Ripples reflect a spectrum of synchronous spiking activity in human anterior temporal lobe

  1. Ai Phuong S Tong
  2. Alex P Vaz
  3. John H Wittig
  4. Sara K Inati
  5. Kareem A Zaghloul

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This study investigates the relationship between neuronal spiking and local field potential (LFP) activity recorded on micro-electrodes, and intracranial EEG activity recorded on macro-electrodes using a data from human patients with epilepsy. The dataset is rare and unique in that it is from the human brain and recording neural activity from multiple scales (single neuron, LFP, iEEG) within the same subjects. The study tackles important questions with regards to how ripples relate to broadband LFP activity as well as single neurons in the human brain. This results will be of interest to a broad audience of neuroscientists, and particularly to readers in the field of memory as well as those who perform electrophysiology.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Direct brain recordings have provided important insights into how high-frequency activity captured through intracranial EEG (iEEG) supports human memory retrieval. The extent to which such activity is comprised of transient fluctuations that reflect the dynamic coordination of underlying neurons, however, remains unclear. Here, we simultaneously record iEEG, local field potential (LFP), and single unit activity in the human temporal cortex. We demonstrate that fast oscillations within the previously identified 80–120 Hz ripple band contribute to broadband high-frequency activity in the human cortex. These ripple oscillations exhibit a spectrum of amplitudes and durations related to the amount of underlying neuronal spiking. Ripples in the macro-scale iEEG are related to the number and synchrony of ripples in the micro-scale LFP, which in turn are related to the synchrony of neuronal spiking. Our data suggest that neural activity in the human temporal lobe is organized into transient bouts of ripple oscillations that reflect underlying bursts of spiking activity.

Article activity feed

  1. Version published to 10.7554/elife.68401 on eLife
  2. eLife

    Author Response:

    Reviewer #1 (Public Review):

    [...] The study tackles important questions with regards to how ripples relate to broadband LFP activity as well as single neurons in the human brain. The authors also include elegant analyses to characterize the timing of spiking activity with respect to high and low frequency activity and relevant control analyses to take into account possible artifacts and epileptic-related activity. My main concern with the study is whether the authors are truly isolating ripple activity in the human brain as claimed. Their threshold for ripple activity is quite low and it thus seems very possible that many of their "ripple" events are rather high frequency activity events that reflect spiking activity. That being said I think this is an important study to share results from in that it provides unique characterization of the relationship between high frequency activity and spiking in the human brain, as well as how it relates to human memory.

    We thank the reviewer for the positive assessment of our manuscript. We agree with the reviewer that there are important concerns regarding whether these events can be truly regarded as ripples, which we address by performing additional analyses to provide stronger support to the possibility that the identified cortical ripples in our recordings are transient and discrete events that reflect underlying bursts of spiking activity. We also acknowledge that these events more likely exist on a continuum, and that there is not always a clear separation between what constitutes one ripple event and the background activity. We think this is a fertile area of investigation and that there are several points to consider regarding this question, which we now address in the revised Discussion.

    Reviewer #2 (Public Review):

    Recordings from human patients with implanted electrodes provide high temporal resolution, localized measurements of brain activity that can reveal neural correlates underlying a wide variety of cognitive functions. These intracranial electroencephalography (iEEG) recordings are typically made with large electrical contacts, however, and thus represent a complex and poorly understood averaging of voltages from the underlying tissue. As such, it is difficult to know exactly what patterns of neural activity these signals correspond to and how to compare them to spiking and local field potential (LFP) recordings more commonly acquired in non-human animals. Tong et al. carried out simultaneous iEEG recordings from surface contacts as well as spiking and LFP recordings from implanted electrode arrays to directly address the relationship between these signals.

    They present quantifications (e.g. Figure 2B) of the relationships between the amount of spiking activity and the amplitude of events detected in the LFP and iEEG. Their results showing the relationships among these signals are very important for the field, and it is very helpful to see that there are clear correlations across scales.

    We thank the reviewer for this positive review of our manuscript.

    The context in which they present these results is problematic, however. They focus on "ripple" events, detected as periods where the power in a 80-120 Hz band exceeds an arbitrary threshold for an arbitrary length of time. To be fair, the application of similarly arbitrary thresholds is common in the human, primate, and rodent literatures, and several important results have arisen from the analysis of these events. These results can be understood as claiming that a set of high amplitude events have certain properties (e.g. they are related to memory retrieval), but should not be understood as establishing that there is some specific threshold that separates real events from others.

    We completely agree with the reviewer that, when considering ripples, there is a continuity of activity and that the central challenge for the field has been how to identify real events and separate them from others. To be clear, the purpose of our manuscript is not to claim that such a threshold exists. Our purpose instead was to offer evidence that even events that fall below arbitrarily defined thresholds still reflect underlying bursts of spiking activity, that these events are still punctate and temporally discrete, and that therefore these events are also likely functionally meaningful.

    Here they go beyond these analyses and make the claim that these ripple events correspond to real, discrete events that, as their title indicates, "reflect a spectrum of synchronous spiking activity." The problem here is that they do not present any criteria for defining a real, discrete event. Indeed, they conclude that "the continuum of activity that [they] observe in [the] data ... suggests that strictly adhering to predefined criteria for what constitutes a ripple may run the risk of overlooking functionally meaningful events". Without a clear definition of what should and should not be considered to be a discrete event, we are left with the current situation where each study uses their own set of criteria, picks out a set of high amplitude events, and uses those for subsequent analyses.

    We agree with the reviewer that the amplitude and strength of the identified ripple events can be quite variable, making it challenging to distinguish these events from baseline activity. We therefore do not claim to identify specific criteria for defining real events. As the reviewer notes, without such criteria we are left with the current situation where future studies will still need to use their own set of criteria. We completely agree. The purpose of our study, however, is to just highlight the point that using these arbitrarily defined criteria risks overlooking these other events that still may be meaningful for the brain. We have addressed this concern by adding several changes to our manuscript and introducing several new analyses. First, we have tempered our claims that these are discrete events that represent separate packets of information. We acknowledge that there is certainly some variability in the size of these ripple events, that making a clean distinction between when these ripple events emerge as entities that are distinct from the background activity is challenging, and that we can never be certain whether arbitrarily small events are functionally meaningful. We have revised our Discussion accordingly to highlight these possibilities, and to discuss the larger point regarding the challenges in identifying these specific thresholds. Second, although we recognize that the data may not be absolutely conclusive, we have supplemented this discussion with additional analyses that, in our opinion, strongly suggest that these events are indeed transient in nature even when failing to meet previous thresholds.

    A second major challenge to understanding the current manuscript is the ambiguity of the physical relationships between the LFP and iEEG recording sites. While it might be obvious to human physiologists, details such as the distance between the LFP and iEEG contacts and the site areas of each type of electrode are critical for interpreting how closely the data from each could be expected to be related.

    We also agree with this very good point. As noted above in the response, we have now introduced several new analyses that examine the relation between the ripples identified using LFP and iEEG recordings.

    We would also like to highlight an instance of a common statistical error in Fig 1 I: the authors conclude that the difference between correct and incorrect is significant in true data and insignificant in the ripple-removed data, and therefore the 70-200Hz power band modulation on correct trials is significantly informed by 80-120Hz ripple events. The statistical problem is further described in Nieuwenhuis et al., Nature Neuroscience 2011.

    We thank the reviewer for pointing out this common statistical error that we have made in concluding that the true and ripple-removed data differ because the difference between correct and incorrect is significant in the true data and not in the ripple-removed data. We agree with the suggestion that a more accurate way to compare the true and ripple-removed data is to compare the effect sizes, or the difference between correct and incorrect trials for the true and ripple-removed data. We have now conducted this analysis. Specifically, we computed the true correlation between the difference in 70-200 Hz power between correct and incorrect trials and the difference in ripple rates between correct and incorrect trials across electrodes and compared this correlation to the correlation present after removing the 80-120 Hz ripples using a paired t-test across participants. We performed this analysis in the subset of six participants who had an MEA and focused on the MTL and ATL electrodes since these regions have the greatest 70-200 Hz power increase with successful retrieval. We found a significant decrease in correlation across patients with ripples removed compared to when we retained the ripples (t(5) = 3.89, p = 0.0115). We now report this new analysis in Fig. 1 – S6. We also compared the two correlations as dependent groups and found a significant difference in correlation (r_true – r_control = 0.172, 95% CI = [0.0691 0.2764], z = 3.2677, p = 0.0011). We accounted for potential interaction effects using the correlation between 70-200 Hz and 70-200 Hz with ripple removed (r = -0.031).

    Finally, we would also like to note the difficulty of characterizing a single deflection in the LFP or iEEG signal as a low frequency oscillation, given the large potential for measurement variability of the frequency of that oscillation as described in Fig. 4. This large deflection is to be expected when a concentrated amount of synaptic input drives a burst of spiking, as we would expect in the case of the increased spiking during ripples. In the hippocampus, this deflection is the sharp wave component of the sharp-wave ripple; it appears to take a similar form in cortical ripples. While unsurprising, it is well worth observing that the iEEG reflects this coincident deflection, but it should not be characterized as a 2-10Hz oscillation.

    We completely agree with the reviewer about this point. As the reviewer points out, the large deflection is often observed with bursts of spiking, which we find with ripples, and we also feel that the iEEG reflects this coincident detection. We have therefore corrected the text and no longer characterize the deflection associated with ripples as 2-10 Hz oscillations. To illustrate this point further, we have also now added a new analysis demonstrating the average iEEG and LFP activity around each ripple, which clearly demonstrates this deflection (Fig. 1). We have, however, retained the discussion about the locking of spikes to 2-10 Hz oscillation, as our analysis includes spikes both within and outside of the spike bursts. While we expect that spikes are associated with the large deflection that reflects a concentrated amount of synaptic input, we also find spikes that are modulated by a 2-10 Hz oscillation, consistent with prior findings of theta-phase locking of spiking neurons.

    Reviewer #3 (Public Review):

    In this study, authors systematically investigated the iEEG ripples, LFP ripples, and their relation with each other and single units from micro channels obtaining LFP. They found that the amplitude of LFP ripples reflects the sum and alignment of underlying spiking activities. Meanwhile, the amplitude of iEEG ripples reflects the number and alignment of LFP ripples. More interestingly, the amplitude of ripple events is functionally relevant. In general, I find that the data analyses and methods are sophisticated and the results are interesting. It extends our understanding of ripple events and is of interest to a wide audience.

    We thank the reviewer for the positive assessment of our manuscript.

  3. eLife

    Evaluation Summary:

    This study investigates the relationship between neuronal spiking and local field potential (LFP) activity recorded on micro-electrodes, and intracranial EEG activity recorded on macro-electrodes using a data from human patients with epilepsy. The dataset is rare and unique in that it is from the human brain and recording neural activity from multiple scales (single neuron, LFP, iEEG) within the same subjects. The study tackles important questions with regards to how ripples relate to broadband LFP activity as well as single neurons in the human brain. This results will be of interest to a broad audience of neuroscientists, and particularly to readers in the field of memory as well as those who perform electrophysiology.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

  4. eLife

    Reviewer #1 (Public Review):

    Tong and colleagues recorded intracranial EEG activity in 21 epilepsy patients to explore the relationship between high frequency LFP activity (high frequency activity - HFA), ripples and neuronal spiking activity during a verbal episodic memory task. Results show an association between ripple amplitude/rates, HFA activity, and successful memory (correct versus incorrect memory recall trials). The dataset is rare and unique in that it is from the human brain and recording neural activity from multiple scales (single neuron, LFP, iEEG) within the same subjects. The study tackles important questions with regards to how ripples relate to broadband LFP activity as well as single neurons in the human brain. The authors also include elegant analyses to characterize the timing of spiking activity with respect to high and low frequency activity and relevant control analyses to take into account possible artifacts and epileptic-related activity. My main concern with the study is whether the authors are truly isolating ripple activity in the human brain as claimed. Their threshold for ripple activity is quite low and it thus seems very possible that many of their "ripple" events are rather high frequency activity events that reflect spiking activity. That being said I think this is an important study to share results from in that it provides unique characterization of the relationship between high frequency activity and spiking in the human brain, as well as how it relates to human memory.

  5. eLife

    Reviewer #2 (Public Review):

    Recordings from human patients with implanted electrodes provide high temporal resolution, localized measurements of brain activity that can reveal neural correlates underlying a wide variety of cognitive functions. These intracranial electroencephalography (iEEG) recordings are typically made with large electrical contacts, however, and thus represent a complex and poorly understood averaging of voltages from the underlying tissue. As such, it is difficult to know exactly what patterns of neural activity these signals correspond to and how to compare them to spiking and local field potential (LFP) recordings more commonly acquired in non-human animals. Tong et al. carried out simultaneous iEEG recordings from surface contacts as well as spiking and LFP recordings from implanted electrode arrays to directly address the relationship between these signals.

    They present quantifications (e.g. Figure 2B) of the relationships between the amount of spiking activity and the amplitude of events detected in the LFP and iEEG. Their results showing the relationships among these signals are very important for the field, and it is very helpful to see that there are clear correlations across scales.

    The context in which they present these results is problematic, however. They focus on "ripple" events, detected as periods where the power in a 80-120 Hz band exceeds an arbitrary threshold for an arbitrary length of time. To be fair, the application of similarly arbitrary thresholds is common in the human, primate, and rodent literatures, and several important results have arisen from the analysis of these events. These results can be understood as claiming that a set of high amplitude events have certain properties (e.g. they are related to memory retrieval), but should not be understood as establishing that there is some specific threshold that separates real events from others.

    Here they go beyond these analyses and make the claim that these ripple events correspond to real, discrete events that, as their title indicates, "reflect a spectrum of synchronous spiking activity." The problem here is that they do not present any criteria for defining a real, discrete event. Indeed, they conclude that "the continuum of activity that [they] observe in [the] data ... suggests that strictly adhering to predefined criteria for what constitutes a ripple may run the risk of overlooking functionally meaningful events". Without a clear definition of what should and should not be considered to be a discrete event, we are left with the current situation where each study uses their own set of criteria, picks out a set of high amplitude events, and uses those for subsequent analyses.

    A second major challenge to understanding the current manuscript is the ambiguity of the physical relationships between the LFP and iEEG recording sites. While it might be obvious to human physiologists, details such as the distance between the LFP and iEEG contacts and the site areas of each type of electrode are critical for interpreting how closely the data from each could be expected to be related.

    We would also like to highlight an instance of a common statistical error in Fig 1 I: the authors conclude that the difference between correct and incorrect is significant in true data and insignificant in the ripple-removed data, and therefore the 70-200Hz power band modulation on correct trials is significantly informed by 80-120Hz ripple events. The statistical problem is further described in Nieuwenhuis et al., Nature Neuroscience 2011.

    Finally, we would also like to note the difficulty of characterizing a single deflection in the LFP or iEEG signal as a low frequency oscillation, given the large potential for measurement variability of the frequency of that oscillation as described in Fig. 4. This large deflection is to be expected when a concentrated amount of synaptic input drives a burst of spiking, as we would expect in the case of the increased spiking during ripples. In the hippocampus, this deflection is the sharp wave component of the sharp-wave ripple; it appears to take a similar form in cortical ripples. While unsurprising, it is well worth observing that the iEEG reflects this coincident deflection, but it should not be characterized as a 2-10Hz oscillation.

  6. eLife

    Reviewer #3 (Public Review):

    In this study, authors systematically investigated the iEEG ripples, LFP ripples, and their relation with each other and single units from micro channels obtaining LFP. They found that the amplitude of LFP ripples reflects the sum and alignment of underlying spiking activities. Meanwhile, the amplitude of iEEG ripples reflects the number and alignment of LFP ripples. More interestingly, the amplitude of ripple events is functionally relevant. In general, I find that the data analyses and methods are sophisticated and the results are interesting. It extends our understanding of ripple events and is of interest to a wide audience.

  7. Version published to 10.1101/2021.03.14.435276v2 on bioRxiv
  8. Version published to 10.1101/2021.03.14.435276v1 on bioRxiv