Auditory and visual sensory representations in human prefrontal cortex as revealed by stimulus-evoked spike-wave complexes

Ultra-rapid sensory responses in the human frontal eye field region

Most of what we know about the human frontal eye field (FEF) is extrapolated from studies in animals. There is ample evidence that this region is crucial for eye movements. However, evidence is accumulating that this region also plays a role in sensory processing and that it belongs to a "fast brain" system. We set out to investigate these issues in humans, using intracerebral recordings in patients with drug-refractory epilepsy. Event-related potential recordings were obtained from 11 epileptic patients from within the FEF region while they passed a series of visual and auditory perceptual tests. No eye movement was required. Ultra-rapid responses were observed, with mean onset latencies at 24 ms after stimulus to auditory stimuli and 45 ms to visual stimuli. Such early responses were compatible with cortical routes as assessed with simultaneous recordings in primary auditory and visual cortices. Components were modulated very early by the sensory characteristics of the stimuli, in the 30-60 ms period for auditory stimuli and in the 45-60 ms period for visual stimuli. Although the frontal lobes in humans are generally viewed as being involved in high-level cognitive processes, these results indicate that the human FEF is a remarkably quickly activated multimodal region that belongs to a network of low-level neocortical sensory areas.

Direct intracranial recording of body-selective responses in human extrastriate visual cortex

We recorded intracranial local field potentials (iLFPs) in right extrastriate visual cortex of a patient prior to surgery for epilepsy. Visual evoked potentials revealed a highly selective response to images of bodies, relative to faces, mammals, and tools, which was restricted to a focal region in the lateral occipitotemporal cortex that corresponds to the location of the extrastriate body area (EBA). Body-selective activity started around 190 ms and peaked 260 ms post-stimulus onset. These findings provide the first direct electrophysiological evidence for an early visual processing stage in human lateral occipitotemporal cortex that is specialized for processing human body shapes.

Analysis of Single-Unit Responses to Emotional Scenes in Human Ventromedial Prefrontal Cortex

Lesion and functional imaging studies in humans have shown that the ventral and medial prefrontal cortex is critically involved in the processing of emotional stimuli, but both of these methods have limited spatiotemporal resolution. Conversely, neurophysiological studies of emotion in nonhuman primates typically rely on stimuli that do not require elaborate cognitive processing. To begin bridging this gap, we recorded from a total of 267 neurons in the left and right orbital and anterior cingulate cortices of four patients who had chronically implanted depth electrodes for monitoring epilepsy. Peristimulus activity was recorded to standardized, complex visual scenes depicting neutral, pleasant, or aversive content. Recording locations were verified with postoperative magnetic resonance imaging. Using a conservative, multistep statistical evaluation, we found significant responses in 56 neurons; 16 of these were selective for only one emotion class, most often aversive. The findings suggest sparse and widely distributed processing of emotional value in the prefrontal cortex, with a predominance of responses to aversive stimuli.

Dissociating the cortical basis of memory for voices, words and tones

Human speech carries both linguistic content and information about the speaker's identity and affect. While neuroimaging has been used extensively to study verbal memory, there has been little attention to the neural basis of memory for voices. Evidence from studies of aphasia and auditory agnosia suggests that voice memory may rely on anatomically distinct areas in the right temporal and parietal lobes regions, but there is little data on the broader neural systems involved in voice memory. The present study tested the hypothesis that the neural systems involved in voice memory are functionally distinct from the systems involved in word recognition and are primarily located in the right cerebral hemisphere. Subjects performed two-back tasks in which they were required to alternately remember the voices speaking (Voice condition), and the words they produced (Word condition). A tone memory condition was also included, as a non-speech comparison. The contrast between the Voice and Word conditions revealed greater Voice-related effects in left temporal, right frontal and right medial parietal areas, while the Word-related effects appeared in left frontal and bilateral parietal areas. These findings map out a partially right-lateralized fronto-parietal network associated with voice memory, which can be distinguished from predominantly left-hemisphere regions associated with verbal working memory. These results provide further evidence that distinct neural systems are associated with the carrier waves of speech and word identity.

Two stages in crossmodal saccadic integration: evidence from a visual-auditory focused attention task

Saccadic reaction time (SRT) toward a visual target stimulus was measured under simultaneous presentation of an auditory non-target (accessory stimulus). Horizontal position of the target was varied (25 degrees left and right of fixation) as well as position and intensity of the auditory accessory. SRT was reduced under the presence of the accessory, and it decreased both with increasing intensity of the auditory accessory and with decreasing distance between target and accessory. The absence of a significant interaction between distance and auditory intensity suggests (1) that the intensity of the accessory stimulus has no direct influence on the process of crossmodal integration, and (2) that spatial position and intensity of the accessory are processed in separate stages. This was supported by a probability inequality test showing that the amount of neural coactivation depends on spatial distance but not on auditory intensity. The results are discussed in the framework of a two-stage model assuming separate processing of unimodal and bimodal characteristics of the stimuli. These results are related to several recent neurophysiological findings.

Visual activation of frontal cortex: segregation from occipital activity

Studies in primates have found visually responsive neurons that are distributed beyond cortical areas typically described as directly involved in vision. Among these areas are premotor cortex, supplementary motor area, dorsolateral prefrontal cortex and frontal eye fields. Given these findings, visual stimulation would be expected to result in activation of human frontal cortex. However, few human studies have described sensory activations in frontal regions in response to simple visual stimulation. Such studies have classically described event-related potential (ERP) components over occipital regions. The present study sought to further characterize the spatiotemporal dynamics of visually-evoked electrocortical responses elicited by simple visual stimuli using scalp current density measures derived from high-density ERP recordings, with particular emphasis on the distribution of stimulus-related activity over frontal cortex. Hemiretinal stimuli were viewed passively and during a simple ipsi- or contramanual (RT) task. The motor requirement was included to investigate the effects of response preparation on premovement frontal activations. The results indicate early frontocentral activation, particularly over the right hemisphere (peak magnitude 124-148 ms) that is independent of input visual field or motor response requirement, and that is clearly separate in timecourse from the posterior responses elicited by visual input. These findings are in accord with the multiplicity of visual inputs to frontal cortex and are discussed in terms of frontal lobe functions as may be required in these tasks.

Event-related fMRI of auditory and visual oddball tasks

Functional magnetic resonance imaging (fMRI) was used to investigate the spatial distribution of cortical activation in frontal and parietal lobes during auditory and visual oddball tasks in 10 healthy subjects. The purpose of the study was to compare activation within auditory and visual modalities and identify common patterns of activation across these modalities. Each subject was scanned eight times, four times each for the auditory and visual conditions. The tasks consisted of a series of trials presented every 1500 ms of which 4-6% were target trials. Subjects kept a silent count of the number of targets detected during each scan. The data were analyzed by correlating the fMRI signal response of each pixel to a reference hemodynamic response function that modeled expected responses to each target stimulus. The auditory and visual targets produced target-related activation in frontal and parietal cortices with high spatial overlap particularly in the middle frontal gyrus and in the anterior cingulate. Similar convergence zones were detected in parietal cortex. Temporal differences were detected in the onset of the activation in frontal and parietal areas with an earlier onset in parietal areas than in the middle frontal areas. Based on consistent findings with previous event-related oddball tasks, the high degree of spatial overlap in frontal and parietal areas appears to be due to modality independent or amodal processes related to procedural aspects of the tasks that may involve memory updating and non-specific response organization.

Intracranial ERPs in humans during a lateralized visual oddball task: II. Temporal, parietal, and frontal recordings

OBJECTIVE

We investigated the relative participation of each cerebral hemisphere during a lateralized task that can be performed by a single hemisphere. While our companion article ( Clarke et al. 1999) focused on visual and motor event-related potentials (ERPs) recorded from occipital and peri-Rolandic sites, the present article is concerned with ERP correlates of intermediate stages of information processing from remaining cerebral regions.

METHODS

Intracranial ERPs were recorded from temporal, parietal and frontal lobe sites in 13 patients with intractable epilepsy while they performed a lateralized visual oddball task.

RESULTS

Visually-responsive N2 and/or P3 effects were recorded from medial temporal and supramarginal sites that appear to reflect activity along ventral and dorsal visual streams, respectively. Likewise, certain prefrontal sites, that are probably involved in working memory, were also visually-responsive. The majority of sites exhibiting N2, P3 and/or slow wave components, however, were unaffected by lateralized visual field or response hand effects. Target-evoked P3-like components were most frequently recorded from medial temporal and prefrontal sites. Post-response slow wave components were pervasive, and polarity reversals were present in the insula/operculum region, apparently reflecting somatosensory activity from SII.

CONCLUSIONS

The general absence of lateralized ERP effects despite lateralized stimulus input and response output suggests the importance of interhemispheric integration over hemispheric independence in the processing of this type of task.

Generators of the late cognitive potentials in auditory and visual oddball tasks

Recordings directly within the brain can establish local evoked potential generation without the ambiguities always associated with extracranial electromagnetic measures. Depth recordings have found that sensory stimuli activate primary cortex and then material-specific encoders. Sensory-specific areas remain active for long periods, but by about 200 ms are joined by activation in widespread brain systems. One system is related to the orientation of attention. It is centered in paralimbic and attentional frontoparietocingular cortex, and associated with the P3a. A second system associated with P3b envelopes cognitive contextual integration. It engages the ventral temporofrontal event-encoding cortices (inferotemporal, perirhinal, and ventrolateral prefrontal), association cortices (superior temporal sulcal and posterior parietal), and the hippocampus. Thus, even in simple tasks, activation is widespread but concentrated in particular multilobar systems. With this information, the late cognitive potentials can be used to monitor the probable location, timing and intensity of brain activation during cognitive tasks.

Dissociation of mnemonic coding and other functional neuronal processing in the monkey prefrontal cortex

Single-neuron activity was recorded in the prefrontal cortex of three monkeys during the performance of a spatial delayed alternation (DA) task and during the presentation of a variety of visual, auditory, and somatosensory stimuli. The aim was to study the relationship between mnemonic neuronal processing and other functional neuronal responsiveness at the single-neuron level in the prefrontal cortex. Recordings were performed in both experimental situations from 152 neurons. The majority of the neurons (92%) was recorded in the prefrontal cortex. Nine of the neurons were recorded in the dorsal bank of the anterior cingulate sulcus and two in the premotor cortex. Of the total number of neurons recorded in the prefrontal area, 32% fired in relation to the DA task performance and 39% were responsive to sensory stimulation or to the movements of the monkey outside of the memory task context. Altogether 42% of the recorded neurons were neither activated by the various stimuli nor by the DA task performance. Three types of task-related neuronal activity were recorded: delay related, delay and movement related, and movement related. The majority of the task-related neurons (n = 33, 73%) fired in relation to the delay period. Of the delay-related neurons, 26 (79%) were spatially selective. The number of spatially selective delay-related neurons of the whole population of recorded neurons was 18%. Twelve task-related neurons (27%) fired in relation to the response period of the DA task. Five of these neurons changed their firing rate during the delay period and were classified as delay/movement-related neurons. Contrary to the delay-related neurons, less than half (42%) of the response-related neurons were spatially selective. The majority (70%) of the delay-related neurons could not be activated by any of the sensory stimuli used and did not fire in relation to the movements of the monkey. The remaining portion of the delay-related neurons was activated by stationary and moving visual stimuli or by visual fixation of an object. In contrast to the delay-related neurons, the majority (66%) of the task-related neurons firing in relation to the movement period were also responsive to sensory stimulation outside of the task context. The majority of these neurons responded to visual stimulation, visual fixation of an object, or tracking eye movements. One neuron gave a somatomotor and another a polysensory response. The majority (n = 37, 67%) of all neurons responding to stimulation outside of the task context did not fire in relation to the DA task performance. The majority of their responses was elicited by visual stimuli or was related to visual fixation of an object or to eye movements. Only six neurons fired in relation to auditory, somatosensory, or somatomotor stimulation. This study provides further evidence about the significance of the dorsolateral prefrontal cortex in spatial working memory processing. Although a considerable number of all DA task-related neurons responded to visual, somatosensory, and auditory stimulation or to the movements of the monkey, most delay-related neurons engaged in the spatial DA task did not respond to extrinsic sensory stimulation. These results indicate that most prefrontal neurons firing selectively during the delay phase of the DA task are highly specialized and process only task-related information.