Do you taste it already?
Brain responses to taste in humans are fast.
Kathrin Ohla, Julie Hudry & Johannes le Coutre
Nestlé Research Center , Lausanne, Switzerland
BACKGROUND
SUMMARY and CONCLUSION
Event-related potentials (ERPs) are commonly used for the objective
evaluation of sensory functions. In the gustatory system, however, studies of
event-related waveforms have yielded little agreement among researchers
[1,2,3]. Numerous approaches failed to provide the onset precision required
for averaging ERPs following lingual stimulation by tastants feeding the
assumption that the first gustatory ERP component is a slow wave peaking
at 300-500 ms. To overcome this difficulty, we employed electrogustometry,
an approach that provides temporally accurate stimulus control, in
combination with ICA. Although this method has been used previously, the
analysis have been hampered by the electrical artifacts that mask the
electroencephalographic (EEG) responses [4,5].
The present approach of electric taste stimulation evoked an unequivocal
taste perception and, moreover, provided the temporal precision required to
explore the resulting event-related potential. We show that the gustatory
evoked potential exhibits a waveform with unambigous peaks the neuronal
generators of which were mainly localised in the insular cortex. Statistical
analyses of the ERP waveforms revealed shorter latencies and higher
amplitudes for electric pulses at maximum intensities. Stimulation of both
sides of the tongue elicited similar cortical responses. Although we can
neither exclude nor quantify the amount of trigeminal activation our results
provide important insights into the temporal and spacial dynamics of
gustatory signal processing.
Aim of the study was to assess the temporal and spacial properties of taste perception in the human brain.
RESULTS
• Mean taste detection threshold: 3.6 db ( 11.5 µA; 0.6 µA/mm2)
• Mean maximum intensity:
33 db (360.5 µA; 18.4 µA/mm2)
• Reported taste quality: metallic (21%), rusty (11%), sour (11%) taste and
mild tingling (12%).
• Mean pleasantness rating: 5.4 (range: 3-8; scale: 1[-]-10[+]).
ERPs. ERP components were similar for stimulations of the right and left side of the
tongue. Electrodewise t-tests exhibited broadly distributed bilateral differences between
stimulus intensities; no effects of side of stimulation were observed. (N=17)
right
left
135 ms
125 ms
135 ms
125 ms
ICA-based artefact correction. While the EEG data (at Cz) were dominated by a
square wave artefact before artefact reduction, the data quality improved considerably
after unmixing. The ERP became apparent even at single trial level (ERPimage, right
panel) and topographical maps clearly show the signal’s propagation over time. (N=1)
150-250 ms
70-170 ms
280-480 ms
ERPs and current source density maps. High intensity pulses evoked higher amplitudes
and shorter peak latencies as compared to low intensity pulses. CSD maps exhibit cortical
sources over the junction of the lateral and central sulcus, the area forming the opercula that
cover the insula, which has been sugested to comprise the primary gustatory cortex. (N=17)
LAURA source estimation. Cortical generators were estimated at the peak latencies of
each ERP component. During the P1 the insular cortices were bilaterally activated. While
the N1 peak showed a prominent activation of the right insula irrespective of the side of
stimulation at low intensities the focus of activation at high intensities was found over the
anterior cingulate area.
METHODS
EEG was recorded from 64 channels (Biosemi) while anodal electrical
pulses at two intensities were applied to participants’ tongues. Stimulus
intensities were determined for each participant individually. With extended
infomax Independent Component Analysis (ICA) electrical stimulus artifacts
and EEG data were separated [6] and the temporal and spatial dynamics of
the electro-gustatory ERPs could be assessed. ERPs, ERP waveform
modulations, current source density (CSD) maps, Global Field Power (GFP),
Global Dissimilarities, and source estimates (LAURA) were computed.
References
[1] Kobal, 1985, Electr. Clin. Neurophysiol., 62, 449-454.
[3] Mizoguchi et al., 2002, Chem. Senses, 27, 629-634.
[5] Yamamoto et al., 2003, Chem. Senses, 28, 245-251.
[2] Kobayakawa et al., 1996, Neuroscience Letters, 212, 155-158
[4] Fitzsimons et al., 1999, Physiol. Meas., 20, 385-400.
[6] Ohla et al., subm. for publication.