Tsunetaka Okita

Memory-search and visual spatial attention - an event-related brain potential analysis

Event-related brain potentials (ERPs) were recorded from subjects as they attended to one diagonal of a visual display. The task was to respond only to memory set items (targets) at the attended diagonal and to ignore stimuli at the other diagonal. The probability that the display contained either an attended or an unattended target was 0.30. The spatial separation between attended and unattended stimuli was 1.6°. The ERP elicited by stimuli at the unattended diagonal contained a sequence of phasic components. The early N170 and P250 components were elicited by the onset of the display and the later components N480 and P550 by the offset of the display. The presence of masks delayed N170 and P250. The ERPs elicited by attended non-targets, in addition, contained an increased N350 (Cz, Fz) and P410 (P3a, Pz, Cz). The ERPs elicited by attended and unattended non-targets started to differ after 200 msec. This finding suggested that selection is relatively late if selection must be based on a conjunction of features (location and orientation) and if the spatial separation between attended and unattended stimuli is small. Memory-set size affected the ERPs after 250 msec. The ERPs elicited by attended stimuli contained a broadly distributed (Fz, Cz, Pz) negative endogeneous component. The amplitude of this component was related to memory-set size. Finally, the ERPs elicited by attended targets contained a large P3b (Pz, Oz) with a peak latency around 600 msec. The ERP results suggested the existence of three processing stages: (1) orienting to the attended stimuli; (2) controlled search, and (3) target decision.

Attention-related negative brain potential for speech words and pure tones

Abstract Event-related potentials (ERPs) were recorded during performance of a selective attention task in which speech words and pure tones were dichotically delivered at random. To estimate the attention-related negative shift, ERPs to unattended stimuli were subtracted from ERPs to attended stimuli. The difference waves for both non-target words and tones demonstrated a biphasic wave form: a fronto-centrally distributed component with a peak latency at about 165 msec and a more frontally focussed component which peaked about 300 msec. The scalp distributions, differential experimental manipulations and developing time-courses suggest that the early negativity manifests the auditory discrimination activity on a task-relevant stimulus for deciding whether it is a target or not, whereas the late negativity reflects modality-nonspecific activation processes which may be associated with ‘momentary effort’ or ‘orienting’ triggered by detection of the task relevancy. Neither the early nor late components, however, presented hemispheric asymmetry corresponding to linguistic processing.

Event-related potentials and selective attention to tones moving in location and pitch: an examination of movement velocity

Event-related potentials (ERPs) were recorded from subjects who listened selectively to one of two classes of tone pips that arose from separate, virtual sound sources moving in spatial location and pitch. In different runs the speeds of the moving sources were varied; for the slow speed, the two classes of tones were easily separable, but for the moderate and fast speeds selective listening was difficult without practice. A prolonged negative shift of the ERP, beginning around 60 ms poststimulus and lasting over 300 ms, was observed when the eliciting tones were selectively attended. However, no differences at any phase of the negative ERPs were found as a function of movement velocity, in contrast to the decline of target detection performance for the fast speed. The results were interpreted as indicating that anticipation, based on contextual information, can play an important role in early input selection.

Specific and non-specific effects of transcranial magnetic stimulation on simple and go/no-go reaction time

Abstract The effects of subthreshold transcranial magnetic stimulation (TMS) on simple and go/no-go reaction time (RT) tasks were studied in seven healthy volunteers. Subjects were asked to respond by abducting the thumb in a warning-imperative signal paradigm. TMS was randomly delivered at variable delays to the imperative signal (IS). Simple RT was significantly shortened when TMS was delivered to the left motor cortex and parietal regions simultaneously with IS. In the go/no-go paradigm, a similar trend to shorter RT was seen at a delay of 0 ms. Additionally, a significant shortening was observed at a delay of 90 ms with TMS over the contralateral motor cortex only. Movement-related potentials (MRPs) in the two paradigms showed a predominantly contralateral negativity approximately 80 ms preceding EMG onset. Our findings support the existence of two differentiated effects of TMS on RT: (1) one non-specific effect, evidenced in both the simple and go/no-go paradigms at a 0 ms delay, which can be at least partially explained by intersensory facilitation; and (2) a motor-specific effect of TMS, unveiled in the go/no-go paradigm at a 90 ms delay.

Selective attention and N400 attenuation with spoken word repetition.

In two experiments, event-related brain potential were recorded to word pairs simultaneously presented to both ears, with instructions to attend to one ear and detect occasional nonwords in that ear. This attentional manipulation yielded four patterns of word repetition on successive trials: first and second presentations attended (AA), both unattended (UU), and across ears (AU and UA). A prominent attenuation of N400 due to immediate repetition of words was observed on AA trials. However, when first presentations were ignored on UU and UA trials, no repetition effect was obtained. These findings indicate that the repetition effect on N400 depends on attentional processing of first presentations.

Event-related potentials reveal memory deficits in Parkinson's disease.

To evaluate whether patients with Parkinsons disease (PD) exhibit deficits in memory tested indirectly, we measured event-related potentials (ERPs) in 23 patients with nondemented PD and 14 age-matched control subjects. The auditory N400 component was measured during an indirect repetition priming paradigm. Some words were repeated immediately after initial presentation (lag 0), while others were repeated after five intervening words (lag 5) or at lags 11-77. Subjects were required to push a button to the occasional nonwords (targets). The N400 amplitudes were quantified in latency window between 300 and 800 ms for the different conditions. The N400 amplitudes for the first presentation were significantly lower in PD patients than controls for both words and nonwords. The N400 in the control group was attenuated for lag 0, lag 5, and lags 11-77 repetitions, while the attenuation in PD patients was noted only for lag 0 repetition. The data suggest that memory assessed in this fashion is impaired in PD patients.

Auditory event-related potentials in an amnesic patient with a left temporal lobe lesion.

We studied auditory event-related potentials (ERPs) in correlation with Reys auditory verbal learning test (AVLT) in an amnesic patient with a left temporal lobe lesion. On the AVLT, the number of words recalled was limited to immediate memory capacity (nine words) throughout five trials, and recall after interference was remarkably decreased. In the ERPs, the P300 component was elicited normally, but no N400 component was apparent. Present findings suggest that these two ERP components may be associated with independent processes and support the view that the N400 may index a neural process involved in transfer of information from immediate memory to long-term storage.

Memory in patients with subcortical infarction : an auditory event-related potential study

Event-related potentials (ERPs) were recorded from 21 patients with subcortical infarction (mean age, 62.1 years) and 14 normal control subjects (mean age, 62.7 years) as they listened to lists of words or pronounceable non-words. Some words were repeated immediately after initial presentation (lag 0), while others were repeated after five intervening words (lag 5), or after 2 to 4 min (lag 11-77). The subjects were asked to push a button upon hearing the occasional non-words. The Auditory-Verbal Learning Test (AVLT) also was administered to the patients to examine explicit memory. The mean N400 amplitude, appearing between 300 and 800 ms after the stimulus, was smaller in patients with subcortical infarction than in control subjects. The N400 in response to repeated words at lags 0, 5 and 11-77 was attenuated for both the patient and control groups. On the AVLT the total number of recalled words and number of words after interference were significantly decreased in patients relative to controls, while recognition was relatively preserved. The results suggest that lexical processing and retrieval mechanism of explicit memory are disturbed in patients with subcortical infarction, but implicit memory measured by this N400 paradigm is relatively preserved.

Effect of subject's family name on visual event-related potential in schizophrenia

Visual event-related potentials (ERPs) were recorded from 15 schizophrenics and 15 age-matched and gender-matched controls, while they performed a modified version of the oddball paradigm. Each subject was required to detect target stimuli among a random sequence of stimuli under two conditions, name and color. In the name condition the stimulus sequence consisted of the subjects family name (deviant 18%), four other family names (standard 73%), and a city name (target 9%). In the color condition the respective stimuli were a pair of solid red circles, four white paired-arrows, and a pair of white plus and minus signs. ERPs elicited by stimuli contained triphasic potentials of P2, N2, and P3. In controls these waves were selectively enhanced for the subjects family name as compared with standards, whereas in schizophrenics no significant difference between the subjects family name and other names was observed. In contrast, selective enhancement for color deviants was observed in both subject groups. These results suggest that impairment of involuntary attention, especially for familiar and significant stimuli such as names in daily life, may underlie disturbances of attentionally controlled central processing in schizophrenia.

The effects of active attention switching between the ears on averaged evoked potentials

Abstract The relationship between averaged vertex evoked potentials and active switching of attention between the ears was studies in 6 trained subjects. A train of tone pips from 6 to 9 on each trial was presented in four modes, of which two modes were monotic presentations to either ear and the other two were dichotic, alternating and random presentations. These pips were delivered at either regular or irregular intervals which averaged 1000 msec. N 1 P 2 amplitudes were found to be larger when attention switching was required between the ears in the dichotic presentations as compared with monotic presentations which allowed subjects to restrict attention to one ear. The random presentation mode produced a further enhancement of N 1 P 2 amplitude as compared to the alternating mode. Within the random mode, attention switching elicited delayed P 2 latency relative to a condition not requiring subjects to switch attention. Irregular stimulus presentation delayed N 1 latency regardless of the attention switching requirement. These results were related to the successive stages of human information processing.