Jw Zhang

BOLD fMRI investigation of the rat auditory pathway and tonotopic organization

Rodents share general anatomical, physiological and behavioral features in the central auditory system with humans. In this study, monaural broadband noise and pure tone sounds are presented to normal rats and the resulting hemodynamic responses are measured with blood oxygenation level-dependent (BOLD) fMRI using a standard spin-echo echo planar imaging sequence (without sparse temporal sampling). The cochlear nucleus (CN), superior olivary complex, lateral lemniscus, inferior colliculus (IC), medial geniculate body and primary auditory cortex, all major auditory structures, are activated by broadband stimulation. The CN and IC BOLD signal changes increase monotonically with sound pressure level. Pure tone stimulation with three distinct frequencies (7, 20 and 40 kHz) reveals the tonotopic organization of the IC. The activated regions shift from dorsolateral to ventromedial IC with increasing frequency. These results agree with electrophysiology and immunohistochemistry findings, indicating the feasibility of auditory fMRI in rats. This is the first fMRI study of the rodent ascending auditory pathway.

High fidelity tonotopic mapping using swept source functional magnetic resonance imaging

Tonotopy, the topographic encoding of sound frequency, is the fundamental property of the auditory system. Invasive techniques lack the spatial coverage or frequency resolution to rigorously investigate tonotopy. Conventional auditory fMRI is corrupted by significant image distortion, sporadic acoustic noise and inadequate frequency resolution. We developed an efficient and high fidelity auditory fMRI method that integrates continuous frequency sweeping stimulus, distortion free MRI sequence with stable scanner noise and Fourier analysis. We demonstrated this swept source imaging (SSI) in the rat inferior colliculus and obtained tonotopic maps with ~2 kHz resolution and 40 kHz bandwidth. The results were vastly superior to those obtained by conventional fMRI mapping approach and in excellent agreement with invasive findings. We applied SSI to examine tonotopic injury following developmental noise exposure and observed that the tonotopic organization was significantly disrupted. With SSI, we also observed the subtle effects of sound pressure level on tonotopic maps, reflecting the complex neuronal responses associated with asymmetric tuning curves. This in vivo and noninvasive technique will greatly facilitate future investigation of tonotopic plasticity and disorders and auditory information processing. SSI can also be adapted to study topographic organization in other sensory systems such as retinotopy and somatotopy.

The inferior colliculus is involved in deviant sound detection as revealed by BOLD fMRI

Rapid detection of deviant sounds is a crucial property of the auditory system because it increases the saliency of biologically important, unexpected sounds. The oddball paradigm in which a deviant sound is randomly interspersed among a train of standard sounds has been traditionally used to study this property in mammals. Currently, most human studies have only revealed the involvement of cortical regions in this property. Recently, several animal electrophysiological studies have reported that neurons in the inferior colliculus (IC) exhibit reduced responses to a standard sound but restore their responses at the occurrence of a deviant sound (i.e., stimulus-specific adaptation or SSA), suggesting that the IC may also be involved in deviance detection. However, by adopting an invasive method, these animal studies examined only a limited number of neurons. Although SSA appears to be more prominent in the external cortical nuclei of the IC for frequency deviant, a thorough investigation of this property throughout the IC using other deviants and efficient imaging techniques may provide more comprehensive information on this important phenomenon. In this study, blood-oxygen-level-dependent (BOLD) fMRI with a large field of view was applied to investigate the role of the IC in deviance detection. Two sound tokens that had identical frequency spectrum but temporally inverted profiles were used as the deviant and standard. A control experiment showed that these two sounds evoked the same responses in the IC when they were separately presented. Two oddball experiments showed that the deviant induced higher responses than the standard (by 0.41±0.09% and 0.41±0.10%, respectively). The most activated voxels were in the medial side of the IC in both oddball experiments. The results clearly demonstrated that the IC is involved in deviance detection. BOLD fMRI detection of increased activities in the medial side of the IC to the deviant revealed the highly adaptive nature of a substantial population of neurons in this region, probably those that belong to the rostral or dorsal cortex of the IC. These findings highlighted the complexity of auditory information processing in the IC and may guide future studies of the functional organizations of this subcortical structure.

Functional magnetic resonance imaging of sound pressure level encoding in the rat central auditory system

Intensity is an important physical property of a sound wave and is customarily reported as sound pressure level (SPL). Invasive techniques such as electrical recordings, which typically examine one brain region at a time, have been used to study neuronal encoding of SPL throughout the central auditory system. Non-invasive functional magnetic resonance imaging (fMRI) with large field of view can simultaneously examine multiple auditory structures. We applied fMRI to measure the hemodynamic responses in the rat brain during sound stimulation at seven SPLs over a 72 dB range. This study used a sparse temporal sampling paradigm to reduce the adverse effects of scanner noise. Hemodynamic responses were measured from the central nucleus of the inferior colliculus (CIC), external cortex of the inferior colliculus (ECIC), lateral lemniscus (LL), medial geniculate body (MGB), and auditory cortex (AC). BOLD signal changes generally increase significantly (p<0.001) with SPL and the dependence is monotonic in CIC, ECIC, and LL. The ECIC has higher BOLD signal change than CIC and LL at high SPLs. The difference between BOLD signal changes at high and low SPLs is less in the MGB and AC. This suggests that the SPL dependences of the LL and IC are different from those in the MGB and AC and the SPL dependence of the CIC is different from that of the ECIC. These observations are likely related to earlier observations that neurons with firing rates that increase monotonically with SPL are dominant in the CIC, ECIC, and LL while non-monotonic neurons are dominant in the MGB and AC. Further, the ICs SPL dependence measured in this study is very similar to that measured in our earlier study using the continuous imaging method. Therefore, sparse temporal sampling may not be a prerequisite in auditory fMRI studies of the IC.

Chronic exposure to broadband noise at moderate sound pressure levels spatially shifts tone-evoked responses in the rat auditory midbrain

Noise-induced hearing disorders are a significant public health concern. One cause of such disorders is exposure to high sound pressure levels (SPLs) above 85 dBA for eight hours/day. High SPL exposures occur in occupational and recreational settings and affect a substantial proportion of the population. However, an even larger proportion is exposed to more moderate SPLs for longer durations. Therefore, there is significant need to better understand the impact of chronic, moderate SPL exposures on auditory processing, especially in the absence of hearing loss. In this study, we applied functional magnetic resonance imaging (fMRI) with tonal acoustic stimulation on an established broadband rat exposure model (65 dB SPL, 30 kHz low-pass, 60 days). The auditory midbrain response of exposed subjects to 7 kHz stimulation (within exposure bandwidth) shifts dorsolaterally to regions that typically respond to lower stimulation frequencies. This shift is quantified by a region of interest analysis that shows that fMRI signals are higher in the dorsolateral midbrain of exposed subjects and in the ventromedial midbrain of control subjects (p<0.05). Also, the center of the responsive region in exposed subjects shifts dorsally relative to that of controls (p<0.05). A similar statistically significant shift (p<0.01) is observed using 40 kHz stimulation (above exposure bandwidth). The results suggest that high frequency midbrain regions above the exposure bandwidth spatially expand due to exposure. This expansion shifts lower frequency regions dorsolaterally. Similar observations have previously been made in the rat auditory cortex. Therefore, moderate SPL exposures affect auditory processing at multiple levels, from the auditory cortex to the midbrain.

Long-term, passive exposure to non-traumatic acoustic noise induces neural adaptation in the adult rat medial geniculate body and auditory cortex

Exposure to loud sounds can lead to permanent hearing loss, i.e., the elevation of hearing thresholds. Exposure at more moderate sound pressure levels (SPLs) (non-traumatic and within occupational limits) may not elevate thresholds, but could in the long-term be detrimental to speech intelligibility by altering its spectrotemporal representation in the central auditory system. In support of this, electrophysiological and behavioral changes following long-term, passive (no conditioned learning) exposure at moderate SPLs have recently been observed in adult animals. To assess the potential effects of moderately loud noise on the entire auditory brain, we employed functional magnetic resonance imaging (fMRI) to study noise-exposed adult rats. We find that passive, pulsed broadband noise exposure for two months at 65 dB SPL leads to a decrease of the sound-evoked blood oxygenation level-dependent fMRI signal in the thalamic medial geniculate body (MGB) and in the auditory cortex (AC). This points to the thalamo-cortex as the site of the neural adaptation to the moderately noisy environment. The signal reduction is statistically significant during 10 Hz pulsed acoustic stimulation (MGB: p<0.05, AC: p<10(-4)), but not during 5 Hz stimulation. This indicates that noise exposure has a greater effect on the processing of higher pulse rate sounds. This study has enhanced our understanding of functional changes following exposure by mapping changes across the entire auditory brain. These findings have important implications for speech processing, which depends on accurate processing of sounds with a wide spectrum of pulse rates.

Noninvasive fMRI investigation of interaural level difference processing in the rat auditory subcortex.

Objective Interaural level difference (ILD) is the difference in sound pressure level (SPL) between the two ears and is one of the key physical cues used by the auditory system in sound localization. Our current understanding of ILD encoding has come primarily from invasive studies of individual structures, which have implicated subcortical structures such as the cochlear nucleus (CN), superior olivary complex (SOC), lateral lemniscus (LL), and inferior colliculus (IC). Noninvasive brain imaging enables studying ILD processing in multiple structures simultaneously. Methods In this study, blood oxygenation level-dependent (BOLD) functional magnetic resonance imaging (fMRI) is used for the first time to measure changes in the hemodynamic responses in the adult Sprague-Dawley rat subcortex during binaural stimulation with different ILDs. Results and Significance Consistent responses are observed in the CN, SOC, LL, and IC in both hemispheres. Voxel-by-voxel analysis of the change of the response amplitude with ILD indicates statistically significant ILD dependence in dorsal LL, IC, and a region containing parts of the SOC and LL. For all three regions, the larger amplitude response is located in the hemisphere contralateral from the higher SPL stimulus. These findings are supported by region of interest analysis. fMRI shows that ILD dependence occurs in both hemispheres and multiple subcortical levels of the auditory system. This study is the first step towards future studies examining subcortical binaural processing and sound localization in animal models of hearing.

BOLD fMRI study of ultrahigh frequency encoding in the inferior colliculus

Many vertebrates communicate with ultrahigh frequency (UHF) vocalizations to limit auditory detection by predators. The mechanisms underlying the neural encoding of such UHF sounds may provide important insights for understanding neural processing of other complex sounds (e.g. human speeches). In the auditory system, sound frequency is normally encoded topographically as tonotopy, which, however, contains very limited representation of UHFs in many species. Instead, electrophysiological studies suggested that two neural mechanisms, both exploiting the interactions between frequencies, may contribute to UHF processing. Neurons can exhibit excitatory or inhibitory responses to a tone when another UHF tone is presented simultaneously (combination sensitivity). They can also respond to such stimulation if they are tuned to the frequency of the cochlear-generated distortion products of the two tones, e.g. their difference frequency (cochlear distortion). Both mechanisms are present in an early station of the auditory pathway, the midbrain inferior colliculus (IC). Currently, it is unclear how prevalent the two mechanisms are and how they are functionally integrated in encoding UHFs. This study investigated these issues with large-view BOLD fMRI in rat auditory system, particularly the IC. UHF vocalizations (above 40kHz), but not pure tones at similar frequencies (45, 55, 65, 75kHz), evoked robust BOLD responses in multiple auditory nuclei, including the IC, reinforcing the sensitivity of the auditory system to UHFs despite limited representation in tonotopy. Furthermore, BOLD responses were detected in the IC when a pair of UHF pure tones was presented simultaneously (45 & 55kHz, 55 & 65kHz, 45 & 65kHz, 45 & 75kHz). For all four pairs, a cluster of voxels in the ventromedial side always showed the strongest responses, displaying combination sensitivity. Meanwhile, voxels in the dorsolateral side that showed strongest secondary responses to each pair of UHF pure tones also showed the strongest responses to a pure tone at their difference frequency, suggesting that they are sensitive to cochlear distortion. These BOLD fMRI results indicated that combination sensitivity and cochlear distortion are employed by large but spatially distinctive neuron populations in the IC to represent UHFs. Our imaging findings provided insights for understanding sound feature encoding in the early stage of the auditory pathway.

Auditory midbrain processing is differentially modulated by auditory and visual cortices: An auditory fMRI study.

The cortex contains extensive descending projections, yet the impact of cortical input on brainstem processing remains poorly understood. In the central auditory system, the auditory cortex contains direct and indirect pathways (via brainstem cholinergic cells) to nuclei of the auditory midbrain, called the inferior colliculus (IC). While these projections modulate auditory processing throughout the IC, single neuron recordings have samples from only a small fraction of cells during stimulation of the corticofugal pathway. Furthermore, assessments of cortical feedback have not been extended to sensory modalities other than audition. To address these issues, we devised blood-oxygen-level-dependent (BOLD) functional magnetic resonance imaging (fMRI) paradigms to measure the sound-evoked responses throughout the rat IC and investigated the effects of bilateral ablation of either auditory or visual cortices. Auditory cortex ablation increased the gain of IC responses to noise stimuli (primarily in the central nucleus of the IC) and decreased response selectivity to forward species-specific vocalizations (versus temporally reversed ones, most prominently in the external cortex of the IC). In contrast, visual cortex ablation decreased the gain and induced a much smaller effect on response selectivity. The results suggest that auditory cortical projections normally exert a large-scale and net suppressive influence on specific IC subnuclei, while visual cortical projections provide a facilitatory influence. Meanwhile, auditory cortical projections enhance the midbrain response selectivity to species-specific vocalizations. We also probed the role of the indirect cholinergic projections in the auditory system in the descending modulation process by pharmacologically blocking muscarinic cholinergic receptors. This manipulation did not affect the gain of IC responses but significantly reduced the response selectivity to vocalizations. The results imply that auditory cortical gain modulation is mediated primarily through direct projections and they point to future investigations of the differential roles of the direct and indirect projections in corticofugal modulation. In summary, our imaging findings demonstrate the large-scale descending influences, from both the auditory and visual cortices, on sound processing in different IC subdivisions. They can guide future studies on the coordinated activity across multiple regions of the auditory network, and its dysfunctions.

Simultaneous spin‐echo and gradient‐echo BOLD measurements by dynamic MRS

This study aimed to dissociate the intravascular and extravascular contributions to spin‐echo (SE) and gradient‐echo (GE) blood oxygenation level‐dependent (BOLD) signals at 7 T, using dynamic diffusion‐weighted MRS. We simultaneously acquired SE and GE data using a point‐resolved spectroscopy sequence with diffusion weightings of 0, 600, and 1200 s/mm2. The BOLD signals were quantified by fitting the free induction decays starting from the SE center to a mono‐exponential decay function. Without diffusion weighting, BOLD signals measured with SE and GE increased by 1.6 ± 0.5% (TESE = 40 ms) and 5.2 ± 1.4% (nominal TEGE = 40 ms) during stimulation, respectively. With diffusion weighting, the BOLD increase during stimulation measured with SE decreased from 1.6 ± 0.5% to 1.3 ± 0.4% (P < 0.001), whereas that measured by GE was unaffected (P > 0.05); the post‐stimulation undershoots in the BOLD signal time courses were largely preserved in both SE and GE measurements. These results demonstrated the feasiblity of simultaneous SE and GE measurements of BOLD signals with and without interleaved diffusion weighting. The results also indicated a predominant extravascular contribution to the BOLD signal time courses, including post‐stimulation undershoots in both SE and GE measurements at 7 T.