Abstract-Over the past decade, many laboratories have begun to explore brain-computer interface (BCI) technology as a radically new communication option for those with neuromuscular impairments that prevent them from using conventional augmentative communication methods. BCI's provide these users with communication channels that do not depend on peripheral nerves and muscles. This article summarizes the first international meeting devoted to BCI research and development. Current BCI's use electroencephalographic (EEG) activity recorded at the scalp or single-unit activity recorded from within cortex to control cursor movement, select letters or icons, or operate a neuroprosthesis. The central element in each BCI is a translation algorithm that converts electrophysiological input from the user into output that controls external devices. BCI operation depends on effective interaction between two adaptive controllers, the user who encodes his or her commands in the electrophysiological input provided to the BCI, and the BCI which recognizes the commands contained in the input and expresses them in device control. Current BCI's have maximum information transfer rates of 5-25 b/min. Achievement of greater speed and accuracy depends on improvements in signal processing, translation algorithms, and user training. These improvements depend on increased interdisciplinary cooperation between neuroscientists, engineers, computer programmers, psychologists, and rehabilitation specialists, and on adoption and widespread application of objective methods for evaluating alternative methods. The practical use of BCI technology depends on the development of appropriate applications, identification of appropriate user groups, and careful attention to the needs and desires of individual users. BCI research and development will also benefit from greater emphasis on peer-reviewed publications, Publisher Item Identifier S 1063-6528(00)04484-0.and from adoption of standard venues for presentations and discussion.Index Terms-Brain-computer interface (BCI), electroencephalography (EEG), augmentative communication.

The second somatosensory area (SII) of awake, untrained cynomolgus monkeys was surveyed with recordings from nearly 1,000 single neurons. A detailed somatotopographic organization could be demonstrated in SII because the majority of these neurons had contralateral, moderate to well-defined receptive fields of < 10 cm2, and because neighboring neurons possessed receptive fields that were only slightly displaced from one another. Different body regions were represented in successive anterior to posterior strips that were oriented across the parietal operculum with an anterolateral to posteromedial slant. Neurons with trigeminal receptive fields were found in the anterior portion of SII; these neurons were the only ones in SII with predominantly bilateral receptive fields (r.f.'s.). Neurons with digit or hand r.f'.s form the largest component of the map, and were located posterior to those with face r.f.s. Most of these neurons had only contralateral activation. The hand and digit region was followed in turn by the arm, the upper and lower trunk, and the hindlimb regions. Although the overall SII orientation was along an anterior-posterior gradient, recordings at individual coronal planes often demonstrated isolated sequences of receptive fields that exhibited a medial-lateral progression. The principle example of this latter gradient was seen in the forelimb region where digits one through five were represented in an overlapping sequence across the parietal operculum. Except for portions of the digit representation, neighboring sequences of neurons in SII do not form a precise topologic map of the body that is comparable to the somatotopic maps observed in areas 3b and 1. The present findings contrast with previous physiological studies of SII in the primate. These discrepancies are discussed in relation to methodological differences and in terms of distinctions used to define the boundaries of SII.

Somatic response properties were determined for over 1,300 neurons isolated within and near the lateral sulci of unanesthetized and unparalyzed cynomolgus monkeys. Somatic stimuli unequivocally activated the majority of units studied in SII (93%) and in cortical fields surrounding SII: area 7b (65%), the retroinsular field (74%), and the granular insula (76%). No activation other than somatic was seen for SII neurons, and noxious somatic stimulation was rarely required. The SII units almost always responded in a rapidly adapting manner to hair or skin stimulation, but not both; however, the submodality distribution seen in SII varied as a function of peripheral receptor locations. Two small zones within SII contained neurons that responded only if the animal actively interacted with the stimulus. In contrast, one-half of the sample of neurons from area 7b unequivocally responded only to somatic stimulation. Although many neurons in the lateral parts of area 7b were vigorously activated by innocuous tactile stimulation, others demonstrated little association with an identifiable somatic submodality, had sluggish responses, required complex, noxious, visual or other non-somatic stimuli for activation, and had labile response properties and receptive fields. Indeed, the responses of some area 7b neurons suggested a possible relationship with the animal's attention towards or anticipation of a noxious or a novel somatic stimulus. Neurons within the retroinsular cortex (Ri), which receives projections from the posterior nucleus (PO), primarily responded to light tactile stimulation of rapidly adapting skin receptors; less than 3% responded to moderate or high threshold mechanical stimulation. The sensitivity to tactile stimulation in Ri closely resembled the responses of SII neurons. Neurons in the granular insula (Ig) often responded to gentle hair deflection within receptive fields covering large areas of the body. Ig and area 7b were the principle loci within the lateral sulcus that contained neurons responding to noxious stimulation. Owing to the great similarity in the somatic response properties within these areas in the awake and unparalyzed animal, the designation of cortical areas could only be made after correlating the recording sites with connectional and cytoarchitectonic analyses in the same animal. Consequently, previous physiological studies may have attributed to SII some of the response characteristics of neurons in neighboring areas.

The boundaries of the second somatic sensory cortex (SII) in primates are difficult to define physiologically because cutaneous stimulation activates several regions around SII that do not receive projections from the ventroposterior nucleus of the thalamus. These cortical regions, which include portions of area 7b, the retroinsular (Ri) and postauditory fields (PA), and the granular insula (Ig) are largely buried within the lateral sulcus and most lie posterior to the caudal end of the insula. The differences in somatic activity in these various cortical fields in the unanesthetized cynomolgus monkey became apparent only after the properties of many neighboring neurons could be compared. Receptive fields for area 7b and Ig neurons were generally large (< 10 cm2), with bilateral, moderately defined boundaries; some neurons in area 7b had receptive fields with labile borders as a function of wakefulness. In contrast, receptive fields for Ri neurons were generally (< 10 cm2 and contralateral, with stable, well-defined boundaries. Taken as an ensemble, the neurons in areas neighboring SII exhibited a very crude topography; but at the level of an individual neuron and its neighbor, there was never a pattern of gradual transition in peripheral receptive field locations between one unit and the next, like that seen in SII. In area 7b, this crude map was organized mediolaterally across the inferior parietal lobule and into the upper bank of the lateral sulcus, with the head represented medially and the lower trunk and hindlimb laterally. In Ri-PA, an anteroposterior organization was noted along the fundus of the lateral sulcus with the head represented anterior to the lower trunk and hindlimb. No organization was apparent in Ig. Additional sensitivity to visual stimuli was noted in the more medial aspects of area 7b that were located on the exposed inferior parietal lobule. Sensitivity to auditory stimuli was principally found in PA and occasionally in Ri. The results, especially from area 7b, are discussed with respect to previous notions about the organization of SII.

The peripheral neuronal correlates of heat pain elicited from normal skin and from skin made hyperalgesic following a mild heat injury were studied by simultaneously recording, in humans, evoked responses in C mechanoheat (CMH) nociceptors and the magnitude estimations of pain obtained from the same subjects. Subjects made continuous magnitude ratings of pain elicited by short-duration stimuli of 39-51 degrees C delivered to the hairy skin of the calf or foot before and at varying intervals of time after a heat injury induced by a conditioning stimulus (CS) of 50 degrees C, 100 s or 48 degrees C, 360 s. The stimuli were applied with a thermode pressed against the nociceptor's receptive field. For heat stimulations of normal skin, that is, uninjured skin, pain thresholds in 14 experiments with nine subjects ranged from 41 to 49 degrees C, whereas response thresholds for most of the 14 CMH nociceptors were 41 degrees C (in two cases, 43 degrees C). The latter suggested that spatial summation of input from many nociceptors was necessary at pain threshold. An intensity-response function was obtained for each CMH by relating the total number of nerve impulses evoked per stimulus to stimulus temperature. A corresponding magnitude scaling function for pain was obtained by relating the maximum rating of pain elicited by each stimulus to stimulus temperature. The relation between the subject's scaling function and the intensity-response function of his CMH nociceptor varied somewhat from one experiment to the next, regardless of whether the results were obtained from the same or from different subjects. However, when averages were computed for all 14 tests, there was a near linear relationship between the mean number of impulses elicited in the CMHs and the median ratings of pain, over the range of 45-51 degrees C. It was concluded that the magnitude of heat pain sensation was more closely related to the magnitude of response in a population of CMH nociceptors than in any individual nociceptor. At 0.5 min after the CS, the pain thresholds of most subjects were elevated, and the magnitude ratings of pain elicited by supra-threshold stimuli were lower than pre-CS values (hypoalgesia). Corresponding changes were seen in the increased thresholds and decreased responses (fatigue) of most CMHs. By 5-10 min after the CS, the pain thresholds of most subjects were lower, and their magnitude ratings of suprathreshold stimuli were greater than pre-CS values (hyperalgesia).(ABSTRACT TRUNCATED AT 400 WORDS)