals, one hypothesis is that novel recognition phenotypes arise from parameter variation in computationally flexible feature detection networks. We test this hypothesis in crickets, where males generate and females recognize the mating songs with a species-specific pulse pattern, by investigating whether the song recognition network in the cricket brain has the computational flexibility to recognize different temporal features. Using electrophysiological recordings from the network that recognizes crucial properties of the pulse pattern on the short timescale in the cricket Gryllus bimaculatus, we built a computational model that reproduces the neuronal and behavioral tuning of that species. An analysis of the model’s parameter space reveals that the network can provide all recognition phenotypes for pulse duration and pause known in crickets and even other insects. Phenotypic diversity in the model is consistent with known preference types in crickets and other insects, and arises from computations that likely evolved to increase energy efficiency and robustness of pattern recognition. The model’s parameter to phenotype mapping is degenerate – different network parameters can create similar changes in the phenotype – which likely supports evolutionary plasticity. Our study suggests that computationally flexible networks underlie the diverse pattern recognition phenotypes, and we reveal network properties that constrain and support behavioral diversity.