Functional magnetic resonance imaging (fMRI) is a tool that can be used to non-invasively measure the activity of the human brain. Active parts of the brain require more oxygen, which increases blood flow to these areas. fMRI can detect these changes, and its signal reflects the coupling between brain activity and changes in blood flow.

The mechanism that couples brain activity to blood flow is known as the ‘hemodynamic response’, and its timing varies across the brain. Therefore, to interpret fMRI signals correctly and use them to measure underlying brain activity, it is necessary to understand how the response changes across the brain.

Current methods for probing hemodynamic response variation are either limited to specific brain regions or require patients to hold their breath – something not all groups of patients can do. To solve this problem, Bailes et al. investigated whether resting-state fMRI signals contain information about how the hemodynamic response changes across the brain. This information could then be used to better infer brain activity from fMRI measurements.

The experiments showed that resting-state fMRI signals can be used to characterize and predict the timing of the hemodynamic response. Specifically, the frequencies in resting-state fMRI signals are impacted by changes in the hemodynamic response and can therefore be used to predict hemodynamic timing. Additionally, Bailes et al. showed that these predictions are better than those obtained in experiments requiring patients to hold their breath, which is the current gold standard. The findings also demonstrate that the information from the frequencies of resting-state fMRI signals should be interpreted carefully, as differences in these frequencies can have a non-neural origin.

Bailes et al. propose a highly generalizable approach for mapping and predicting variations of the hemodynamic response across the whole brain. These findings provide insights into the time-related properties of fMRI signals that are crucial for accurate analyses. This will be of particular importance as the field moves towards fMRI studies focused on rapid neural dynamics and higher-level cognition.

Functional magnetic resonance imaging (fMRI) enables non-invasive measurement of human brain activity via the hemodynamic response. When activity in a population of neurons changes, these changes give rise to the blood-oxygenation-level-dependent (BOLD) signal measured in most fMRI studies (Ogawa et al., 1990). Thus far, however, BOLD fMRI has exhibited relatively limited ability to provide the fine-grained temporal information necessary for deepening our understanding of brain dynamics. This is due to the fact that the signals obtained from BOLD fMRI are not direct measures of neural activity, but rather reflect the coupling between neuronal activity and the hemodynamic response, which evolves on a time course of seconds (Kwong et al., 1992; Ogawa et al., 1990). This coupling between neural activity and the BOLD signal can be represented by the hemodynamic response function (HRF) (Aguirre et al., 1998; Handwerker et al., 2004). The properties of the HRF depend on many interconnected factors, including the effects of local vascular architecture and cerebrovascular dynamics, that vary substantially across the brain and between individuals (Aguirre et al., 1998; Handwerker et al., 2004; Len and Neary, 2011; Logothetis et al., 2001). The relative timing and shape of the HRF, therefore, also varies considerably across brain regions and even between neighboring voxels (Buckner et al., 1998; Lee et al., 1995; Birn et al., 2001; Siero et al., 2011; Miezin et al., 2000; Siero et al., 2015). Hemodynamic response temporal lag variation is substantially larger than many neural effects of interest, introducing variability on the order of several seconds (Buckner et al., 1998; Lee et al., 1995; Birn et al., 2001; Siero et al., 2011). Thus, to enable inferences about the relative timing of neural activity using signals obtained from BOLD fMRI, it is crucial to understand the variations in HRF timing across the brain.

Advances in acquisition technology now allow high-resolution whole brain fMRI data to be acquired at fast (<500ms) rates (Polimeni and Lewis, 2021; Chen et al., 2019; Barth et al., 2016; Chiew et al., 2018; Hennig et al., 2007; Setsompop et al., 2016), suggesting that fMRI could provide a unique tool to noninvasively track temporal sequences of neural activity (Menon et al., 1998) across the entire brain. Indeed, recent studies have revealed highly structured temporal dynamics using fMRI and suggest that fMRI can enable whole-brain mapping of temporal sequences (Lee et al., 2013; Setzer et al., 2021; Raut et al., 2021; Mitra et al., 2016; Vizioli et al., 2018). Furthermore, hemodynamic signals have been shown to contain more information about fast and high-frequency activity than previously thought (Lee et al., 2013; Smith et al., 2012; Chen and Glover, 2015; Lewis et al., 2016; Sasai et al., 2021). Studies examining individual brain regions have demonstrated that fMRI can achieve impressive temporal precision within regions, on the order of 100ms (Lin et al., 2013), meaning that high fidelity temporal information is present within these hemodynamic signals. However, a key remaining challenge is that the hemodynamic differences across the brain confound our ability to infer the timing of the underlying neural activity from BOLD fMRI. Specifically, if a given brain region shows earlier BOLD activity, it could be due to faster neural activity or simply due to a faster hemodynamic response in that region. Fully exploiting the higher temporal resolution provided by fast fMRI techniques will therefore ultimately require accounting for differences in the temporal dynamics of the hemodynamic response across the whole brain.

Despite the well-known heterogeneity of hemodynamic timing, most common analysis approaches for BOLD fMRI data assume a standard, canonical HRF shape throughout the brain (Glover, 1999). This approach is understandable, since the true HRF is not known, but it nevertheless cannot account for the vascular confound introduced by hemodynamic response variability. Incorrect assumptions about the shape and timing of the HRF can lead to incorrect inferences regarding the underlying neural activity (Gonzalez-Castillo et al., 2012; Lindquist et al., 2009; Handwerker et al., 2012), and studies that assume a whole brain canonical HRF are unable to decouple the neural and vascular components of the BOLD signal (Handwerker et al., 2004; Rangaprakash et al., 2018; Rangaprakash et al., 2017; Deshpande et al., 2010; Chang et al., 2008). Even when using flexible modeling approaches, such as basis sets or finite impulse response models, it is not possible to determine whether a given region’s faster fMRI response reflects fast neural activity, or simply faster local neurovascular coupling (Handwerker et al., 2012).

The fact that most studies do not account for variations in HRF dynamics is largely due to methodological challenges. Previous work has demonstrated that it is possible to quantify hemodynamic lags across brain regions, and even on a voxel-wise level, to detect the relative order of BOLD responses with high temporal precision (Siero et al., 2011; Miezin et al., 2000; Siero et al., 2015; Lin et al., 2013; Chang et al., 2008; Sicard and Duong, 2005; Lin et al., 2018; Misaki et al., 2013; Kastrup et al., 1999; Liu et al., 2017; Bright et al., 2009; Chen et al., 2021; Wu et al., 2013; Wu et al., 2021). One such method is the use of a stimulus paradigm that drives activity in particular brain regions where the neuronal response properties are relatively well understood and controlled such as primary sensory or motor cortices (Handwerker et al., 2004; Birn et al., 2001; Siero et al., 2011; Miezin et al., 2000; Lin et al., 2018). However, this approach cannot be applied to the majority of the brain, where the neuronal response properties are not known ahead of time. Alternatively, using a breath hold or similar hypercapnic challenge can modulate cerebral blood flow (CBF) to all vascularized regions with minimal changes in cerebral metabolic rate of oxygen (CMRO₂), allowing mapping of vascular latencies (Chang et al., 2008; Sicard and Duong, 2005; Kastrup et al., 1999; Liu et al., 2017; Bright et al., 2009; Chen et al., 2021; Pinto et al., 2021). However, breath hold tasks are not suitable for all subject populations, as some patients may have difficulty complying with the breath hold task. Furthermore, breath hold tasks modulate and measure cerebrovascular reactivity (CVR), which contributes to neurovascular coupling but is a distinct process. While neurovascular coupling reflects the alterations in local hemodynamics that occur in response to changes in neural activity, CVR is specifically a measure of a blood vessel’s capacity to dilate and constrict in response to a vasoactive stimulus, and does not include the extensive metabolic and molecular factors that also drive neurovascular coupling (Pinto et al., 2021; D’Esposito et al., 2003; Iadecola, 2017). In fact, there is evidence that while CVR is affected by healthy aging, some metrics of neurovascular coupling are not, hinting that distinct mechanisms may shape these two patterns (Stefanidis et al., 2019; West et al., 2019; Grinband et al., 2017).

A task-free, neurovascular-based approach for detecting the lags of intrinsic neurovascular coupling would therefore be broadly relevant for analyzing fMRI data. A potential alternative route towards identifying local hemodynamic properties is to examine the properties of resting-state fMRI data. Resting-state fMRI signals reflect neurovascular coupling induced by spontaneous neural activity (Mateo et al., 2017; Ma et al., 2016) and confer the additional benefit of being task-independent, which makes it a viable scan type for patient populations. Moreover, unlike stimulus- or task-based paradigms for mapping local hemodynamic response timings in specific brain regions, resting-state fMRI can be used to examine the HRF across the whole brain. These advantages have prompted past research into the utility of resting-state fMRI signals to estimate the HRF itself with prior work utilizing deconvolution approaches to explore HRF timing in resting-state data (Wu et al., 2013; Wu et al., 2021; Sreenivasan et al., 2015). However, these approaches require assumptions about the underlying neural events, which are not known. We therefore investigated whether intrinsic signatures of local neurovascular coupling dynamics are present in the resting-state fMRI signal.

Our goal was to understand whether information about the temporal dynamics of the hemodynamic response is present in resting-state fMRI data. We first used simulations of the BOLD response to illustrate how distinct, physiologically relevant HRF shapes should produce marked differences in the frequency content of resting-state signals. Next, we verified this result in fast fMRI data collected at 7 Tesla (T) using visual stimulation to induce a neural response with known timing in primary visual cortex (V1). We quantified the temporal delay of voxels in V1 in response to a controlled, oscillating visual stimulus, and found that voxels with fast and slow hemodynamic responses exhibited distinct resting-state spectral features. We further extended our analyses to the visual thalamus (lateral geniculate nucleus, LGN) and found that this principle generalized to subcortex. To understand the potential of this information as a tool to predict the temporal properties of individual voxels, we then trained classifiers to use information from the resting-state spectrum to classify voxels as being fast or slow cortical voxels, or even faster LGN voxels. We found that resting-state signals were better predictors of voxel-wise differences in relative hemodynamic timing than latencies measured from a gold standard breath hold task. Our results establish that information about hemodynamic timing can be extracted from the frequency spectrum of resting-state fMRI signals. This demonstrates that resting-state fMRI can provide a way to understand and predict the temporal dynamics of the HRF across the brain, which is critical for interpreting neural activity using BOLD fMRI.

We conclude that the frequency spectrum of resting-state fMRI signals contains rich information about local hemodynamic timing. Our simulations first illustrated the rationale for this effect: HRFs with faster temporal dynamics produce less power in the low frequency bands and a shallower attenuation of power at high frequencies. This effect is preserved even when we account for the higher amplitude of slower HRFs and the 1 /f decay of the amplitude of spontaneous neural activity. Following this observation, we constructed several quantitative features based on the resting-state spectrum, each capturing a similar property: the relative amplitude of high vs. low frequencies, and when used as predictors for a SVM were able to classify individual voxels as fast or slow. Our findings demonstrate that the temporal properties of the HRF affect the spectral features of resting-state fMRI signals and present a framework for characterizing the temporal properties of the hemodynamic response across voxels, which is crucial for accurate fMRI analyses.

The work presented here has broad implications for fMRI studies using the frequency content of the BOLD signal to make inferences about intrinsic brain activity. Prior work has identified changes in the spectral content of fMRI signals in certain clinical populations, including major depressive disorder (Wang et al., 2016), mild cognitive impairment (Han et al., 2011), and Alzheimer’s disease (Yang et al., 2020), and interpreted these differences as changes in intrinsic brain activity. However, as seen here, differences in hemodynamic response timings will also alter the frequency content of fMRI signals. Furthermore, prior work has shown that intrinsic timescales of BOLD activity vary across brain regions and can be used to predict individual subject patterns (Fallon et al., 2020). Our work suggests hemodynamic differences may contribute to these observations. Therefore, changes in the spectral content of fMRI signals can arise not just from differences in intrinsic brain activity but could also be indicators of different hemodynamic response timing.

A common theme across the spectral features we constructed was that they were sensitive to the relative contributions of low- and high-frequency power. Two features we selected (the slope of the spectrum and the exponent of the aperiodic 1 /f fit) are direct measures of the attenuation in power towards higher frequencies. Similar information is contained in the commonly used metrics ALFF and fALFF (Zou et al., 2008): ALFF is a marker of the power in low frequency bands and fALFF is a measure of the ratio of high to low frequency power. While each of these features showed significant differences between fast and slow cortical voxels, only the slope and fALFF exhibited significant differences within each individual subject. These features explicitly capture the relative difference in high-frequency vs. low-frequency power, corresponding to the main prediction of our simulations. Conversely, ALFF had the poorest sensitivity within subjects which could be due to the fact that ALFF is the only feature that does not include information from both low and high frequency ranges and is instead limited to a relatively narrow band of frequencies, 0.01–0.08 Hz. Overall, our results suggest that while some information is contained in the magnitude of the fMRI signal alone, capturing the relative power in high versus low frequencies is the most important metric for predicting hemodynamic timing. Moreover, using the resting-state spectrum itself did not perform significantly better than our chosen features in classifying fast and slow voxels, suggesting that the relative power is indeed the primary contributor when predicting variations in local HRF timings.

Our results highlight the substantial variability in hemodynamic timing even within a single cortical area, and provide a way to predict this speed, but the specific reasons why individual voxels are fast or slow are not yet clear. A multitude of neurovascular and anatomical factors contribute to the variability in hemodynamic response timing across the brain. For example, the anatomical organization of the vascular network inherently introduces temporal dispersion to the hemodynamic response. As blood travels through the vascular network in response to changes in neuronal activity it not only encounters a range of vessel diameters, but also many branch points which each introduce delays. Prior work has shown that the timing and duration of the hemodynamic response both increase towards the cortical surface while the fastest, narrowest hemodynamic responses occurred in deeper gray matter (Siero et al., 2011; de Zwart et al., 2005; Duvernoy et al., 1981). This suggests the possibility that the fast responding voxels identified in our study are more concentrated in deeper gray matter, although the spatial resolution of our images was not sufficient to separate voxels by cortical depth in this study. Vascular elasticity and reactivity also change across the vascular network and vary between regions further contributing to the variable timing of the hemodynamic response across the brain (Uludağ and Blinder, 2018; Drew, 2019). Arteries are surrounded by smooth muscle cells that rapidly relax and expand their diameter in response to neural activity while veins dilate much more slowly (Drew, 2019). Local differences in the relative concentrations of arteries and veins could therefore also introduce differences in the temporal dynamics of the BOLD signal (Siero et al., 2011; de Zwart et al., 2005; Taylor et al., 2018), suggesting that the slow responding voxels may be more likely to contain veins. These relationships between vascular anatomy and timing further highlight that fMRI studies analyzing spectral power or temporal autocorrelation should take into consideration how local vascular anatomy affects these metrics.

A unique advantage of fMRI is its ability to image throughout the whole brain, and our results suggest the potential for extracting more information from fMRI studies of subcortex. Due to significant improvements in both the sensitivity and spatial resolution of fMRI, an increasing number of studies are utilizing fMRI to study small, deep brain structures such as the thalamus and brainstem (Beissner et al., 2014; Sclocco et al., 2018; Saranathan et al., 2021; Beissner and Baudrexel, 2014). The ability to image these deeper brain structures in humans opens the door to studying diverse aspects of cognition associated with these deeper brain regions (Beissner et al., 2014; Sclocco et al., 2018; Saranathan et al., 2021; Beissner and Baudrexel, 2014; Sherman, 2007). However, these deep brain structures also have unique physiological and anatomical properties that alter their vascular dynamics (Lau et al., 2011; Yen et al., 2011; Lewis et al., 2018; Duvernoy, 2009; Devonshire et al., 2012). Faster hemodynamic responses are frequently reported in subcortex, but hemodynamic responses in these regions are less well characterized than in the cortex. Our analyses found that variations in HRF timing are reflected in the frequency spectra of thalamic voxels as well, despite their differing signal-to-noise ratio. This result demonstrates the utility of our approach in subcortex and could benefit neuroimaging studies of structures such as the thalamus, a target of increasing interest in fMRI.

Improvements in the temporal resolution of BOLD fMRI have also sparked interest in detecting neural sequences at sub-second timescales, which are highly relevant for many studies of cognition. Recent studies have leveraged fast fMRI to detect rapid sequences of neural events related to visual sequence detection (Wittkuhn and Schuck, 2021), auditory dynamics (Frühholz et al., 2020), and changes in arousal state (Setzer et al., 2021). As we continue to identify these rapid neural sequences, it will become even more crucial to consider how the hemodynamic response varies across regions to determine whether a given sequence represents regional differences in neuronal or in hemodynamic timing. Considering spectral signatures can support inference of precise timing of neural activity by providing information about relative hemodynamic latencies between voxels and regions.

We found that the resting-state signals predicted voxel-wise differences in relative hemodynamic response timing significantly better using a breath hold to estimate vascular delays, the current gold standard. This observation could be explained by multiple factors. One potential factor is subject motion (Figure 5—figure supplement 1). There was significantly more motion during the breath hold task compared to the visual stimulus (Wilcoxon rank-sum test, p=0.01), which could lead to breath tasks providing less reliable information. Additionally, the models trained using the breath hold latencies were only able to leverage one piece of information: the measure of vascular latency derived from cross-correlation. The information from taking the cross-correlation in the time domain may also be more sensitive to noise, especially at the fast sampling rate of our scans. However, a more important factor may be the biological mechanisms generating each signal. The breath hold task directly modulates cerebrovascular reactivity with minimal accompanying changes in CMRO₂, allowing for assessment of local cerebral vascular reactivity uncoupled from neuronal activation (Kastrup et al., 1999). CVR is important to assess in many clinical applications, and the breath hold-based approach remains a gold standard for CVR mapping (Kastrup et al., 1999; Liu et al., 2017; Bright et al., 2009; Pinto et al., 2021). However, while CVR is a significant modulator of the hemodynamic response, it is only one component of neurovascular coupling, and there are many other factors and signaling pathways that also contribute. In particular, local metabolic factors and feedback mechanisms also modulate of blood flow in the brain and are not replicated in the breath hold task (Iadecola, 2017; Hosford and Gourine, 2019). The hemodynamic responses induced by neural activity may, therefore, not be identical in timing to those induced by a purely vascular signal. By contrast, the signals obtained from resting-state fMRI are coupled to underlying neural activity in an analogous manner as during a task condition (Mateo et al., 2017; Ma et al., 2016; Fox and Raichle, 2007). The improved performance of the resting-state-based prediction may therefore reflect that it is intrinsically more similar to a task than the breath hold condition, as the underlying biological origins of resting-state signals share common mechanisms with task-induced neurovascular coupling. Future applications may therefore benefit from continuing to use breath tasks as a gold standard to assess CVR, whereas resting-state analyses may be a better metric of neurovascular coupling.

One limitation of our study is that our approach has thus far only been validated in the visual system, which is the basis for the majority of our knowledge about the HRF. Given that the HRF has been shown to vary across the brain, the generalizability of our approach to different brain regions has yet to be established. However, studies that have empirically derived the HRF in other brain regions, such as the motor cortex (Siero et al., 2011) or somatosensory cortex (de Zwart et al., 2005), have produced HRFs with similar shapes and parameters to the double-gamma HRF utilized in our simulations (Handwerker et al., 2004; Siero et al., 2011; de Zwart et al., 2005; Taylor et al., 2018) providing evidence that this approach may be generalizable across the brain. Still, future work could investigate this approach in other primary sensory systems such as the auditory system where there are reliable stimulation paradigms. Additionally, this study used data collected at an ultra-high magnetic field strength (7T) which affords a higher signal-to-noise ratio (SNR) compared to more traditional 3T fMRI. However, we also performed the same spectral analysis in an independent dataset acquired at 3T and robustly replicated our result, demonstrating that this pattern is preserved despite a lower SNR and different acquisition parameters (Figure 3—figure supplement 1).

The work discussed here suggests a wide range of neuroscience applications for our approach to measuring hemodynamic timing. One logical next step would be to use the ability to characterize temporal variation in the HRF to not just predict, but to correct for vascular delays. Previous work demonstrated that correcting for varying hemodynamic latencies across the brain can affect functional connectivity analyses (Rangaprakash et al., 2018; Chang et al., 2008). Our results could further enhance removal of non-neural latency differences that confound functional connectivity metrics, both static and dynamic, and can increase confidence that the networks we are analyzing are derived from neuronal dynamics. This has become increasingly of interest as more studies use functional connectivity, particularly in resting-state, to study dynamics underlying diseases such as PTSD (Jin et al., 2017; Rabinak et al., 2011; Maron-Katz et al., 2020; Zhu et al., 2017), Alzheimer’s Disease (Greicius et al., 2004; Bai et al., 2008; Agosta et al., 2012; Koch et al., 2012), Parkinson’s Disease (Helmich et al., 2010; Hacker et al., 2012; Agosta et al., 2014; Baggio et al., 2015; Putcha et al., 2015), and others (Filippi et al., 2019). Since changes in neurovascular function have been observed in many disorders (de la Torre and Mecocci, 2018; Gutteridge et al., 2020; Burrage et al., 2018), analyzing spectral dynamics may help interpret functional connectivity differences in clinical populations.

Although we focused on the utility of resting-state spectral information to classify voxels as having relatively fast or slow hemodynamic response timings, we did also examine the correlation with absolute timing measures. We found that most subjects had significant positive correlations between each of the spectral features and the task response latency. Despite this fact, when estimating continuous predictions of absolute hemodynamic response lag, prediction performance was poor. Importantly, previous evidence has shown that the absolute timing of the HRF varies between task and resting-state conditions (Chen and Glover, 2015), suggesting that caution is needed in using absolute measures of hemodynamic latencies, as these may not generalize across conditions. Measuring relative differences in HRF timing might therefore be a more effective method to correct for variations in the hemodynamic response across brain regions, since task state can modulate the absolute timing of the hemodynamic response. Furthermore, the relative differences provide the key information necessary to interpret sequences of activity across brain regions, enabling examination of whether purely vascular differences are present across those regions.

Together, our results demonstrate that the resting-state fMRI signal contains information about local hemodynamic response speeds. This approach can help understand brain-wide variations in HRF dynamics, which is critical as the field moves toward a new era of fMRI studies utilizing fast fMRI to study rapid neuronal dynamics and higher level cognition.

The data used in this paper has been deposited on OpenNeuro (https://doi.org/10.18112/openneuro.ds004645.v1.0.0).

1. 1. Bailes SM
   2. Gomez DEP
   3. Setzer B
   4. Lewis LD

   (2023) OpenNeuro

   Resting-state fMRI signals contain spectral signatures of local hemodynamic response timing.

   https://doi.org/10.18112/openneuro.ds004645.v1.0.0

1. 1. Agosta F
   2. Pievani M
   3. Geroldi C
   4. Copetti M
   5. Frisoni GB
   6. Filippi M

   (2012) Resting state fMRI in Alzheimer’s disease: beyond the default mode network

   Neurobiology of Aging 33:1564–1578.

   https://doi.org/10.1016/j.neurobiolaging.2011.06.007

   • Google Scholar
2. 1. Agosta F
   2. Caso F
   3. Stankovic I
   4. Inuggi A
   5. Petrovic I
   6. Svetel M
   7. Kostic VS
   8. Filippi M

   (2014) Cortico-striatal-thalamic network functional connectivity in hemiparkinsonism

   Neurobiology of Aging 35:2592–2602.

   https://doi.org/10.1016/j.neurobiolaging.2014.05.032

   • Google Scholar
3. 1. Agrawal U
   2. Brown EN
   3. Lewis LD

   (2020) Model-based physiological noise removal in fast fMRI

   NeuroImage 205:116231.

   https://doi.org/10.1016/j.neuroimage.2019.116231

   • PubMed
   • Google Scholar
4. 1. Aguirre GK
   2. Zarahn E
   3. D’esposito M

   (1998) The variability of human, BOLD hemodynamic responses

   NeuroImage 8:360–369.

   https://doi.org/10.1006/nimg.1998.0369

   • PubMed
   • Google Scholar
5. 1. Baggio H-C
   2. Segura B
   3. Sala-Llonch R
   4. Marti M-J
   5. Valldeoriola F
   6. Compta Y
   7. Tolosa E
   8. Junqué C

   (2015) Cognitive impairment and resting-state network connectivity in Parkinson’s disease

   Human Brain Mapping 36:199–212.

   https://doi.org/10.1002/hbm.22622

   • PubMed
   • Google Scholar
6. 1. Bai F
   2. Zhang Z
   3. Yu H
   4. Shi Y
   5. Yuan Y
   6. Zhu W
   7. Zhang X
   8. Qian Y

   (2008) Default-mode network activity distinguishes amnestic type mild cognitive impairment from healthy aging: A combined structural and resting-state functional MRI study

   Neuroscience Letters 438:111–115.

   https://doi.org/10.1016/j.neulet.2008.04.021

   • Google Scholar
7. 1. Barth M
   2. Breuer F
   3. Koopmans PJ
   4. Norris DG
   5. Poser BA

   (2016) Simultaneous multislice (SMS) imaging techniques

   Magnetic Resonance in Medicine 75:63–81.

   https://doi.org/10.1002/mrm.25897

   • Google Scholar
8. 1. Beissner F
   2. Baudrexel S

   (2014) Investigating the human brainstem with structural and functional MRI

   Frontiers in Human Neuroscience 8:116.

   https://doi.org/10.3389/fnhum.2014.00116

   • PubMed
   • Google Scholar
9. 1. Beissner F
   2. Schumann A
   3. Brunn F
   4. Eisenträger D
   5. Bär KJ

   (2014) Advances in functional magnetic resonance imaging of the human brainstem

   NeuroImage 86:91–98.

   https://doi.org/10.1016/j.neuroimage.2013.07.081

   • PubMed
   • Google Scholar
10. 1. Birn RM
    2. Saad ZS
    3. Bandettini PA

    (2001) Spatial heterogeneity of the nonlinear dynamics in the FMRI BOLD response

    NeuroImage 14:817–826.

    https://doi.org/10.1006/nimg.2001.0873

    • PubMed
    • Google Scholar
11. 1. Bokil H
    2. Andrews P
    3. Kulkarni JE
    4. Mehta S
    5. Mitra PP

    (2010) Chronux: A platform for analyzing neural signals

    Journal of Neuroscience Methods 192:146–151.

    https://doi.org/10.1016/j.jneumeth.2010.06.020

    • PubMed
    • Google Scholar
12. 1. Brainard DH

    (1997)

    The Psychophysics Toolbox

    Spatial Vision 10:433–436.

    • PubMed
    • Google Scholar
13. 1. Bright MG
    2. Bulte DP
    3. Jezzard P
    4. Duyn JH

    (2009) Characterization of regional heterogeneity in cerebrovascular reactivity dynamics using novel hypocapnia task and BOLD fMRI

    NeuroImage 48:166–175.

    https://doi.org/10.1016/j.neuroimage.2009.05.026

    • PubMed
    • Google Scholar
14. 1. Buckner RL
    2. Koutstaal W
    3. Schacter DL
    4. Dale AM
    5. Rotte M
    6. Rosen BR

    (1998) Functional-anatomic study of episodic retrieval. II. Selective averaging of event-related fMRI trials to test the retrieval success hypothesis

    NeuroImage 7:163–175.

    https://doi.org/10.1006/nimg.1998.0328

    • PubMed
    • Google Scholar
15. 1. Burrage E
    2. Marshall KL
    3. Santanam N
    4. Chantler PD

    (2018) Cerebrovascular dysfunction with stress and depression

    Brain Circulation 4:43–53.

    https://doi.org/10.4103/bc.bc_6_18

    • PubMed
    • Google Scholar
16. 1. Chang C
    2. Thomason ME
    3. Glover GH

    (2008) Mapping and correction of vascular hemodynamic latency in the BOLD signal

    NeuroImage 43:90–102.

    https://doi.org/10.1016/j.neuroimage.2008.06.030

    • PubMed
    • Google Scholar
17. 1. Chen JE
    2. Glover GH

    (2015) BOLD fractional contribution to resting-state functional connectivity above 0.1 Hz

    NeuroImage 107:207–218.

    https://doi.org/10.1016/j.neuroimage.2014.12.012

    • PubMed
    • Google Scholar
18. 1. Chen JE
    2. Polimeni JR
    3. Bollmann S
    4. Glover GH

    (2019) On the analysis of rapidly sampled fMRI data

    NeuroImage 188:807–820.

    https://doi.org/10.1016/j.neuroimage.2019.02.008

    • PubMed
    • Google Scholar
19. 1. Chen K
    2. Yang H
    3. Zhang H
    4. Meng C
    5. Becker B
    6. Biswal B

    (2021) Altered cerebrovascular reactivity due to respiratory rate and breath holding: a BOLD-fMRI study on healthy adults

    Brain Structure & Function 226:1229–1239.

    https://doi.org/10.1007/s00429-021-02236-5

    • PubMed
    • Google Scholar
20. 1. Chiew M
    2. Graedel NN
    3. Miller KL

    (2018) Recovering task fMRI signals from highly under-sampled data with low-rank and temporal subspace constraints

    NeuroImage 174:97–110.

    https://doi.org/10.1016/j.neuroimage.2018.02.062

    • PubMed
    • Google Scholar
21. 1. de la Torre J
    2. Mecocci P

    (2018) The vascular hypothesis of alzheimer’s disease: a key to preclinical prediction of dementia using neuroimaging

    Journal of Alzheimer’s Disease 63:35–52.

    https://doi.org/10.3233/JAD-180004

    • Google Scholar
22. 1. de Zwart JA
    2. Silva AC
    3. van Gelderen P
    4. Kellman P
    5. Fukunaga M
    6. Chu R
    7. Koretsky AP
    8. Frank JA
    9. Duyn JH

    (2005) Temporal dynamics of the BOLD fMRI impulse response

    NeuroImage 24:667–677.

    https://doi.org/10.1016/j.neuroimage.2004.09.013

    • PubMed
    • Google Scholar
23. 1. Deshpande G
    2. Sathian K
    3. Hu X

    (2010) Effect of hemodynamic variability on Granger causality analysis of fMRI

    NeuroImage 52:884–896.

    https://doi.org/10.1016/j.neuroimage.2009.11.060

    • Google Scholar
24. 1. D’Esposito M
    2. Deouell LY
    3. Gazzaley A

    (2003) Alterations in the BOLD fMRI signal with ageing and disease: a challenge for neuroimaging

    Nature Reviews. Neuroscience 4:863–872.

    https://doi.org/10.1038/nrn1246

    • PubMed
    • Google Scholar
25. 1. Devonshire IM
    2. Papadakis NG
    3. Port M
    4. Berwick J
    5. Kennerley AJ
    6. Mayhew JEW
    7. Overton PG

    (2012) Neurovascular coupling is brain region-dependent

    NeuroImage 59:1997–2006.

    https://doi.org/10.1016/j.neuroimage.2011.09.050

    • Google Scholar
26. 1. Drew PJ

    (2019) Vascular and neural basis of the BOLD signal

    Current Opinion in Neurobiology 58:61–69.

    https://doi.org/10.1016/j.conb.2019.06.004

    • Google Scholar
27. 1. Duvernoy HM
    2. Delon S
    3. Vannson JL

    (1981) Cortical blood vessels of the human brain

    Brain Research Bulletin 7:519–579.

    https://doi.org/10.1016/0361-9230(81)90007-1

    • PubMed
    • Google Scholar
28. Book

    1. Duvernoy H

    (2009)

    The Human Brain

    Springer-Verlag.

    • Google Scholar
29. 1. Fallon J
    2. Ward PGD
    3. Parkes L
    4. Oldham S
    5. Arnatkevičiūtė A
    6. Fornito A
    7. Fulcher BD

    (2020) Timescales of spontaneous fMRI fluctuations relate to structural connectivity in the brain

    Network Neuroscience 4:788–806.

    https://doi.org/10.1162/netn_a_00151

    • Google Scholar
30. 1. Filippi M
    2. Spinelli EG
    3. Cividini C
    4. Agosta F

    (2019) Resting state dynamic functional connectivity in neurodegenerative conditions: a review of magnetic resonance imaging findings

    Frontiers in Neuroscience 13:657.

    https://doi.org/10.3389/fnins.2019.00657

    • Google Scholar
31. 1. Fischl B

    (2012) FreeSurfer

    NeuroImage 62:774–781.

    https://doi.org/10.1016/j.neuroimage.2012.01.021

    • PubMed
    • Google Scholar
32. 1. Fox MD
    2. Raichle ME

    (2007) Spontaneous fluctuations in brain activity observed with functional magnetic resonance imaging

    Nature Reviews. Neuroscience 8:700–711.

    https://doi.org/10.1038/nrn2201

    • PubMed
    • Google Scholar
33. 1. Frühholz S
    2. Trost W
    3. Grandjean D
    4. Belin P

    (2020) Neural oscillations in human auditory cortex revealed by fast fMRI during auditory perception

    NeuroImage 207:116401.

    https://doi.org/10.1016/j.neuroimage.2019.116401

    • PubMed
    • Google Scholar
34. 1. Glover GH

    (1999) Deconvolution of impulse response in event-related BOLD fMRI

    NeuroImage 9:416–429.

    https://doi.org/10.1006/nimg.1998.0419

    • PubMed
    • Google Scholar
35. 1. Glover GH
    2. Li TQ
    3. Ress D

    (2000) Image-based method for retrospective correction of physiological motion effects in fMRI: RETROICOR

    Magnetic Resonance in Medicine 44:162–167.

    https://doi.org/10.1002/1522-2594(200007)44:1<162::aid-mrm23>3.0.co;2-e

    • PubMed
    • Google Scholar
36. 1. Gonzalez-Castillo J
    2. Saad ZS
    3. Handwerker DA
    4. Inati SJ
    5. Brenowitz N
    6. Bandettini PA

    (2012) Whole-brain, time-locked activation with simple tasks revealed using massive averaging and model-free analysis

    PNAS 109:5487–5492.

    https://doi.org/10.1073/pnas.1121049109

    • PubMed
    • Google Scholar
37. 1. Greicius MD
    2. Srivastava G
    3. Reiss AL
    4. Menon V

    (2004) Default-mode network activity distinguishes Alzheimer’s disease from healthy aging: Evidence from functional MRI

    PNAS 101:4637–4642.

    https://doi.org/10.1073/pnas.0308627101

    • Google Scholar
38. 1. Greve DN
    2. Fischl B

    (2009) Accurate and robust brain image alignment using boundary-based registration

    NeuroImage 48:63–72.

    https://doi.org/10.1016/j.neuroimage.2009.06.060

    • PubMed
    • Google Scholar
39. 1. Grinband J
    2. Steffener J
    3. Razlighi QR
    4. Stern Y

    (2017) BOLD neurovascular coupling does not change significantly with normal aging

    Human Brain Mapping 38:3538–3551.

    https://doi.org/10.1002/hbm.23608

    • PubMed
    • Google Scholar
40. 1. Gutteridge DS
    2. Saredakis D
    3. Badcock NA
    4. Collins-Praino LE
    5. Keage HAD

    (2020) Cerebrovascular function during cognition in Parkinson’s disease: A functional transcranial Doppler sonography study

    Journal of the Neurological Sciences 408:116578.

    https://doi.org/10.1016/j.jns.2019.116578

    • PubMed
    • Google Scholar
41. 1. Hacker CD
    2. Perlmutter JS
    3. Criswell SR
    4. Ances BM
    5. Snyder AZ

    (2012) Resting state functional connectivity of the striatum in Parkinson’s disease

    Brain 135:3699–3711.

    https://doi.org/10.1093/brain/aws281

    • PubMed
    • Google Scholar
42. 1. Han Y
    2. Wang J
    3. Zhao Z
    4. Min B
    5. Lu J
    6. Li K
    7. He Y
    8. Jia J

    (2011) Frequency-dependent changes in the amplitude of low-frequency fluctuations in amnestic mild cognitive impairment: A resting-state fMRI study

    NeuroImage 55:287–295.

    https://doi.org/10.1016/j.neuroimage.2010.11.059

    • PubMed
    • Google Scholar
43. 1. Handwerker DA
    2. Ollinger JM
    3. D’Esposito M

    (2004) Variation of BOLD hemodynamic responses across subjects and brain regions and their effects on statistical analyses

    NeuroImage 21:1639–1651.

    https://doi.org/10.1016/j.neuroimage.2003.11.029

    • PubMed
    • Google Scholar
44. 1. Handwerker DA
    2. Gonzalez-Castillo J
    3. D’Esposito M
    4. Bandettini PA

    (2012) The continuing challenge of understanding and modeling hemodynamic variation in fMRI

    NeuroImage 62:1017–1023.

    https://doi.org/10.1016/j.neuroimage.2012.02.015

    • Google Scholar
45. 1. Helmich RC
    2. Derikx LC
    3. Bakker M
    4. Scheeringa R
    5. Bloem BR
    6. Toni I

    (2010) Spatial remapping of cortico-striatal connectivity in Parkinson’s disease

    Cerebral Cortex 20:1175–1186.

    https://doi.org/10.1093/cercor/bhp178

    • PubMed
    • Google Scholar
46. 1. Hennig J
    2. Zhong K
    3. Speck O

    (2007) MR-Encephalography: Fast multi-channel monitoring of brain physiology with magnetic resonance

    NeuroImage 34:212–219.

    https://doi.org/10.1016/j.neuroimage.2006.08.036

    • Google Scholar
47. 1. Hosford PS
    2. Gourine AV

    (2019) What is the key mediator of the neurovascular coupling response?

    Neuroscience and Biobehavioral Reviews 96:174–181.

    https://doi.org/10.1016/j.neubiorev.2018.11.011

    • PubMed
    • Google Scholar
48. 1. Iadecola C

    (2017) The neurovascular unit coming of age: a journey through neurovascular coupling in health and disease

    Neuron 96:17–42.

    https://doi.org/10.1016/j.neuron.2017.07.030

    • PubMed
    • Google Scholar
49. 1. Iglesias JE
    2. Insausti R
    3. Lerma-Usabiaga G
    4. Bocchetta M
    5. Van Leemput K
    6. Greve DN
    7. van der Kouwe A
    8. Fischl B
    9. Caballero-Gaudes C
    10. Paz-Alonso PM

    (2018) A probabilistic atlas of the human thalamic nuclei combining ex vivo MRI and histology

    NeuroImage 183:314–326.

    https://doi.org/10.1016/j.neuroimage.2018.08.012

    • PubMed
    • Google Scholar
50. 1. Jin C
    2. Jia H
    3. Lanka P
    4. Rangaprakash D
    5. Li L
    6. Liu T
    7. Hu X
    8. Deshpande G

    (2017) Dynamic brain connectivity is a better predictor of PTSD than static connectivity

    Human Brain Mapping 38:4479–4496.

    https://doi.org/10.1002/hbm.23676

    • PubMed
    • Google Scholar
51. 1. Kastrup A
    2. Krüger G
    3. Glover GH
    4. Moseley ME

    (1999) Assessment of cerebral oxidative metabolism with breath holding and fMRI

    Magnetic Resonance in Medicine 42:608–611.

    https://doi.org/10.1002/(sici)1522-2594(199909)42:3<608::aid-mrm26>3.0.co;2-i

    • PubMed
    • Google Scholar
52. 1. Koch W
    2. Teipel S
    3. Mueller S
    4. Benninghoff J
    5. Wagner M
    6. Bokde ALW
    7. Hampel H
    8. Coates U
    9. Reiser M
    10. Meindl T

    (2012) Diagnostic power of default mode network resting state fMRI in the detection of Alzheimer’s disease

    Neurobiology of Aging 33:466–478.

    https://doi.org/10.1016/j.neurobiolaging.2010.04.013

    • PubMed
    • Google Scholar
53. 1. Kwong KK
    2. Belliveau JW
    3. Chesler DA
    4. Goldberg IE
    5. Weisskoff RM
    6. Poncelet BP
    7. Kennedy DN
    8. Hoppel BE
    9. Cohen MS
    10. Turner R

    (1992) Dynamic magnetic resonance imaging of human brain activity during primary sensory stimulation

    PNAS 89:5675–5679.

    https://doi.org/10.1073/pnas.89.12.5675

    • PubMed
    • Google Scholar
54. 1. Lau C
    2. Zhang JW
    3. Xing KK
    4. Zhou IY
    5. Cheung MM
    6. Chan KC
    7. Wu EX

    (2011) BOLD responses in the superior colliculus and lateral geniculate nucleus of the rat viewing an apparent motion stimulus

    NeuroImage 58:878–884.

    https://doi.org/10.1016/j.neuroimage.2011.06.055

    • PubMed
    • Google Scholar
55. 1. Lee AT
    2. Glover GH
    3. Meyer CH

    (1995) Discrimination of large venous vessels in time-course spiral blood-oxygen-level-dependent magnetic-resonance functional neuroimaging

    Magnetic Resonance in Medicine 33:745–754.

    https://doi.org/10.1002/mrm.1910330602

    • Google Scholar
56. 1. Lee HL
    2. Zahneisen B
    3. Hugger T
    4. LeVan P
    5. Hennig J

    (2013) Tracking dynamic resting-state networks at higher frequencies using MR-encephalography

    NeuroImage 65:216–222.

    https://doi.org/10.1016/j.neuroimage.2012.10.015

    • PubMed
    • Google Scholar
57. 1. Len TK
    2. Neary JP

    (2011) Cerebrovascular pathophysiology following mild traumatic brain injury

    Clinical Physiology and Functional Imaging 31:85–93.

    https://doi.org/10.1111/j.1475-097X.2010.00990.x

    • PubMed
    • Google Scholar
58. 1. Lewis LD
    2. Setsompop K
    3. Rosen BR
    4. Polimeni JR

    (2016) Fast fMRI can detect oscillatory neural activity in humans

    PNAS 113:E6679–E6685.

    https://doi.org/10.1073/pnas.1608117113

    • PubMed
    • Google Scholar
59. 1. Lewis LD
    2. Setsompop K
    3. Rosen BR
    4. Polimeni JR

    (2018) Stimulus-dependent hemodynamic response timing across the human subcortical-cortical visual pathway identified through high spatiotemporal resolution 7T fMRI

    NeuroImage 181:279–291.

    https://doi.org/10.1016/j.neuroimage.2018.06.056

    • PubMed
    • Google Scholar
60. 1. Lin F-H
    2. Witzel T
    3. Raij T
    4. Ahveninen J
    5. Wen-Kai Tsai K
    6. Chu Y-H
    7. Chang W-T
    8. Nummenmaa A
    9. Polimeni JR
    10. Kuo W-J
    11. Hsieh J-C
    12. Rosen BR
    13. Belliveau JW

    (2013) fMRI hemodynamics accurately reflects neuronal timing in the human brain measured by MEG

    NeuroImage 78:372–384.

    https://doi.org/10.1016/j.neuroimage.2013.04.017

    • Google Scholar
61. 1. Lin F-H
    2. Polimeni JR
    3. Lin J-FL
    4. Tsai KW-K
    5. Chu Y-H
    6. Wu P-Y
    7. Li Y-T
    8. Hsu Y-C
    9. Tsai S-Y
    10. Kuo W-J

    (2018) Relative latency and temporal variability of hemodynamic responses at the human primary visual cortex

    NeuroImage 164:194–201.

    https://doi.org/10.1016/j.neuroimage.2017.01.041

    • Google Scholar
62. 1. Lindquist MA
    2. Meng Loh J
    3. Atlas LY
    4. Wager TD

    (2009) Modeling the hemodynamic response function in fMRI: Efficiency, bias and mis-modeling

    NeuroImage 45:S187–S198.

    https://doi.org/10.1016/j.neuroimage.2008.10.065

    • Google Scholar
63. 1. Liu P
    2. Li Y
    3. Pinho M
    4. Park DC
    5. Welch BG
    6. Lu H

    (2017) Cerebrovascular reactivity mapping without gas challenges

    NeuroImage 146:320–326.

    https://doi.org/10.1016/j.neuroimage.2016.11.054

    • PubMed
    • Google Scholar
64. 1. Logothetis NK
    2. Pauls J
    3. Augath M
    4. Trinath T
    5. Oeltermann A

    (2001) Neurophysiological investigation of the basis of the fMRI signal

    Nature 412:150–157.

    https://doi.org/10.1038/35084005

    • PubMed
    • Google Scholar
65. 1. Ma Y
    2. Shaik MA
    3. Kozberg MG
    4. Kim SH
    5. Portes JP
    6. Timerman D
    7. Hillman EMC

    (2016) Resting-state hemodynamics are spatiotemporally coupled to synchronized and symmetric neural activity in excitatory neurons

    PNAS 113:E8463–E8471.

    https://doi.org/10.1073/pnas.1525369113

    • Google Scholar
66. 1. Maron-Katz A
    2. Zhang Y
    3. Narayan M
    4. Wu W
    5. Toll RT
    6. Naparstek S
    7. De Los Angeles C
    8. Longwell P
    9. Shpigel E
    10. Newman J
    11. Abu-Amara D
    12. Marmar C
    13. Etkin A

    (2020) Individual Patterns of Abnormality in Resting-State Functional Connectivity Reveal Two Data-Driven PTSD Subgroups

    American Journal of Psychiatry 177:244–253.

    https://doi.org/10.1176/appi.ajp.2019.19010060

    • Google Scholar
67. 1. Mateo C
    2. Knutsen PM
    3. Tsai PS
    4. Shih AY
    5. Kleinfeld D

    (2017) Entrainment of arteriole vasomotor fluctuations by neural activity is a basis of blood-oxygenation-level-dependent “resting-state” connectivity

    Neuron 96:936–948.

    https://doi.org/10.1016/j.neuron.2017.10.012

    • PubMed
    • Google Scholar
68. 1. Menon RS
    2. Luknowsky DC
    3. Gati JS

    (1998) Mental chronometry using latency-resolved functional MRI

    PNAS 95:10902–10907.

    https://doi.org/10.1073/pnas.95.18.10902

    • Google Scholar
69. 1. Miezin FM
    2. Maccotta L
    3. Ollinger JM
    4. Petersen SE
    5. Buckner RL

    (2000) Characterizing the hemodynamic response: effects of presentation rate, sampling procedure, and the possibility of ordering brain activity based on relative timing

    NeuroImage 11:735–759.

    https://doi.org/10.1006/nimg.2000.0568

    • PubMed
    • Google Scholar
70. 1. Misaki M
    2. Luh WM
    3. Bandettini PA

    (2013) Accurate decoding of sub-TR timing differences in stimulations of sub-voxel regions from multi-voxel response patterns

    NeuroImage 66:623–633.

    https://doi.org/10.1016/j.neuroimage.2012.10.069

    • PubMed
    • Google Scholar
71. 1. Mitra A
    2. Snyder AZ
    3. Hacker CD
    4. Pahwa M
    5. Tagliazucchi E
    6. Laufs H
    7. Leuthardt EC
    8. Raichle ME

    (2016) Human cortical–hippocampal dialogue in wake and slow-wave sleep

    PNAS 113:E6868–E6876.

    https://doi.org/10.1073/pnas.1607289113

    • Google Scholar
72. 1. Ogawa S
    2. Lee TM
    3. Kay AR
    4. Tank DW

    (1990) Brain magnetic resonance imaging with contrast dependent on blood oxygenation

    PNAS 87:9868–9872.

    https://doi.org/10.1073/pnas.87.24.9868

    • PubMed
    • Google Scholar
73. 1. Pedregosa F

    (2011)

    Scikit-learn: machine learning in python

    Journal of Machine Learning Research: JMLR 12:2825–2830.

    • Google Scholar
74. 1. Pinto J
    2. Bright MG
    3. Bulte DP
    4. Figueiredo P

    (2021) Cerebrovascular reactivity mapping without gas challenges: a methodological guide

    Frontiers in Physiology 11:1711.

    https://doi.org/10.3389/fphys.2020.608475

    • Google Scholar
75. 1. Polimeni JR
    2. Lewis LD

    (2021) Imaging faster neural dynamics with fast fMRI: A need for updated models of the hemodynamic response

    Progress in Neurobiology 207:102174.

    https://doi.org/10.1016/j.pneurobio.2021.102174

    • PubMed
    • Google Scholar
76. 1. Putcha D
    2. Ross RS
    3. Cronin-Golomb A
    4. Janes AC
    5. Stern CE

    (2015) Altered intrinsic functional coupling between core neurocognitive networks in Parkinson’s disease

    NeuroImage. Clinical 7:449–455.

    https://doi.org/10.1016/j.nicl.2015.01.012

    • PubMed
    • Google Scholar
77. 1. Rabinak CA
    2. Angstadt M
    3. Welsh RC
    4. Kenndy AE
    5. Lyubkin M
    6. Martis B
    7. Phan KL

    (2011) Altered amygdala resting-state functional connectivity in post-traumatic stress disorder

    Frontiers in Psychiatry 2:62.

    https://doi.org/10.3389/fpsyt.2011.00062

    • PubMed
    • Google Scholar
78. 1. Rangaprakash D
    2. Dretsch MN
    3. Yan W
    4. Katz JS
    5. Denney TS Jr
    6. Deshpande G

    (2017) Hemodynamic variability in soldiers with trauma: Implications for functional MRI connectivity studies

    NeuroImage. Clinical 16:409–417.

    https://doi.org/10.1016/j.nicl.2017.07.016

    • PubMed
    • Google Scholar
79. 1. Rangaprakash D
    2. Wu GR
    3. Marinazzo D
    4. Hu X
    5. Deshpande G

    (2018) Hemodynamic response function (HRF) variability confounds resting-state fMRI functional connectivity

    Magnetic Resonance in Medicine 80:1697–1713.

    https://doi.org/10.1002/mrm.27146

    • PubMed
    • Google Scholar
80. 1. Raut RV
    2. Snyder AZ
    3. Mitra A
    4. Yellin D
    5. Fujii N
    6. Malach R
    7. Raichle ME

    (2021) Global waves synchronize the brain’s functional systems with fluctuating arousal

    Science Advances 7:eabf2709.

    https://doi.org/10.1126/sciadv.abf2709

    • PubMed
    • Google Scholar
81. 1. Saranathan M
    2. Iglehart C
    3. Monti M
    4. Tourdias T
    5. Rutt B

    (2021) In vivo high-resolution structural MRI-based atlas of human thalamic nuclei

    Scientific Data 8:275.

    https://doi.org/10.1038/s41597-021-01062-y

    • PubMed
    • Google Scholar
82. 1. Sasai S
    2. Koike T
    3. Sugawara SK
    4. Hamano YH
    5. Sumiya M
    6. Okazaki S
    7. Takahashi HK
    8. Taga G
    9. Sadato N

    (2021) Frequency-specific task modulation of human brain functional networks: A fast fMRI study

    NeuroImage 224:117375.

    https://doi.org/10.1016/j.neuroimage.2020.117375

    • PubMed
    • Google Scholar
83. 1. Sclocco R
    2. Beissner F
    3. Bianciardi M
    4. Polimeni JR
    5. Napadow V

    (2018) Challenges and opportunities for brainstem neuroimaging with ultrahigh field MRI

    NeuroImage 168:412–426.

    https://doi.org/10.1016/j.neuroimage.2017.02.052

    • Google Scholar
84. 1. Setsompop K
    2. Gagoski BA
    3. Polimeni JR
    4. Witzel T
    5. Wedeen VJ
    6. Wald LL

    (2012) Blipped-controlled aliasing in parallel imaging for simultaneous multislice echo planar imaging with reduced g-factor penalty

    Magnetic Resonance in Medicine 67:1210–1224.

    https://doi.org/10.1002/mrm.23097

    • PubMed
    • Google Scholar
85. 1. Setsompop K
    2. Feinberg DA
    3. Polimeni JR

    (2016) Rapid brain MRI acquisition techniques at ultra-high fields

    NMR in Biomedicine 29:1198–1221.

    https://doi.org/10.1002/nbm.3478

    • Google Scholar
86. 1. Setzer B
    2. Fultz NE
    3. Gomez DEP
    4. Williams SD
    5. Bonmassar G
    6. Polimeni JR
    7. Lewis LD

    (2021) A temporal sequence of thalamic activity unfolds at transitions in behavioral arousal state

    Neuroscience 13:5442.

    https://doi.org/10.1101/2021.12.01.470627

    • Google Scholar
87. 1. Sherman SM

    (2007) The thalamus is more than just a relay

    Current Opinion in Neurobiology 17:417–422.

    https://doi.org/10.1016/j.conb.2007.07.003

    • Google Scholar
88. 1. Sicard KM
    2. Duong TQ

    (2005) Effects of hypoxia, hyperoxia, and hypercapnia on baseline and stimulus-evoked BOLD, CBF, and CMRO2 in spontaneously breathing animals

    NeuroImage 25:850–858.

    https://doi.org/10.1016/j.neuroimage.2004.12.010

    • PubMed
    • Google Scholar
89. 1. Siero JCW
    2. Petridou N
    3. Hoogduin H
    4. Luijten PR
    5. Ramsey NF

    (2011) Cortical depth-dependent temporal dynamics of the BOLD response in the human brain

    Journal of Cerebral Blood Flow and Metabolism 31:1999–2008.

    https://doi.org/10.1038/jcbfm.2011.57

    • PubMed
    • Google Scholar
90. 1. Siero JCW
    2. Hendrikse J
    3. Hoogduin H
    4. Petridou N
    5. Luijten P
    6. Donahue MJ

    (2015) Cortical depth dependence of the BOLD initial dip and poststimulus undershoot in human visual cortex at 7 Tesla

    Magnetic Resonance in Medicine 73:2283–2295.

    https://doi.org/10.1002/mrm.25349

    • PubMed
    • Google Scholar
91. 1. Smith SM
    2. Miller KL
    3. Moeller S
    4. Xu J
    5. Auerbach EJ
    6. Woolrich MW
    7. Beckmann CF
    8. Jenkinson M
    9. Andersson J
    10. Glasser MF
    11. Van Essen DC
    12. Feinberg DA
    13. Yacoub ES
    14. Ugurbil K

    (2012) Temporally-independent functional modes of spontaneous brain activity

    PNAS 109:3131–3136.

    https://doi.org/10.1073/pnas.1121329109

    • PubMed
    • Google Scholar
92. 1. Sreenivasan KR
    2. Havlicek M
    3. Deshpande G

    (2015) Nonparametric hemodynamic deconvolution of FMRI using homomorphic filtering

    IEEE Transactions on Medical Imaging 34:1155–1163.

    https://doi.org/10.1109/TMI.2014.2379914

    • PubMed
    • Google Scholar
93. 1. Stefanidis KB
    2. Askew CD
    3. Klein T
    4. Lagopoulos J
    5. Summers MJ

    (2019) Healthy aging affects cerebrovascular reactivity and pressure-flow responses, but not neurovascular coupling: A cross-sectional study

    PLOS ONE 14:e0217082.

    https://doi.org/10.1371/journal.pone.0217082

    • PubMed
    • Google Scholar
94. 1. Taylor AJ
    2. Kim JH
    3. Ress D

    (2018) Characterization of the hemodynamic response function across the majority of human cerebral cortex

    NeuroImage 173:322–331.

    https://doi.org/10.1016/j.neuroimage.2018.02.061

    • PubMed
    • Google Scholar
95. 1. Uludağ K
    2. Blinder P

    (2018) Linking brain vascular physiology to hemodynamic response in ultra-high field MRI

    NeuroImage 168:279–295.

    https://doi.org/10.1016/j.neuroimage.2017.02.063

    • Google Scholar
96. 1. van der Kouwe AJW
    2. Benner T
    3. Salat DH
    4. Fischl B

    (2008) Brain morphometry with multiecho MPRAGE

    NeuroImage 40:559–569.

    https://doi.org/10.1016/j.neuroimage.2007.12.025

    • PubMed
    • Google Scholar
97. 1. Vizioli L
    2. Bratch A
    3. Lao J
    4. Ugurbil K
    5. Muckli L
    6. Yacoub E

    (2018) Temporal multivariate pattern analysis (tMVPA): A single trial approach exploring the temporal dynamics of the BOLD signal

    Journal of Neuroscience Methods 308:74–87.

    https://doi.org/10.1016/j.jneumeth.2018.06.029

    • PubMed
    • Google Scholar
98. 1. Wang L
    2. Kong Q
    3. Li K
    4. Su Y
    5. Zeng Y
    6. Zhang Q
    7. Dai W
    8. Xia M
    9. Wang G
    10. Jin Z
    11. Yu X
    12. Si T

    (2016) Frequency-dependent changes in amplitude of low-frequency oscillations in depression: A resting-state fMRI study

    Neuroscience Letters 614:105–111.

    https://doi.org/10.1016/j.neulet.2016.01.012

    • Google Scholar
99. 1. West KL
    2. Zuppichini MD
    3. Turner MP
    4. Sivakolundu DK
    5. Zhao Y
    6. Abdelkarim D
    7. Spence JS
    8. Rypma B

    (2019) BOLD hemodynamic response function changes significantly with healthy aging

    NeuroImage 188:198–207.

    https://doi.org/10.1016/j.neuroimage.2018.12.012

    • PubMed
    • Google Scholar
100. 1. Williams SD
     2. Setzer B
     3. Fultz NE
     4. Valdiviezo Z
     5. Tacugue N
     6. Diamandis Z
     7. Lewis LD

     (2023) Correction: Neural activity induced by sensory stimulation can drive large-scale cerebrospinal fluid flow during wakefulness in humans

     PLOS Biology 21:e3002035.

     https://doi.org/10.1371/journal.pbio.3002035

     • Google Scholar
101. 1. Wittkuhn L
     2. Schuck NW

     (2021) Dynamics of fMRI patterns reflect sub-second activation sequences and reveal replay in human visual cortex

     Nature Communications 12:1795.

     https://doi.org/10.1038/s41467-021-21970-2

     • PubMed
     • Google Scholar
102. 1. Wu G-R
     2. Liao W
     3. Stramaglia S
     4. Ding J-R
     5. Chen H
     6. Marinazzo D

     (2013) A blind deconvolution approach to recover effective connectivity brain networks from resting state fMRI data

     Medical Image Analysis 17:365–374.

     https://doi.org/10.1016/j.media.2013.01.003

     • PubMed
     • Google Scholar
103. 1. Wu G-R
     2. Colenbier N
     3. Van Den Bossche S
     4. Clauw K
     5. Johri A
     6. Tandon M
     7. Marinazzo D

     (2021) rsHRF: A toolbox for resting-state HRF estimation and deconvolution

     NeuroImage 244:118591.

     https://doi.org/10.1016/j.neuroimage.2021.118591

     • Google Scholar
104. 1. Yang L
     2. Yan Y
     3. Li Y
     4. Hu X
     5. Lu J
     6. Chan P
     7. Yan T
     8. Han Y

     (2020) Frequency-dependent changes in fractional amplitude of low-frequency oscillations in Alzheimer’s disease: a resting-state fMRI study

     Brain Imaging and Behavior 14:2187–2201.

     https://doi.org/10.1007/s11682-019-00169-6

     • PubMed
     • Google Scholar
105. 1. Yen CCC
     2. Fukuda M
     3. Kim SG

     (2011) BOLD responses to different temporal frequency stimuli in the lateral geniculate nucleus and visual cortex: insights into the neural basis of fMRI

     NeuroImage 58:82–90.

     https://doi.org/10.1016/j.neuroimage.2011.06.022

     • PubMed
     • Google Scholar
106. 1. Zang Y-F
     2. He Y
     3. Zhu C-Z
     4. Cao Q-J
     5. Sui M-Q
     6. Liang M
     7. Tian L-X
     8. Jiang T-Z
     9. Wang Y-F

     (2007) Altered baseline brain activity in children with ADHD revealed by resting-state functional MRI

     Brain & Development 29:83–91.

     https://doi.org/10.1016/j.braindev.2006.07.002

     • PubMed
     • Google Scholar
107. 1. Zhu X
     2. Helpman L
     3. Papini S
     4. Schneier F
     5. Markowitz JC
     6. Van Meter PE
     7. Lindquist MA
     8. Wager TD
     9. Neria Y

     (2017) Altered resting state functional connectivity of fear and reward circuitry in comorbid PTSD and major depression

     Depression and Anxiety 34:641–650.

     https://doi.org/10.1002/da.22594

     • PubMed
     • Google Scholar
108. 1. Zou Q-H
     2. Zhu C-Z
     3. Yang Y
     4. Zuo X-N
     5. Long X-Y
     6. Cao Q-J
     7. Wang Y-F
     8. Zang Y-F

     (2008) An improved approach to detection of amplitude of low-frequency fluctuation (ALFF) for resting-state fMRI: fractional ALFF

     Journal of Neuroscience Methods 172:137–141.

     https://doi.org/10.1016/j.jneumeth.2008.04.012

     • PubMed
     • Google Scholar

Author details

1. Sydney M Bailes

   Department of Biomedical Engineering, Boston University, Boston, United States

   Contribution

   Conceptualization, Data curation, Software, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-1712-9233

2. Daniel EP Gomez

   1. Department of Biomedical Engineering, Boston University, Boston, United States
   2. Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, United States
   3. Department of Radiology, Harvard Medical School, Boston, United States

   Contribution

   Software, Investigation, Writing - review and editing

   Competing interests

   No competing interests declared

3. Beverly Setzer

   1. Department of Biomedical Engineering, Boston University, Boston, United States
   2. Graduate Program for Neuroscience, Boston University, Boston, United States

   Contribution

   Software, Investigation, Writing - review and editing

   Competing interests

   No competing interests declared

4. Laura D Lewis

   1. Department of Biomedical Engineering, Boston University, Boston, United States
   2. Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, United States
   3. Institute for Medical Engineering and Science, Massachusetts Institute of Technology, Cambridge, United States
   4. Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, United States

   Contribution

   Conceptualization, Resources, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Project administration, Writing - review and editing

   For correspondence

   ldlewis@mit.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-4003-0277

• 1,263

  views

• 157

  downloads

• 6

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.