This review highlights recent developments in research on human cortical oscillations in the gamma-band range (30–100 Hz). Electroencephalography has demonstrated a role of these signals for cognitive functions including visual perception, attention, learning and memory. During auditory processing, magnetoencephalogram has identified oscillatory activity in higher frequency ranges and with a more discrete localization than electroencephalogram. Gamma-band activity increases have been observed in the putative auditory dorsal and ventral processing streams during the processing of auditory spatial and pattern information, respectively. Additional gamma-band activity has been found over the frontal cortex during top-down tasks. Oscillatory activity in the gamma range may serve to assess the temporal dynamics of cortical networks and their interactions.

The representation of visual objects in short-term memory has been shown to be related to increased gamma-band activity in the electroencephalogram. Using a similar paradigm, we investigated oscillatory magnetoencephalographic activity in human subjects during a delayed matching-to-sample task requiring working memory of auditory spatial information. The memory task involved same–different judgments about the lateralization angle of pairs of filtered noise stimuli (S1 and S2) separated by 800 msec delays of background noise. This was compared with a control condition requiring the detection of a possible change in the background noise volume appearing instead of S2 (volume task). Statistical probability mapping revealed increased spectral activity at 59 Hz over left parietal cortex during the delay phase of the memory condition. In addition, 59 Hz coherence was enhanced between left parietal and right frontal sensors. During the end of the delay and during the presentation of S2, enhanced gamma-band activity at 67 Hz was observed over right frontal and later over midline parietal areas. In contrast, the volume task was characterized by increased left inferior frontotemporal 59 Hz spectral amplitude after S1. Apparently representation of the spatial position of a sound source is associated both with synchronization of networks in parietal areas involved in the auditory dorsal stream and with increased coupling between networks serving representation of audiospatial information and frontal executive systems. The comparison with S2 seemed to activate frontal and parietal neuronal ensembles. Gamma-band activity during the volume task may reflect auditory pattern encoding in auditory ventral stream areas.

In animals, physical activity has been shown to induce functional and structural changes especially in the hippocampus and to improve memory, probably by upregulating the release of neurotrophic factors. In humans, results on the effect of acute exercise on memory are inconsistent so far. Therefore, the aim of the present study was to assess the effects of a single bout of physical exercise on memory consolidation and the underlying neuroendocrinological mechanisms in young adults. Participants encoded a list of German-Polish vocabulary before exercising for 30 minutes with either high intensity or low intensity or before a relaxing phase. Retention of the vocabulary was assessed 20 minutes after the intervention as well as 24 hours later. Serum BDNF and salivary cortisol were measured at baseline, after learning, and after the intervention. The high-intensity exercise group showed an increase in BDNF and cortisol after exercising compared to baseline. Exercise after learning did not enhance the absolute number of recalled words. Participants of the high-intensity exercise group, however, forgot less vocabulary than the relaxing group 24 hours after learning. There was no robust relationship between memory scores and the increase in BDNF and cortisol, respectively, suggesting that further parameters have to be taken into account to explain the effects of exercise on memory in humans.

Cortical processing of change in direction of a perceived sound source was investigated in 12 human subjects using whole-head magnetoencephalography. The German word “da” was presented either with or without 0.7 msec interaural time delays to create the impression of right- or left-lateralized or midline sources, respectively. Midline stimuli served as standards, and lateralized stimuli served as deviants in a mismatch paradigm. Two symmetrically linked dipoles fitted to the mismatch fields showed stronger moments in the hemisphere contralateral to the side of the deviant. The right dipole displayed equal latencies to both left and right deviants, whereas left dipole latencies were longer for ipsilateral than contralateral deviants. Frequency analysis between 20–70 Hz and statistical probability mapping revealed increased induced gamma-band activity at 53 ± 2.5 Hz to both types of deviants. Right deviants elicited spectral amplitude enhancements in this frequency range, peaking at latencies of 160 and 240 msec. These effects were localized bilaterally over the angular gyri and posterior temporal regions. Coherence analysis suggested the existence of two separate interhemispheric networks. For left-lateralized deviants, both spectral amplitude enhancements at 110 and 220 msec and coherence increases were restricted to the right hemisphere. In conclusion, both mismatch dipole latencies at the supratemporal plane and gamma-band activity in posterior parietotemporal areas suggested a right hemisphere engagement in the processing of bidirectional sound-source shifts. In contrast, left-hemisphere regions responded predominantly to contralateral events. These findings may help to elucidate phenomena such as unilateral auditory neglect.