Sensitivity and specificity considerations for fMRI encoding, decoding, and mapping of auditory cortex at ultra-high field

Linguistic modulation of the neural encoding of phonemes

Speech comprehension entails the neural mapping of the acoustic speech signal onto learned linguistic units. This acousto-linguistic transformation is bi-directional, whereby higher-level linguistic processes (e.g. semantics) modulate the acoustic analysis of individual linguistic units. Here, we investigated the cortical topography and linguistic modulation of the most fundamental linguistic unit, the phoneme. We presented natural speech and "phoneme quilts" (pseudo-randomly shuffled phonemes) in either a familiar (English) or unfamiliar (Korean) language to native English speakers while recording functional magnetic resonance imaging. This allowed us to dissociate the contribution of acoustic vs. linguistic processes toward phoneme analysis. We show that (i) the acoustic analysis of phonemes is modulated by linguistic analysis and (ii) that for this modulation, both of acoustic and phonetic information need to be incorporated. These results suggest that the linguistic modulation of cortical sensitivity to phoneme classes minimizes prediction error during natural speech perception, thereby aiding speech comprehension in challenging listening situations.

Characterisation of laminar and vascular spatiotemporal dynamics of CBV and BOLD signals using VASO and ME-GRE at 7T in humans

Interpretation of cortical laminar functional magnetic resonance imaging (fMRI) activity requires detailed knowledge of the spatiotemporal haemodynamic response across vascular compartments due to the well-known vascular biases (e.g. the draining veins). Further complications arise from the spatiotemporal hemodynamic response that differs depending on the duration of stimulation. This information is crucial for future studies using depth-dependent cerebral blood volume (CBV) measurements, which promise higher specificity for the cortical microvasculature than the blood oxygenation level dependent (BOLD) contrast. To date, direct information about CBV dynamics with respect to stimulus duration, cortical depth and vasculature is missing in humans. Therefore, we characterized the cortical depth-dependent CBV-haemodynamic responses across a wide set of stimulus durations with 0.9 mm isotropic spatial and 0.785 seconds effective temporal resolution in humans using slice-selective slab-inversion vascular space occupancy (SS-SI VASO). Additionally, we investigated signal contributions from macrovascular compartments using fine-scale vascular information from multi-echo gradient-echo (ME-GRE) data at 0.35 mm isotropic resolution. In total, this resulted in >7.5h of scanning per participant (n=5). We have three major findings: (I) While we could demonstrate that 1 second stimulation is viable using VASO, more than 12 seconds stimulation provides better CBV responses in terms of specificity to microvasculature, but durations beyond 24 seconds of stimulation may be wasteful for certain applications. (II) We observe that CBV responses show dilation patterns across the cortex. (III) While we found increasingly strong BOLD signal responses in vessel-dominated voxels with longer stimulation durations, we found increasingly strong CBV signal responses in vessel-dominated voxels only until 4 second stimulation durations. After 4 seconds, only the signal from non-vessel dominated voxels kept increasing. This might explain why CBV responses are more specific to the underlying neuronal activity for long stimulus durations.

Evaluating the effect of denoising submillimeter auditory fMRI data with NORDIC

Functional magnetic resonance imaging (fMRI) has emerged as an essential tool for exploring human brain function. Submillimeter fMRI, in particular, has emerged as a tool to study mesoscopic computations. The inherently low signal-to-noise ratio (SNR) at submillimeter resolutions warrants the use of denoising approaches tailored at reducing thermal noise - the dominant contributing noise component in high resolution fMRI. NORDIC PCA is one of such approaches, and has been benchmarked against other approaches in several applications. Here, we investigate the effects that two versions of NORDIC denoising have on auditory submillimeter data. As investigating auditory functional responses poses unique challenges, we anticipated that the benefit of this technique would be especially pronounced. Our results show that NORDIC denoising improves the detection sensitivity and the reliability of estimates in submillimeter auditory fMRI data. These effects can be explained by the reduction of the noise-induced signal variability. However, we also observed a reduction in the average response amplitude (percent signal), which may suggest that a small amount of signal was also removed. We conclude that, while evaluating the effects of the signal reduction induced by NORDIC may be necessary for each application, using NORDIC in high resolution auditory fMRI studies may be advantageous because of the large reduction in variability of the estimated responses.

Microvascular Specificity of Spin Echo BOLD fMRI: Impact of EPI Echo Train Length

A spatially specific fMRI acquisition requires specificity to the microvasculature that serves active neuronal sites. Macrovascular contributions will reduce the microvascular specificity but can be reduced by using spin echo (SE) sequences that use a π pulse to refocus static field inhomogeneities near large veins. The microvascular specificity of a SE-echo planar imaging (SE-EPI) scan depends on the echo train length (ETL)-duration, but the dependence is not well-characterized in humans at 7T. To determine how microvascular-specific SE-EPI BOLD is in humans at 7T, we developed a Monte Carlo voxel model that computes the signal of a proton ensemble residing in a vasculature subjected to a SE-EPI pulse sequence. We characterized the ETL-duration dependence of the microvascular specificity by simulating the BOLD signal as a function of ETL, the range adhering to experimentally realistic readouts. We performed a validation experiment for our simulation observations, in which we acquired a set of SE-EPI BOLD time series with varying ETL during a hyperoxic gas challenge. Both our simulations and measurements show an increase in macrovascular contamination as a function of ETL, with an increase of 30% according to our simulation and 60% according to our validation experiment between the shortest and longest ETL durations (23.1 - 49.7 ms). We conclude that the microvascular specificity decreases heavily with increasing ETL-durations. We recommend reducing the ETL-duration as much as possible to minimize macrovascular contamination in SE-EPI BOLD experiments. We additionally recommend scanning at high resolutions to minimize partial volume effects with CSF. CSF voxels show a large BOLD response, which can be attributed to both the presence of large veins (high blood volume) and molecular oxygen-induced T1-shortening (significant in a hyperoxia experiment). The magnified BOLD signal in a GM-CSF partial volume voxel reduces the desired microvascular specificity and, therefore, will hinder the interpretation of functional MRI activation patterns.

Stacked regressions and structured variance partitioning for interpretable brain maps

Relating brain activity associated with a complex stimulus to different properties of that stimulus is a powerful approach for constructing functional brain maps. However, when stimuli are naturalistic, their properties are often correlated (e.g., visual and semantic features of natural images, or different layers of a convolutional neural network that are used as features of images). Correlated properties can act as confounders for each other and complicate the interpretability of brain maps, and can impact the robustness of statistical estimators. Here, we present an approach for brain mapping based on two proposed methods: stacking different encoding models and structured variance partitioning. Our stacking algorithm combines encoding models that each use as input a feature space that describes a different stimulus attribute. The algorithm learns to predict the activity of a voxel as a linear combination of the outputs of different encoding models. We show that the resulting combined model can predict held-out brain activity better or at least as well as the individual encoding models. Further, the weights of the linear combination are readily interpretable; they show the importance of each feature space for predicting a voxel. We then build on our stacking models to introduce structured variance partitioning, a new type of variance partitioning that takes into account the known relationships between features. Our approach constrains the size of the hypothesis space and allows us to ask targeted questions about the similarity between feature spaces and brain regions even in the presence of correlations between the feature spaces. We validate our approach in simulation, showcase its brain mapping potential on fMRI data, and release a Python package. Our methods can be useful for researchers interested in aligning brain activity with different layers of a neural network, or with other types of correlated feature spaces.

Combining arterial blood contrast with BOLD increases fMRI intracortical contrast

BOLD fMRI is widely applied in human neuroscience but is limited in its spatial specificity due to a cortical-depth-dependent venous bias. This reduces its localization specificity with respect to neuronal responses, a disadvantage for neuroscientific research. Here, we modified a submillimeter BOLD protocol to selectively reduce venous and tissue signal and increase cerebral blood volume weighting through a pulsed saturation scheme (dubbed Arterial Blood Contrast) at 7 T. Adding Arterial Blood Contrast on top of the existing BOLD contrast modulated the intracortical contrast. Isolating the Arterial Blood Contrast showed a response free of pial-surface bias. The results suggest that Arterial Blood Contrast can modulate the typical fMRI spatial specificity, with important applications in in-vivo neuroscience.

Improved sensitivity and microvascular weighting of 3T laminar fMRI with GE-BOLD using NORDIC and phase regression

INTRODUCTION

Functional MRI with spatial resolution in the submillimeter domain enables measurements of activation across cortical layers in humans. This is valuable as different types of cortical computations, e.g., feedforward versus feedback related activity, take place in different cortical layers. Laminar fMRI studies have almost exclusively employed 7T scanners to overcome the reduced signal stability associated with small voxels. However, such systems are relatively rare and only a subset of those are clinically approved. In the present study, we examined if the feasibility of laminar fMRI at 3T could be improved by use of NORDIC denoising and phase regression.

METHODS

5 healthy subjects were scanned on a Siemens MAGNETOM Prisma 3T scanner. To assess across-session reliability, each subject was scanned in 3-8 sessions on 3-4 consecutive days. A 3D gradient echo EPI (GE-EPI) sequence was used for BOLD acquisitions (voxel size 0.82 mm isotopic, TR = 2.2 s) using a block design finger tapping paradigm. NORDIC denoising was applied to the magnitude and phase time series to overcome limitations in temporal signal-to-noise ratio (tSNR) and the denoised phase time series were subsequently used to correct for large vein contamination through phase regression.

RESULTS AND CONCLUSION

NORDIC denoising resulted in tSNR values comparable to or higher than commonly observed at 7T. Layer-dependent activation profiles could thus be extracted robustly, within and across sessions, from regions of interest located in the hand knob of the primary motor cortex (M1). Phase regression led to substantially reduced superficial bias in obtained layer profiles, although residual macrovascular contribution remained. We believe the present results support an improved feasibility of laminar fMRI at 3T.

Cerebral blood volume sensitive layer-fMRI in the human auditory cortex at 7T: Challenges and capabilities

The development of ultra high field fMRI signal readout strategies and contrasts has led to the possibility of imaging the human brain in vivo and non-invasively at increasingly higher spatial resolutions of cortical layers and columns. One emergent layer-fMRI acquisition method with increasing popularity is the cerebral blood volume sensitive sequence named vascular space occupancy (VASO). This approach has been shown to be mostly sensitive to locally-specific changes of laminar microvasculature, without unwanted biases of trans-laminar draining veins. Until now, however, VASO has not been applied in the technically challenging cortical area of the auditory cortex. Here, we describe the main challenges we encountered when developing a VASO protocol for auditory neuroscientific applications and the solutions we have adopted. With the resulting protocol, we present preliminary results of laminar responses to sounds and as a proof of concept for future investigations, we map the topographic representation of frequency preference (tonotopy) in the auditory cortex.

Towards functional spin-echo BOLD line-scanning in humans at 7T

OBJECTIVE

Neurons cluster into sub-millimeter spatial structures and neural activity occurs at millisecond resolutions; hence, ultimately, high spatial and high temporal resolutions are required for functional MRI. In this work, we implemented a spin-echo line-scanning (SELINE) sequence to use in high spatial and temporal resolution fMRI.

MATERIALS AND METHODS

A line is formed by simply rotating the spin-echo refocusing gradient to a plane perpendicular to the excited slice and by removing the phase-encoding gradient. This technique promises a combination of high spatial and temporal resolution (250 μm, 500 ms) and microvascular specificity of functional responses. We compared SELINE data to a corresponding gradient-echo version (GELINE).

RESULTS

We demonstrate that SELINE showed much-improved line selection (i.e. a sharper line profile) compared to GELINE, albeit at the cost of a significant drop in functional sensitivity.

DISCUSSION

This low functional sensitivity needs to be addressed before SELINE can be applied for neuroscientific purposes.

Mesoscopic in vivo human T2* dataset acquired using quantitative MRI at 7 Tesla

Mesoscopic (0.1-0.5 mm) interrogation of the living human brain is critical for advancing neuroscience and bridging the resolution gap with animal models. Despite the variety of MRI contrasts measured in recent years at the mesoscopic scale, in vivo quantitative imaging of T₂^* has not been performed. Here we provide a dataset containing empirical T₂^* measurements acquired at 0.35 × 0.35 × 0.35 mm³ voxel resolution using 7 Tesla MRI. To demonstrate unique features and high quality of this dataset, we generate flat map visualizations that reveal fine-scale cortical substructures such as layers and vessels, and we report quantitative depth-dependent T₂^* (as well as R₂^*) values in primary visual cortex and auditory cortex that are highly consistent across subjects. This dataset is freely available at https://doi.org/10.17605/OSF.IO/N5BJ7, and may prove useful for anatomical investigations of the human brain, as well as for improving our understanding of the basis of the T₂^*-weighted (f)MRI signal.

On the encoding of natural music in computational models and human brains

This article discusses recent developments and advances in the neuroscience of music to understand the nature of musical emotion. In particular, it highlights how system identification techniques and computational models of music have advanced our understanding of how the human brain processes the textures and structures of music and how the processed information evokes emotions. Musical models relate physical properties of stimuli to internal representations called features, and predictive models relate features to neural or behavioral responses and test their predictions against independent unseen data. The new frameworks do not require orthogonalized stimuli in controlled experiments to establish reproducible knowledge, which has opened up a new wave of naturalistic neuroscience. The current review focuses on how this trend has transformed the domain of the neuroscience of music.

Using high spatial resolution fMRI to understand representation in the auditory network

Following rapid methodological advances, ultra-high field (UHF) functional and anatomical magnetic resonance imaging (MRI) has been repeatedly and successfully used for the investigation of the human auditory system in recent years. Here, we review this work and argue that UHF MRI is uniquely suited to shed light on how sounds are represented throughout the network of auditory brain regions. That is, the provided gain in spatial resolution at UHF can be used to study the functional role of the small subcortical auditory processing stages and details of cortical processing. Further, by combining high spatial resolution with the versatility of MRI contrasts, UHF MRI has the potential to localize the primary auditory cortex in individual hemispheres. This is a prerequisite to study how sound representation in higher-level auditory cortex evolves from that in early (primary) auditory cortex. Finally, the access to independent signals across auditory cortical depths, as afforded by UHF, may reveal the computations that underlie the emergence of an abstract, categorical sound representation based on low-level acoustic feature processing. Efforts on these research topics are underway. Here we discuss promises as well as challenges that come with studying these research questions using UHF MRI, and provide a future outlook.

Investigating mechanisms of fast BOLD responses: The effects of stimulus intensity and of spatial heterogeneity of hemodynamics

Recent studies have demonstrated that fast fMRI can track neural activity well above the temporal limit predicted by the canonical hemodynamic response model. While these findings are promising, the biophysical mechanisms underlying these fast fMRI phenomena remain underexplored. In this study, we discuss two aspects of the hemodynamic response, complementary to several existing hypotheses, that can accommodate faster fMRI dynamics beyond those predicted by the canonical model. First, we demonstrate, using both visual and somatosensory paradigms, that the timing and shape of hemodynamic response functions (HRFs) vary across graded levels of stimulus intensity-with lower-intensity stimulation eliciting faster and narrower HRFs. Second, we show that as the spatial resolution of fMRI increases, voxel-wise HRFs begin to deviate from the canonical model, with a considerable portion of voxels exhibiting faster temporal dynamics than predicted by the canonical HRF. Collectively, both stimulus/task intensity and image resolution can affect the sensitivity of fMRI to fast brain activity, which may partly explain recent observations of fast fMRI signals. It is further noteworthy that, while the present investigations focus on fast neural responses, our findings suggest that a revised hemodynamic model may benefit the many fMRI studies using paradigms with wide ranges of contrast levels (e.g., resting or naturalistic conditions) or with modern, high-resolution MR acquisitions.

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

Fast fMRI enables the detection of neural dynamics over timescales of hundreds of milliseconds, suggesting it may provide a new avenue for studying subsecond neural processes in the human brain. The magnitudes of these fast fMRI dynamics are far greater than predicted by canonical models of the hemodynamic response. Several studies have established nonlinear properties of the hemodynamic response that have significant implications for fast fMRI. We first review nonlinear properties of the hemodynamic response function that may underlie fast fMRI signals. We then illustrate the breakdown of canonical hemodynamic response models in the context of fast neural dynamics. We will then argue that the canonical hemodynamic response function is not likely to reflect the BOLD response to neuronal activity driven by sparse or naturalistic stimuli or perhaps to spontaneous neuronal fluctuations in the resting state. These properties suggest that fast fMRI is capable of tracking surprisingly fast neuronal dynamics, and we discuss the neuroscientific questions that could be addressed using this approach.

LayNii: A software suite for layer-fMRI

High-resolution fMRI in the sub-millimeter regime allows researchers to resolve brain activity across cortical layers and columns non-invasively. While these high-resolution data make it possible to address novel questions of directional information flow within and across brain circuits, the corresponding data analyses are challenged by MRI artifacts, including image blurring, image distortions, low SNR, and restricted coverage. These challenges often result in insufficient spatial accuracy of conventional analysis pipelines. Here we introduce a new software suite that is specifically designed for layer-specific functional MRI: LayNii. This toolbox is a collection of command-line executable programs written in C/C++ and is distributed opensource and as pre-compiled binaries for Linux, Windows, and macOS. LayNii is designed for layer-fMRI data that suffer from SNR and coverage constraints and thus cannot be straightforwardly analyzed in alternative software packages. Some of the most popular programs of LayNii contain 'layerification' and columnarization in the native voxel space of functional data as well as many other layer-fMRI specific analysis tasks: layer-specific smoothing, model-based vein mitigation of GE-BOLD data, quality assessment of artifact dominated sub-millimeter fMRI, as well as analyses of VASO data.

Functional neuroimaging in hearing research and audiology

The various methods of medical imaging are essential for many diagnostic issues in clinical routine, e.g., for the diagnostics and localisation of tumorous diseases, or for the clarification of other lesions in the central nervous system. In addition to these classical roles both positron emission tomography (PET) and magnetic resonance imaging (MRI) allow for the investigation of functional processes in the human brain, when used in a specific way. The last 25 years have seen great progress, especially with respect to functional MRI, in terms of the available experimental paradigms as well as the data analysis strategies, so that a directed investigation of neurophysiological correlates of psychoacoustic performance is possible. This covers fundamental measures of sound perception like loudness and pitch, specific audiological symptoms like tinnitus, which often accompanies hearing disorders, but it also includes experiments on speech perception or on virtual acoustic environments. One important aspect common to many auditory neuroimaging studies is the central question at what stage in the human auditory pathway the sensory coding of the incoming sound is transformed into a universal and context-dependent perceptual representation, which is the basis for what we hear. This overview summarises findings from the literature as well as a few studies from our lab, to discuss the possibilities and the limits of the adoption of functional neuroimaging methods in audiology. Up to this stage, most auditory neuroimaging studies have investigated basic processes in normal hearing listeners. However, the hitherto existing results suggest that the methods of auditory functional neuroimaging - possibly complemented by electrophysiological methods like EEG and MEG - have a great potential to contribute to a deeper understanding of the processes and the impact of hearing disorders.

Determining laminar neuronal activity from BOLD fMRI using a generative model

Laminar fMRI using the BOLD contrast enables the non-invasive investigation of mesoscopic functional circuits in the human brain. However, the laminar neuronal activity is spatiotemporally biased in the observed cortical depth profiles of the BOLD signal. In this study, we propose a generative fMRI signal model, comprehensively covering the relationship between cortical depth-dependent changes in excitatory and inhibitory neuronal activity with the sampling of the BOLD signal with finite voxels. The generative model allowed us to investigate pertinent questions regarding the accuracy of the laminar BOLD signal relative to the neuronal activity, and we found that: a) condition differences in laminar BOLD signals may be more reflective of neuronal activity than single condition BOLD signal depth profiles; b) angular dependence of the BOLD signal induces significant signal variability, which can mask underlying activity profiles; c) even if only three neuronal depths are of interest, more BOLD signal depths should be considered in the analysis. In addition, we recommend that the laminar BOLD data should be displayed using the centroid method to appreciate its spatial distribution in the original resolution. Finally, we showed that Bayesian model inversion of the generative model can improve sensitivity and specificity of assessing depth-dependent neuronal changes both for steady-state and dynamically.

High-Resolution Neurovascular Imaging at 7T: Arterial Spin Labeling Perfusion, 4-Dimensional MR Angiography, and Black Blood MR Imaging

Ultrahigh field offers increased resolution and contrast for neurovascular imaging. Arterial spin labeling methods benefit from an increased intrinsic signal-to-noise ratio of MR imaging signal and a prolonged tracer half-life at ultrahigh field, allowing the visualization of layer-dependent microvascular perfusion. Arterial spin labeling-based time-resolved 4-dimensional MR angiography at 7T provides a detailed depiction of the vascular architecture and dynamic blood flow pattern with high spatial and temporal resolutions. High-resolution black blood MR imaging at 7T allows detailed characterization of small perforating arteries such as lenticulostriate arteries. All techniques benefit from advances in parallel radiofrequency transmission technologies at ultrahigh field.

Zooming-in on higher-level vision: High-resolution fMRI for understanding visual perception and awareness

One of the central questions in visual neuroscience is how the sparse retinal signals leaving our eyes are transformed into a rich subjective visual experience of the world. Invasive physiology studies, which offers the highest spatial resolution, have revealed many facts about the processing of simple visual features like contrast, color, and orientation, focusing on the early visual areas. At the same time, standard human fMRI studies with comparably coarser spatial resolution have revealed more complex, functionally specialized, and category-selective responses in higher visual areas. Although the visual system is the best understood among the sensory modalities, these two areas of research remain largely segregated. High-resolution fMRI opens up a possibility for linking them. On the one hand, it allows studying how the higher-level visual functions affect the fine-scale activity in early visual areas. On the other hand, it allows discovering the fine-scale functional organization of higher visual areas and exploring their functional connectivity with visual areas lower in the hierarchy. In this review, I will discuss examples of successful work undertaken in these directions using high-resolution fMRI and discuss where this method could be applied in the future to advance our understanding of the complexity of higher-level visual processing.

Effects of phase regression on high-resolution functional MRI of the primary visual cortex

High-resolution functional MRI studies have become a powerful tool to non-invasively probe the sub-millimeter functional organization of the human cortex. Advances in MR hardware, imaging techniques and sophisticated post-processing methods have allowed high resolution fMRI to be used in both the clinical and academic neurosciences. However, consensus within the community regarding the use of gradient echo (GE) or spin echo (SE) based acquisition remains largely divided. On one hand, GE provides a high temporal signal-to-noise ratio (tSNR) technique sensitive to both the macro- and micro-vascular signal while SE based methods are more specific to microvasculature but suffer from lower tSNR and specific absorption rate limitations, especially at high field and with short repetition times. Fortunately, the phase of the GE-EPI signal is sensitive to vessel size and this provides a potential avenue to reduce the macrovascular weighting of the signal (phase regression, Menon 2002). In order to determine the efficacy of this technique at high-resolution, phase regression was applied to GE-EPI timeseries and compared to SE-EPI to determine if GE-EPI's specificity to the microvascular compartment improved. To do this, functional data was collected from seven subjects on a neuro-optimized 7 T system at 800 μm isotropic resolution with both GE-EPI and SE-EPI while observing an 8 Hz contrast reversing checkerboard. Phase data from the GE-EPI was used to create a microvasculature-weighted time series (GE-EPI-PR). Anatomical imaging (MP2RAGE) was also collected to allow for surface segmentation so that the functional results could be projected onto a surface. A multi-echo gradient echo sequence was collected and used to identify venous vasculature. The GE-EPI-PR surface activation maps showed a high qualitative similarity with SE-EPI and also produced laminar activity profiles similar to SE-EPI. When the GE-EPI and GE-EPI-PR distributions were compared to SE-EPI it was shown that GE-EPI-PR had similar distribution characteristics to SE-EPI (p < 0.05) across the top 60% of cortex. Furthermore, it was shown that GE-EPI-PR has a higher contrast-to-noise ratio (0.5 ± 0.2, mean ± std. dev. across layers) than SE-EPI (0.27 ± 0.07) demonstrating the technique has higher sensitivity than SE-EPI. Taken together this evidence suggests phase regression is a useful method in low SNR studies such as high-resolution fMRI.

New acquisition techniques and their prospects for the achievable resolution of fMRI

This work reviews recent advances in technologies for functional magnetic resonance imaging (fMRI) of the human brain and highlights the push for higher functional specificity based on increased spatial resolution and specific MR contrasts to reveal previously undetectable functional properties of small-scale cortical structures. We discuss how the combination of MR hardware, advanced acquisition techniques and various MR contrast mechanisms have enabled recent progress in functional neuroimaging. However, these advanced fMRI practices have only been applied to a handful of neuroscience questions to date, with the majority of the neuroscience community still using conventional imaging techniques. We thus discuss upcoming challenges and possibilities for fMRI technology development in human neuroscience. We hope that readers interested in functional brain imaging acquire an understanding of current and novel developments and potential future applications, even if they don't have a background in MR physics or engineering. We summarize the capabilities of standard fMRI acquisition schemes with pointers to relevant literature and comprehensive reviews and introduce more recent developments.

A temporal decomposition method for identifying venous effects in task-based fMRI

The spatial resolution of functional magnetic resonance imaging (fMRI) is fundamentally limited by effects from large draining veins. Here we describe an analysis method that provides data-driven estimates of these effects in task-based fMRI. The method involves fitting a one-dimensional manifold that characterizes variation in response timecourses observed in a given dataset, and then using identified early and late timecourses as basis functions for decomposing responses into components related to the microvasculature (capillaries and small venules) and the macrovasculature (large veins), respectively. We show the removal of late components substantially reduces the superficial cortical depth bias of fMRI responses and helps eliminate artifacts in cortical activity maps. This method provides insight into the origins of the fMRI signal and can be used to improve the spatial accuracy of fMRI.

Comparison of BOLD and CBV using 3D EPI and 3D GRASE for cortical layer functional MRI at 7 T

PURPOSE

Functional MRI (fMRI) at the mesoscale of cortical layers and columns requires both sensitivity and specificity, the latter of which can be compromised if the imaging method is affected by vascular artifacts, particularly cortical draining veins at the pial surface. Recent studies have shown that cerebral blood volume (CBV) imaging is more specific to the actual laminar locus of neural activity than BOLD imaging using standard gradient-echo EPI sequences. Gradient and spin-echo (GRASE) BOLD imaging has also shown greater specificity when compared with standard gradient-echo EPI BOLD. Here we directly compare CBV and BOLD contrasts in high-resolution imaging of the primary motor cortex for laminar functional MRI in four combinations of signal labeling, CBV using slice-selective slab-inversion vascular space occupancy (VASO) and BOLD, each with 3D gradient-echo EPI and zoomed 3D-GRASE image readouts.

METHODS

Activations were measured using each sequence and contrast combination during a motor task. Activation profiles across cortical depth were measured to assess the sensitivity and specificity (pial bias) of each method.

RESULTS

Both CBV imaging using gradient-echo 3D-EPI and BOLD imaging using 3D-GRASE show similar specificity and sensitivity and are therefore useful tools for mesoscopic functional MRI in the human cortex. The combination of GRASE and VASO did not demonstrate high levels of sensitivity, nor show increased specificity.

CONCLUSION

Three-dimensional EPI with VASO contrast and 3D-GRASE with BOLD contrast both demonstrate sufficient sensitivity and specificity for laminar functional MRI to be used by neuroscientists in a wide range of investigations of depth-dependent neural circuitry in the human brain.

A probabilistic atlas of finger dominance in the primary somatosensory cortex

With the advent of ultra-high field (7T), high spatial resolution functional MRI (fMRI) has allowed the differentiation of the cortical representations of each of the digits at an individual-subject level in human primary somatosensory cortex (S1). Here we generate a probabilistic atlas of the contralateral SI representations of the digits of both the left and right hand in a group of 22 right-handed individuals. The atlas is generated in both volume and surface standardised spaces from somatotopic maps obtained by delivering vibrotactile stimulation to each distal phalangeal digit using a travelling wave paradigm. Metrics quantify the likelihood of a given position being assigned to a digit (full probability map) and the most probable digit for a given spatial location (maximum probability map). The atlas is validated using a leave-one-out cross validation procedure. Anatomical variance across the somatotopic map is also assessed to investigate whether the functional variability across subjects is coupled to structural differences. This probabilistic atlas quantifies the variability in digit representations in healthy subjects, finding some quantifiable separability between digits 2, 3 and 4, a complex overlapping relationship between digits 1 and 2, and little agreement of digit 5 across subjects. The atlas and constituent subject maps are available online for use as a reference in future neuroimaging studies.

Cortical Activation Patterns of Different Masking Noises and Correlation With Their Masking Efficacy, Determined by Functional Near-Infrared Spectroscopy

Acoustic therapy in tinnitus treatment is poorly characterized, and efficacy assessment depends on subjective descriptions. Narrow-band noise, notched sound, and white noise have positive therapeutic effects on monotonous tinnitus. Considering the tonotopic characteristics of the auditory system and the spectral characteristics of these three masking sounds, the activation pattern of the auditory cortex and the mechanism of inhibiting tinnitus may be different. This study aimed to compare the activation patterns of three spectrally different masking noises and study the correlation between the masking effects and variational amplitude of oxygenated hemoglobin (HbO) in the corresponding cortical regions. We also assessed near-infrared spectroscopy brain function imaging (NIRS) as an objective assessment tool in acoustic therapy. Patients with persistent non-pulsatile tinnitus and control volunteers without tinnitus were enrolled in this study. The subjects were seated in a sound-proof room, with two optode arrays covering the bilateral temporal lobe. Auditory stimuli were presented; stimulation sequences followed the block design: different noises appeared randomly and repeated in five cycles. Tinnitus match and residual inhibition were performed in the tinnitus group. The data analyses were conducted using the NIRS_SPM toolbox. The group analysis results showed that the narrow-band noise caused a marginally significant decrease in HbO signal in the Brodmann 21 region (BA21), while white noise caused a significant increase in HbO signal in BA21. Notched sound did not cause significant changes in the HbO signal in the temporal cortex. And none of the three masking noises caused significant changes in the HbR signal in the temporal cortex. The depth of residual inhibition induced by the narrow-band noise and white noise significantly correlated with ΔHbO in the region of interest (ROI). However, neither the depth nor duration of the residual inhibition induced by notched sound correlated with the ΔHbO. Thus, NIRS showed three cortical activation patterns induced by three different masking noises, and correlations between residual inhibition effects and change of HbO amplitude were found. NIRS could therefore be applied in objective assessment of acoustic therapy.

Four new cytoarchitectonic areas surrounding the primary and early auditory cortex in human brains

The architectonical organization of putatively higher auditory areas in the human superior temporal gyrus and sulcus is not yet well understood. To provide a coherent map of this part of the brain, which is involved in language and other functions, we examined the cytoarchitecture and cortical parcellation of this region in histological sections of ten human postmortem brains using an observer-independent mapping approach. Two new areas were identified in the temporo-insular region (areas TeI, TI). TeI is medially adjacent to the primary auditory cortex (area Te1). TI is located between TeI and the insular cortex. Laterally adjacent to previously mapped areas Te2 and Te3, two new areas (STS1, STS2) were identified in the superior temporal sulcus. All four areas were mapped over their whole extent in serial, cell-body stained sections, and their cytoarchitecture was analyzed using quantitative image analysis and multivariate statistics. Interestingly, area TeI, which is located between area Te1 and area TI at the transition to the insula, was more similar in cytoarchitecture to lateral area Te2.1 than to the directly adjacent areas TI and Te1. Such structural similarity of areas medially and laterally to Te1 would be in line with the core-belt-parabelt concept in macaques. The cytoarchitectonic probabilistic maps of all areas show the localization of the areas and their interindividual variability. The new maps are publicly available and provide a basis to further explore structural-functional relationship of the language network in the temporal cortex.

Early Blindness Shapes Cortical Representations of Auditory Frequency within Auditory Cortex

Early loss of vision is classically linked to large-scale cross-modal plasticity within occipital cortex. Much less is known about the effects of early blindness on auditory cortex. Here, we examine the effects of early blindness on the cortical representation of auditory frequency within human primary and secondary auditory areas using fMRI. We observe that 4 individuals with early blindness (2 females), and a group of 5 individuals with anophthalmia (1 female), a condition in which both eyes fail to develop, have lower response amplitudes and narrower voxelwise tuning bandwidths compared with a group of typically sighted individuals. These results provide some of the first evidence in human participants for compensatory plasticity within nondeprived sensory areas as a result of sensory loss.SIGNIFICANCE STATEMENT Early blindness has been linked to enhanced perception of the auditory world, including auditory localization and pitch perception. Here we used fMRI to compare neural responses with auditory stimuli within auditory cortex across sighted, early blind, and anophthalmic individuals, in whom both eyes fail to develop. We find more refined frequency tuning in blind subjects, providing some of the first evidence in human subjects for compensation within nondeprived primary sensory areas as a result of blindness early in life.

Processing complexity increases in superficial layers of human primary auditory cortex

The layers of the neocortex each have a unique anatomical connectivity and functional role. Their exploration in the human brain, however, has been severely restricted by the limited spatial resolution of non-invasive measurement techniques. Here, we exploit the sensitivity and specificity of ultra-high field fMRI at 7 Tesla to investigate responses to natural sounds at deep, middle, and superficial cortical depths of the human auditory cortex. Specifically, we compare the performance of computational models that represent different hypotheses on sound processing inside and outside the primary auditory cortex (PAC). We observe that while BOLD responses in deep and middle PAC layers are equally well represented by a simple frequency model and a more complex spectrotemporal modulation model, responses in superficial PAC are better represented by the more complex model. This indicates an increase in processing complexity in superficial PAC, which remains present throughout cortical depths in the non-primary auditory cortex. These results suggest that a relevant transformation in sound processing takes place between the thalamo-recipient middle PAC layers and superficial PAC. This transformation may be a first computational step towards sound abstraction and perception, serving to form an increasingly more complex representation of the physical input.

A critical assessment of data quality and venous effects in sub-millimeter fMRI

Advances in hardware, pulse sequences, and reconstruction techniques have made it possible to perform functional magnetic resonance imaging (fMRI) at sub-millimeter resolution while maintaining high spatial coverage and acceptable signal-to-noise ratio. Here, we examine whether sub-millimeter fMRI can be used as a routine method for obtaining accurate measurements of fine-scale local neural activity. We conducted fMRI in human visual cortex during a simple event-related visual experiment (7 T, gradient-echo EPI, 0.8-mm isotropic voxels, 2.2-s sampling rate, 84 slices), and developed analysis and visualization tools to assess the quality of the data. Our results fall along three lines of inquiry. First, we find that the acquired fMRI images, combined with appropriate surface-based processing, provide reliable and accurate measurements of fine-scale blood oxygenation level dependent (BOLD) activity patterns. Second, we show that the highly folded structure of cortex causes substantial biases on spatial resolution and data visualization. Third, we examine the well-recognized issue of venous contributions to fMRI signals. In a systematic assessment of large sections of cortex measured at a fine scale, we show that time-averaged T₂*-weighted EPI intensity is a simple, robust marker of venous effects. These venous effects are unevenly distributed across cortex, are more pronounced in gyri and outer cortical depths, and are, to a certain degree, in consistent locations across subjects relative to cortical folding. Furthermore, we show that these venous effects are strongly correlated with BOLD responses evoked by the experiment. We conclude that sub-millimeter fMRI can provide robust information about fine-scale BOLD activity patterns, but special care must be exercised in visualizing and interpreting these patterns, especially with regards to the confounding influence of the brain's vasculature. To help translate these methodological findings to neuroscience research, we provide practical suggestions for both high-resolution and standard-resolution fMRI studies.

Resolving laminar activation in human V1 using ultra-high spatial resolution fMRI at 7T

The mesoscopic organization of the human neocortex is of great interest for cognitive neuroscience. However, fMRI in humans typically maps the functional units of cognitive processing on a macroscopic level. With the advent of ultra-high field MRI (≥7T), it has become possible to acquire fMRI data with sub-millimetre resolution, enabling probing the laminar and columnar circuitry in humans. Currently, laminar BOLD responses are not directly observed but inferred via data analysis, due to coarse spatial resolution of fMRI (e.g. 0.7-0.8 mm isotropic) relative to the extent of histological laminae. In this study, we introduce a novel approach for mapping the cortical BOLD response at the spatial scale of cortical layers and columns at 7T (an unprecedented 0.1 mm, either in the laminar or columnar direction). We demonstrate experimentally and using simulations, the superiority of the novel approach compared to standard approaches for human laminar fMRI in terms of effective spatial resolution in either laminar or columnar direction. In addition, we provide evidence that the laminar BOLD signal profile is not homogeneous even over short patches of cortex. In summary, the proposed novel approach affords the ability to directly study the mesoscopic organization of the human cortex, thus, bridging the gap between human cognitive neuroscience and invasive animal studies.

Laminar Organization of Working Memory Signals in Human Visual Cortex

The human primary visual cortex (V1) is not only activated by incoming visual information but is also engaged by top-down cognitive processes, such as visual working memory, even in the absence of visual input [1-3]. This feedback may be critical to our ability to visualize specific visual features, as higher-order regions lack the selectivity to represent such information [4]. Clearly, such internally generated signals do not trigger genuine perception of the remembered stimulus, meaning they must be organized in a manner that is different to bottom-up-driven signals. Internally generated signals may be kept separate from incoming sensory data by virtue of the laminar organization of inter-area cortical connections. Namely, bottom-up driving connections target layer 4, located in the middle of the cortical column, and feedback connections target deep and superficial layers and avoid layer 4 [5-7]. Using lamina-resolved fMRI, we simultaneously measured the activity in three early visual cortical areas (V1-V3) that are recruited to represent stimulus information during visual working memory [8]. We observed item-specific working memory signals in early visual cortex. In V1, this item-specific activity was selectively present at deep and superficial cortical depths, avoiding the middle layers, and working-memory-related activity was present at all depths in V2 and V3. These results show for the first time the laminar organization of internally generated signals during visual working memory in the human visual system and provide new insights into how bottom-up and top-down signals in visual cortex are deployed.

Evaluating the Columnar Stability of Acoustic Processing in the Human Auditory Cortex

Using ultra-high field fMRI, we explored the cortical depth-dependent stability of acoustic feature preference in human auditory cortex. We collected responses from human auditory cortex (subjects from either sex) to a large number of natural sounds at submillimeter spatial resolution, and observed that these responses were well explained by a model that assumes neuronal population tuning to frequency-specific spectrotemporal modulations. We observed a relatively stable (columnar) tuning to frequency and temporal modulations. However, spectral modulation tuning was variable throughout the cortical depth. This difference in columnar stability between feature maps could not be explained by a difference in map smoothness, as the preference along the cortical sheet varied in a similar manner for the different feature maps. Furthermore, tuning to all three features was more columnar in primary than nonprimary auditory cortex. The observed overall lack of overlapping columnar regions across acoustic feature maps suggests, especially for primary auditory cortex, a coding strategy in which across cortical depths tuning to some features is kept stable, whereas tuning to other features systematically varies.SIGNIFICANCE STATEMENT In the human auditory cortex, sound aspects are processed in large-scale maps. Invasive animal studies show that an additional processing organization may be implemented orthogonal to the cortical sheet (i.e., in the columnar direction), but it is unknown whether observed organizational principles apply to the human auditory cortex. Combining ultra-high field fMRI with natural sounds, we explore the columnar organization of various sound aspects. Our results suggest that the human auditory cortex contains a modular coding strategy, where, for each module, several sound aspects act as an anchor along which computations are performed while the processing of another sound aspect undergoes a transformation. This strategy may serve to optimally represent the content of our complex acoustic natural environment.

Feature-dependent intrinsic functional connectivity across cortical depths in the human auditory cortex

Frequency preference and spectral tuning are two cardinal features of information processing in the auditory cortex. However, sounds should not only be processed in separate frequency bands because information needs to be integrated to be meaningful. One way to better understand the integration of acoustic information is to examine the functional connectivity across cortical depths, as neurons are already connected differently across laminar layers. Using a tailored receiver array and surface-based cortical depth analysis, we revealed the frequency-preference as well as tuning-width dependent intrinsic functional connectivity (iFC) across cortical depths in the human auditory cortex using functional magnetic resonance imaging (fMRI). We demonstrated feature-dependent iFC in both core and noncore regions at all cortical depths. The selectivity of frequency-preference dependent iFC was higher at deeper depths than at intermediate and superficial depths in the core region. Both the selectivity of frequency-preference and tuning-width dependent iFC were stronger in the core than in the noncore region at deep cortical depths. Taken together, our findings provide evidence for a cortical depth-specific feature-dependent functional connectivity in the human auditory cortex.

Keeping track of sound objects in space: The contribution of early-stage auditory areas

The influential dual-stream model of auditory processing stipulates that information pertaining to the meaning and to the position of a given sound object is processed in parallel along two distinct pathways, the ventral and dorsal auditory streams. Functional independence of the two processing pathways is well documented by conscious experience of patients with focal hemispheric lesions. On the other hand there is growing evidence that the meaning and the position of a sound are combined early in the processing pathway, possibly already at the level of early-stage auditory areas. Here, we investigated how early auditory areas integrate sound object meaning and space (simulated by interaural time differences) using a repetition suppression fMRI paradigm at 7 T. Subjects listen passively to environmental sounds presented in blocks of repetitions of the same sound object (same category) or different sounds objects (different categories), perceived either in the left or right space (no change within block) or shifted left-to-right or right-to-left halfway in the block (change within block). Environmental sounds activated bilaterally the superior temporal gyrus, middle temporal gyrus, inferior frontal gyrus, and right precentral cortex. Repetitions suppression effects were measured within bilateral early-stage auditory areas in the lateral portion of the Heschl's gyrus and posterior superior temporal plane. Left lateral early-stages areas showed significant effects for position and change, interactions Category x Initial Position and Category x Change in Position, while right lateral areas showed main effect of category and interaction Category x Change in Position. The combined evidence from our study and from previous studies speaks in favour of a position-linked representation of sound objects, which is independent from semantic encoding within the ventral stream and from spatial encoding within the dorsal stream. We argue for a third auditory stream, which has its origin in lateral belt areas and tracks sound objects across space.

Functional connectivity corresponding to the tonotopic differentiation of the human auditory cortex

Recent research has demonstrated that resting-state functional connectivity (RS-FC) within the human auditory cortex (HAC) is frequency-selective, but whether RS-FC between the HAC and other brain areas is differentiated by frequency remains unclear. Three types of data were collected in this study, including resting-state functional magnetic resonance imaging (fMRI) data, task-based fMRI data using six pure tone stimuli (200, 400, 800, 1,600, 3,200, and 6,400 Hz), and structural imaging data. We first used task-based fMRI to identify frequency-selective cortical regions in the HAC. Six regions of interest (ROIs) were defined based on the responses of 50 participants to the six pure tone stimuli. Then, these ROIs were used as seeds to determine RS-FC between the HAC and other brain regions. The results showed that there was RS-FC between the HAC and brain regions that included the superior temporal gyrus, dorsolateral prefrontal cortex (DL-PFC), parietal cortex, occipital lobe, and subcortical structures. Importantly, significant differences in FC were observed among most of the brain regions that showed RS-FC with the HAC. Specifically, there was stronger RS-FC between (1) low-frequency (200 and 400 Hz) regions and brain regions including the premotor cortex, somatosensory/-association cortex, and DL-PFC; (2) intermediate-frequency (800 and 1,600 Hz) regions and brain regions including the anterior/posterior superior temporal sulcus, supramarginal gyrus, and inferior frontal cortex; (3) intermediate/low-frequency regions and vision-related regions; (4) high-frequency (3,200 and 6,400 Hz) regions and the anterior cingulate cortex or left DL-PFC. These findings demonstrate that RS-FC between the HAC and other brain areas is frequency selective.

Reconstructing Tone Sequences from Functional Magnetic Resonance Imaging Blood-Oxygen Level Dependent Responses within Human Primary Auditory Cortex

Here we show that, using functional magnetic resonance imaging (fMRI) blood-oxygen level dependent (BOLD) responses in human primary auditory cortex, it is possible to reconstruct the sequence of tones that a person has been listening to over time. First, we characterized the tonotopic organization of each subject’s auditory cortex by measuring auditory responses to randomized pure tone stimuli and modeling the frequency tuning of each fMRI voxel as a Gaussian in log frequency space. Then, we tested our model by examining its ability to work in reverse. Auditory responses were re-collected in the same subjects, except this time they listened to sequences of frequencies taken from simple songs (e.g., “Somewhere Over the Rainbow”). By finding the frequency that minimized the difference between the model’s prediction of BOLD responses and actual BOLD responses, we were able to reconstruct tone sequences, with mean frequency estimation errors of half an octave or less, and little evidence of systematic biases.