Involvement of right piriform cortex in olfactory familiarity judgments

www.elsevier.com/locate/ynimg NeuroImage 24 (2005) 1032 – 1041

Involvement of right piriform cortex in olfactory familiarity judgments Jane Plailly,a,* Moustafa Bensafi,a Mathilde Pachot-Clouard,b Chantal Delon-Martin,b David A. Kareken,c Catherine Rouby,a Christoph Segebarth,b and Jean-P. Royeta,d a

Neurosciences et Syste`mes Sensoriels, Universite´ Claude Bernard Lyon 1, UMR CNRS 5020, IFR 19, Institut Fe´de´ratif des Neurosciences de Lyon, 50 Avenue Tony Garnier, 69366 Lyon Cedex 07, France b Unite´ mixte INSERM/Universite´ Joseph Fourier U594, LRC-CEA, Hoˆpital Michallon, 38043 Grenoble, France c Neuropsychology Section, Department of Neurology, Indiana University School of Medicine, Indianapolis, IN 46202, USA d CERMEP, 69003 Lyon, France Received 16 April 2004; revised 12 October 2004; accepted 26 October 2004 Available online 9 December 2004 Previous studies have shown activation of right orbitofrontal cortex during judgments of odor familiarity. In the present study, we sought to extend our knowledge about the neural circuits involved in such a task by exploring the involvement of the right prefrontal areas and limbic/ primary olfactory structures. Fourteen right-handed male subjects were tested using fMRI with a single functional run of two olfactory conditions (odor detection and familiarity judgments). Each condition included three epochs. During the familiarity condition, subjects rated whether odors were familiar or unfamiliar. During the detection condition, participants decided if odors were present. When contrasting the familiarity with the detection conditions, activated areas were found mainly in the right piriform cortex (PC) and hippocampus, the left inferior frontal gyrus and amygdala, and bilaterally in the mid-fusiform gyrus. Further analyses demonstrated that the right PC was more strongly activated than the left PC. This result supports the notion that the right PC is preferentially involved in judgments of odor familiarity. D 2004 Elsevier Inc. All rights reserved. Keywords: Olfaction; Familiarity judgment; Recognition memory; Piriform cortex; fMRI

Introduction Hemispheric asymmetry is well-established for high-level brain functions such as language and spatial attention (e.g., Broca, 1863; Weintraub and Mesulam, 1987). Hemispheric predominance also exists in sensory functions such as hand somatosensory representation (Soros et al., 1999) and temporal and spectral auditory resolution (Zatorre et al., 2002). Studies in olfaction lead to similar conclusions. Early cerebral imaging studies showed functional * Corresponding author. Fax: +33 4 37 28 76 01. E-mail address: [email protected] (J. Plailly). Available online on ScienceDirect (www.sciencedirect.com). 1053-8119/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2004.10.028

lateralization of olfactory processes in the right hemisphere, especially in orbitofrontal cortex (OFC) (Zatorre et al., 1992); most subsequent studies confirm this result (Dade et al., 1998; Sobel et al., 1998; Yousem et al., 1997). However, Zald and colleagues reported stronger activation in the left OFC and amygdala for very aversive odors, pointing to these areas as being important in the emotional processing of olfactory information (Zald and Pardo, 1997; Zald et al., 1998). Using positron emission tomography (PET), Royet et al. (1999, 2001) found that judgments of odor familiarity preferentially activated right OFC, whereas hedonic judgments principally activated left OFC. Beyond OFC, recent cerebral imaging data have extended these observations to piriform cortex (PC). In a functional magnetic resonance imaging (fMRI) study, we showed that left piriform– amygdala region activation was associated with subjects’ ratings of the odorants’ emotional intensities (Royet et al., 2003), a result consistent with Gottfried et al.’s (2002) and Anderson et al.’s (2003) findings. Convergent results from Dade et al.’s (2002) lesion and PET studies further show that the piriform region mediates olfactory long-term recognition memory, giving support to the notion that this area may be more than primary sensory cortex (e.g., Schoenbaum and Eichenbaum, 1995). Specifically, Dade et al. (2002) found that the extent of piriform activity corresponded with different cognitive demands, in which PC activity followed a continuum between mnemonic encoding (no significant activity), to short-term recognition (weak bilateral activity), to long-term recognition (strong bilateral activity). These authors further suggested that piriform activity could be related to odor familiarity, which would require that subjects compare odors with previously stored olfactory representations, and thus represent a type of long-term olfactory reference memory. In the present study, we explored that question by specifically asking whether PC is involved in the processing of odor familiarity, and if so, whether familiarity-evoked activations are lateralized, as previously suggested (Royet et al., 1999, 2001). We studied

J. Plailly et al. / NeuroImage 24 (2005) 1032–1041

familiarity judgments and their relation to PC activation with fMRI using a classical block paradigm design. Familiar and unfamiliar odors were presented in a same epoch allowing subjects to rate familiarity in a binary fashion by pressing one of two buttons. A control condition was employed in which subjects had to judge the presence or absence of an odor. The contrast between both conditions allowed us to identify the areas specifically involved in the familiarity judgment of odors.

Materials and methods Subjects Fourteen healthy right-handed men (18–32 years old) participated in the study. Participation required a medical screening. Exclusion criteria were rhinal disorders (colds, active allergies, history of nasal-sinus surgery, or asthma), neurologic disease, ferrous implants (e.g., pacemakers, cochlear implants), or claustrophobia. Participants scored at least 87% correct in a forcedchoice suprathreshold detection test, and had a breathing cycle mean duration of 3.88 s (F0.72). All participants provided written informed consent as approved by the local Institutional Review Board, and according to French regulations on biomedical experiments on healthy volunteers. Odorous stimuli One hundred and eight stimuli were used, 27 of which were employed before imaging sessions and 81 of which were used

1033

during scanning. For fMRI, 54 odorants were used for the familiarity (F) condition and 27 odorants for the detection (D) condition (Table 1). For F conditions, three sets (Fa, Fb, Fc) contained nine familiar and nine unfamiliar odorants selected so as to provide high- and low-familiarity scores from data as derived from previous work (Royet et al., 1999). Analysis of variance (ANOVA) indicated that familiarity scores were significantly higher for familiar than for unfamiliar odorants [ F(1,48) = 129.173, P b 0.0001]. For D conditions, three sets (Da, Db, Dc) contained nine odorants with low familiarity and nine bottles with odorless air. Odors with low familiarity were selected to avoid implicit familiarity judgments. For training, three sets of nine low familiarity odorants and nine bottles with odorless air were used. In each set, the presentation order of stimuli was pseudorandomized, but identical for all subjects. Odorants were diluted to a concentration of 10% using mineral oil (Sigma Aldrich, France). For presentation, 5 ml of this solution was absorbed into compressed polypropylene filaments inside of a 100 ml white polyethylene squeeze-bottles equipped with a dropper (Osi, France). Stimulating and recording materials Odors were presented using an airflow olfactometer, which allowed synchronizing stimulation with breathing. The stimulation equipment was essentially the one used in a previous PET study (Royet et al., 1999), but adapted so as to avoid interference with the static magnetic field of the scanner (Royet et al., 2003). Briefly, the apparatus was split into two modules: the electronic part of the olfactometer positioned outside the magnet room (shielded with a Faraday cage), and the nonferrous (DuraluminR) air-dilution

Table 1 List of odorants selected for the Da, Db, Dc, Fa, Fb, and Fc epochs 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Da

Db

Dc

Fa

Fb

Fc

Plum

Sage Turpentine

Acacia

Blackcurrant

Apricot Bergamot Orange Diethyl ether Pine Bornyl acetate Caramel Grass Strawberry Pepper Oyster Gardenia Celery Hazelnut Tar Methyl acetate Biscuit Tobacco 3-Methyl anisol

Raspberry Tetralin Mint Honey Patchouli Jasmine Basil Toluene Green lily Banana Vanilla Lily Bitter almond Caprylic aldehyde Passion fruit Eucalyptus Coconut Clove

Citronella trans-2-Hexenal Cypress Geranium Vienna bread Anise Iris Incense Lavender Gingerbread Apple Garlic PPA Rose Thyme Lime 1-Octen-3-ol Camomile

4.29 (0.92) 3.09–5.17

4.56 (0.28) 4.03–4.91

4.38 (0.39) 3.41–4.69

4.37 (0.23) 3.89–4.61

6.53 (0.75) 5.13–7.27

6.28 (0.67) 4.89–7.24

6.24 (0.23) 4.96–6.91

Tarragon Parsley Jonquil

Guaiacol Camphor Neroli Musk Pine needle

Carrot 1,4-Dichlorobutane

EBA Butanol Eglantine

Cherry Acetol

Liqueur wine 2-Octanol

Liquorice

Unfamiliar odors Mean (SD) 4.00 (0.72) Range 3.14–5.11 Familiar odors Mean (SD) Range

Orange Acetophenone

4.11 (0.87) 3.19–5.16

2-Bromophenol Tangerine

Note. EBA, ethyl benzoyl acetate; PPA, phenyl propionaldehyde; SD, standard deviation. Italic, familiar odorants. Odorants were rated on a 10-point rating scale, with 1 representing low familiarity, and 10 representing high familiarity (see Royet et al., 1999).

1034

J. Plailly et al. / NeuroImage 24 (2005) 1032–1041

injection head placed within the stray-field of the magnet. Compressed air (10 l/min) was pumped into the olfactometer, and delivered continuously through a standard anesthesia mask. A detailed description was recently given (Vigouroux et al., in press). The ventral breathing rhythm of the subject was recorded with the aid of a foot bellows in polyvinyl chloride (Herga Electric Limited, Suffolk, UK) held on the stomach with a weaved cotton belt. Movements of the abdominal wall produced variations in the internal volume of the foot bellows. The flow was transformed into an electrical signal that was amplified, and then successively transmitted to the acquisition system and to a headphone via a voltage-tofrequency converter. The experimenter could therefore listen to the progressive frequency variations that accompanied respiration (high frequency during inspiration and low frequency during expiration). This method made respiration easily detected for the experimenter and allowed easy timing of the stimulus delivery. During scanning, the subject was instructed to avoid sniffing or blocking his breathing, but instead to breathe regularly, thus allowing the experimenter to anticipate the beginning of an inspiration phase. One stimulus was then injected into the olfactometer by squeezing one bottle into the injection head, so that the odor (or odorless air) was carried to the subject’s anesthesia mask. Subjects rated familiarity or judged odor presence by pressing one of two buttons. The response signal was then transmitted outside the radiofrequency shielded room by fiber optics to analogto-digital converters powered by nickel–cadmium batteries. Behavioral data were recorded on line (100 Hz sampling rate) using a NEC PC computer equipped with a digital acquisition board DAQCard-500 (National Instruments, USA). LabView 5.0 software (National Instruments) was used to acquire, store, and read data. Data analysis was performed with the WinDaq Waveform Browser 1.91 software (DataQ Instruments, USA). Experimental procedure A single functional run was presented in blocks that consisted of two olfactory conditions (F and D) alternating with odorless rest (R) epochs (Fig. 1). Each epoch lasted 60 s. Both F and D conditions were presented three times each, either three F followed by three D conditions or vice versa. Within the same condition, the presentation order of the three sets (a, b, c) was counterbalanced across subjects according to a balanced experimental (Latin square) design. For olfactory conditions, subjects were asked to judge whether or not they smelled an odor (D condition) or whether the odor was familiar or unfamiliar (F condition). Subjects were then asked to make a dyesT or dnoT rating using the two buttons with

Fig. 1. Experimental procedure showing the functional run including 12 epochs of 60 s each. Two olfactory conditions were performed: one detection condition with three epochs (Da, Db, Dc) and one familiarity condition with three epochs (Fa, Fb, Fc). Example of an epoch (Da) for which 15 stimuli (from S1 to S15) were delivered. R, rest.

their dominant hand. For half of the subjects, dyesT and dnoT responses were obtained with the index and the middle fingers, respectively. For the other half of the subjects, the meaning of the two key-press buttons was reversed. For the R condition, no stimulation was provided and the subjects were instructed not to use the key-press buttons. General instructions were provided to subjects before the functional run. During the run, and 3 s before each experimental condition (F, D or R), subjects were instructed orally by means of specific keywords (dfamiliarity,T ddetection,T and drestT) which task was to be performed next. Subjects wore earplugs to protect hearing from excessive scanner noise and kept their eyes closed during scanning. The day before the fMRI examination, subjects were trained outside the MR facility to breathe regularly, to detect odorants without sniffing during normal inspiration, and to give the most rapid possible response (odor vs. no odor) using the buttons. Imaging parameters Functional MR imaging was performed on a 1.5-T MR imager (Philips NT). Twenty-five adjacent 5-mm-thick axial slices were imaged. The imaging volume covered the whole brain and was oriented parallel to the bicommissural plane. The image planes were positioned from scout images acquired in the sagittal plane. A 3D three-shot PRESTO MR imaging sequence (Liu et al., 1993) was used with the following parameters: TR = 26 ms, TE = 38 ms, flip angle = 148, field-of-view = 256  205 mm2, imaging matrix = 64  51 (voxel size of 4  4  5 mm3). This sequence is less prone to magnetic susceptibility artifacts than the usual echo planar imaging (EPI) sequence (Van Gelderen et al., 1995), particularly in the OFC and mesial temporal region (Zald and Pardo, 2000). During the functional run, the volume of interest was scanned 144 times successively. The signal was averaged three times, leading to an acquisition time per volume of 5 s. A high-resolution anatomical 3D T1-weighted MR scan was acquired before the functional run. Data processing and statistical analyses Functional images were analyzed using SMP99 (Wellcome Department of Cognitive Neurology, London, UK). Image processing included interscan realignment, spatial normalization to stereotactic space as defined by the Montreal Neurological Institute (MNI) reference brain template, and image smoothing with a threedimensional Gaussian kernel (FWMH: 8  8  10 mm3) to overcome residual anatomical variability during group analysis, increase the signal-to-noise ratio and conform to Random Field Theory assumptions underlying the statistical analysis (Friston et al., 1995a). A boxcar reference function was convolved with SPM99’s dcanonical hemodynamicT response function. A high-pass filter (cutoff frequency of 1/720 Hz) was used to eliminate instrumental and physiological very low frequency signal fluctuations. Global differences in BOLD signal were covaried out from all voxels, and comparisons across conditions were effected with t tests. The statistical significance of signal differences was assessed through Z scores in an omnibus sense, using an uncorrected height threshold (P b 0.001). Only clusters of more than 10 adjacent activated voxels were taken into account as a significant hemodynamic response (Z N 3.20 at voxel level). Duvernoy’s (1991) and Mai et al.’s (1997) anatomic atlases were used to localize and describe activated anatomic regions as the often used Talairach’s atlas (Talairach and Tournoux, 1988) describes subcortical and

J. Plailly et al. / NeuroImage 24 (2005) 1032–1041

1035

Table 2 Behavioral data recorded for the three sets of odors (a, b, c) during the detection and familiarity judgment tasks Parameter

Task

Mean number of stimulations

Detection Familiarity Detection Detection

Response accuracy Reaction time

Familiarity Proportion of familiar odors

Response

Yes No Yes No

Familiarity

Set a 13.33 12.89 0.962 1.597 2.125 2.141 2.220 0.545

Set b F F F F F F F F

1.87 2.20 0.050 0.462 0.602 0.668 0.652 0.214

14.11 F 13.33 F 0.857 F 1.698 F 2.092 F 1.985 F 2.346 F 0.520 F

Set c 3.02 2.24 0.104 0.624 0.758 0.540 0.730 0.188

13.89 13.44 0.930 1.625 2.033 2.153 2.359 0.509

F F F F F F F F

1.96 2.65 0.078 0.487 0.620 0.635 0.819 0.180

For reaction time, data are given according to whether the odors were detected or not (Yes or No), or whether they were recognized as being familiar or not (Yes or No).

limbic olfactory regions with much less detail. Activated areas were indicated using the MNI coordinate system. Specific effects for the familiarity judgment task were calculated by comparing the signals during the F and D conditions using the general linear model (Friston et al., 1995b). Intrasubject analyses were first performed, followed by a random effects analysis which extent statistical inferences into the healthy population. This twostage analysis accounted first for intrasubject variance (scan-toscan), and second for intersubject variance. In the first step, scan-toscan variance was separately modeled for each subject by creating a summary contrast image from weighted parameter estimates that represented each scan condition. In the second step, these contrast images were then analyzed using a basic model one-sample t tests to assess the F–D contrast against a null hypothesis. A cluster analysis was further performed to compare activation between the right and left PC. A region of interest (ROI) corresponding to the right PC was defined by selecting an 8mm-diameter sphere centered on coordinates (30, 2, 16) of the activation cluster obtained in the F–D contrast image from the group analysis. An identical ROI was centered on the contralateral coordinate in the left hemisphere ( 30, 2, 16). Using the MarsBar SPM toolbox (Brett et al., 2002), we then obtained a mean activity level within both ROIs for each one of 12 subjects. A statistical analysis was then performed to compare the activity levels of left and right PC.

indicating that odors of the Da and Dc sets were more easily detected than those of the Db set. For the F condition, the number of odorants judged as being familiar or unfamiliar by the subjects was determined for the three odor sets (Fa, Fb, and Fc). For each subject, data were normalized with respect to the number of stimulations per epoch. The mean number of stimulations and the mean ratios of familiar odorants delivered per epoch are given in Table 2. A one-way ANOVA with repeated measurements performed on the ratios of familiar odorants showed no significant effect of set factor [ F(2,22) = 0.455, P = 0.640], indicating that the same proportion of familiar odorants was found in the three odor sets. Subjects’ reaction times for odor detection and familiarity judgments were also calculated (Table 2). The number of stimulations delivered per epoch depended on the subject’s breathing rhythm. As this factor could affect reaction times by creating a shorter interstimulus response period in those who breathe more rapidly, the data were normalized with respect to the number of stimulations, and analyzed as a function of task, odor sets, and response (Yes or No) factors. A three-way ANOVA with repeated measurements showed a significant effect of the judgment task [ F(1,11) = 33.380, P b 0.0001], of response [ F(1,11) = 65.965, P b 0.0001], but no significant effect of odor set factor [ F(2,22) = 1.020, P = 0.3772]. Significant task  response [ F(2,22) = 10.153, P = 0.0087] (see Fig. 2) and task  set 

Results Behavioral data Response accuracy was determined for the detection task only, since the familiarity judgment depends on personal experience. During the scanning day, two subjects scored very low response accuracy (65% and 68%, respectively), due to a very small number of odors detected (29% and 41%, respectively). The aberration of both these values was rated with the Grubbs’ test (Dagnelie, 1975), which indicated that they were indeed outliers (t = 5.414 and t = 5.052, respectively, for a theoretical value of t 0.99815 = 3.663). One of these subjects also did not provide behavioral responses during the three epochs of the familiarity task. These two subjects were therefore excluded from further analysis. For the 12 remaining subjects, the mean numbers and response accuracies of odors delivered per epoch for the three odor sets (Da, Db, Dc) of the detection task are given in Table 2. A one-way ANOVA with repeated measurements performed on response accuracy showed a significant main effect of set factor [ F(2,22) = 5.985, P = 0.0084],

Fig. 2. Reaction times represented as a function of the olfactory task (Detection vs. Familiarity) and of the type of response (Yes vs. No). Data were normalized with respect to the number of stimulations per epoch. The vertical bars show the standard errors of the means. *, significant difference (P b 0.007).

1036

J. Plailly et al. / NeuroImage 24 (2005) 1032–1041

response [ F(2,22) = 3.833, P = 0.0373] interactions were noted. Mean comparisons indicated that reaction times in the D condition were longer when no odor was detected (P b 0.0001), longer in the F condition when odors were perceived as more unfamiliar (P b 0.0070), and longer with unfamiliar odors during familiarity judgments (as compared to detection judgments; P b 0.0001). Breathing data Breathing changes could be expected as a function of epoch, as during F-epochs odors were delivered at each inspiration, and during D epochs odors were delivered on 50% of inspirations. Inspiratory airflow measures were therefore analyzed as a function of task (Detection and Familiarity), three epochs (a, b, and c), and subject responses (Yes and No) (Fig. 3). A three-way ANOVA with repeated measurements showed no significant effect of task [ F(1,22) = 2.169, P = 0.169], epochs [ F(2,22) = 0.415, P = 0.665], and response [ F(1,22) = 1.409, P = 0.260] factors. fMRI data F–D contrast Familiarity-specific responses (F–D) were present in the right temporal piriform area (30, 2, 16; Z = 3.75), spanning the cortico-amygdaloid transition area, the preamygdalar claustrum, the periamygdalar area, and the lateral amygdaloid nucleus (22, 16, 10; Z = 3.85; Table 3 and Fig. 4A). PC activation was not detected in the left hemisphere (using the same statistical significance threshold), but strong activation was present in the amygdala ( 24, 6, 12; Z = 4.29) spanning approximate areas for the basomedial, basolateral, central, lateral and medial amygdaloid nuclei, and the anterior amygdaloid area. Familiarityrelated activation in the right PC extended into the right midfusiform gyrus (40, 40, 16; Z = 3.93) and hippocampal region (Fig. 4B; 22, 22, 8; Z = 3.84). The familiarity judgment task also activated the left cingulate gyrus ( 8, 16, 40; Z = 4.21) and the left inferior frontal gyrus in its opercular part ( 50, 28, 2; Z = 3.27), as depicted in Fig. 4C. Finally, we noted significant activation in the left occipital gyrus ( 16, 100, 14; Z = 3.79) and the right middle frontal gyrus (44, 20, 18; Z = 3.40).

Table 3 Areas activated in the F–D contrast Brain region

L/R

k

Z value

MNI coordinates

Amygdala Cingulate gyrus Mid-fusiform gyrus Amygdala Hippocampal region (CA3) Temporal piriform cortex Parahippocampal gyrus Posterior Insula Superior occipital gyrus Inferior frontal gyrus, pars orbitalis

L L R R R

54 67 432

4.29 4.21 3.93 3.85 3.84

24 8 40 20 22

6 16 40 16 20

12 40 16 10 8

R

3.75

30

2

16

R

3.61

42

16

10

x

y

z

R L

62

3.29 3.79

46 16

10 100

6 14

L

13

3.27

50

28

2

Note. F, familiarity; D, detection; L, left; R, right; k, size of the cluster in number of connected voxels; x, y, z, MNI coordinates in mm of the maximum in the Montreal Neurological Institute Brain template; CA, Cornu Ammonis.

Comparison of activations between the right and left PC Mean activity levels were measured in the right and left PC using ROIs. A one-way ANOVA with repeated measurements on these data showed significantly higher right than left PC activation [ F(1,11) = 4.846, P = 0.0499].

Discussion The aim of this fMRI study was to determine the cerebral regions that mediate judgments of odor familiarity. With such an approach, we identified odor-evoked neural responses in several olfactory and limbic regions, including piriform cortex, amygdala, hippocampus, the pars orbitalis of the inferior frontal gyrus, and the mid-fusiform gyrus. Activation of the mesial temporal region

Fig. 3. Inspiratory airflow as a function of the olfactory task (Detection vs. Familiarity) and of the type of response (Yes vs. No). Data were normalized with respect to the number of stimulations per epoch. The vertical bars show the standard errors of the means.

Olfactory-related activation of mesial temporal regions remains inconsistent across studies. Whereas some authors have reported activation in PC in PET studies (e.g., Kareken et al., 2001, 2003, 2004; Savic et al., 2000; Small et al., 1997; Zatorre et al., 1992), we did not find any mesial temporal activation in our previous studies (Royet et al., 1999, 2001). Piriform cortex activation has also been inconsistently detected across fMRI studies (e.g., Sobel et al., 1998; Yousem et al., 1999). A well-known problem with the EPI pulse sequence usually applied in fMRI is magnetic susceptibility artefact, which induces signal loss in these regions (Zald and Pardo, 2000). The PRESTO sequence used in the current study appears to reduce these artifacts and to provide stronger signal in the regions affected by susceptibility differences. Habituation may also contribute to signal loss in these ventral regions, particularly with blocked designs and the use of only one or two odorants (Poellinger et al., 2001; Sobel et al., 2000). Since we used a different odorant on each breathing cycle (from 12 to 20 different odorants per 60-s epoch, depending on the duration of breathing cycle of the subject), we minimized both self-adaptation

J. Plailly et al. / NeuroImage 24 (2005) 1032–1041

1037

Fig. 4. Localization of task-specific activations in the F–D contrasts. (A) Piriform cortex; (B) hippocampus; (C) inferior frontal gyrus. Neural responses are overlaid on coronal and axial sections from a subject’s normalized T1-weighted brain image. Clusters were thresholded at t = 3.10.

(Engen, 1982) and sensory habituation (De´monet et al., 1993). As in our previous fMRI study (Royet et al., 2003), the present data suggest that the pulse sequence and stimulation paradigm provided sufficient sensitivity to detect activation in primary olfactory areas. A PC activation that is stronger in the F than D condition could be explained by different factors. A possible confound lies first of all in the higher proportion of odorous stimulations delivered in the F than D condition. We decided for this particular design because we considered it was important to keep the subjects’ attention focused during the detection task (in which they had to respond dyesT and dnoT equivalently). At the same time, if we had presented blanks during the familiarity task, this might have confused the subjects and this would have introduced a third type of event (no odor in addition to the familiar and the unfamiliar

odors). Further experiments are therefore needed to rule out this specific confound. It would be of interest, in particular, to design an event-related fMRI study in which familiar and unfamiliar odors would be distinguished as discrete events in both conditions. Unfamiliar odors of the F and D conditions could then be contrasted and any PC activation could then be specifically associated with the F task. It should nevertheless be noted that the piriform cortex may become activated in retrieving odor associations alone, without any direct chemosensory stimulation (Gottfried et al., 2004). This strongly suggests that the task performed by subjects may be a decisive factor in activating the piriform cortex. A second source of potential confound may lie in the different respiratory patterns across conditions, as it has

1038

J. Plailly et al. / NeuroImage 24 (2005) 1032–1041

previously been shown that sniffing alone can induce piriform activation (Sobel et al., 1998; but also see Kareken et al., 2004 for contradictory results), and that odor imagery alone can lead to breathing differences (Bensafi et al., 2003). In the present study, however, inspiratory airflow was not higher in the F condition than in the D condition. Finally, a third source of potential confound is identified in the differences in odor familiarity, since odors selected a priori as being familiar were only delivered in the F condition. Nevertheless, the same number of a priori unfamiliar odors were presented in both conditions. We then noted that reaction times were much longer when the unfamiliar odors were presented in the F condition, and even longer than reaction times of familiar odors presented in this same condition. This suggests that subjects performed different judgments across conditions, and that activation differences in PC could be better explained by the type of task than by perceptual differences in odor familiarity. This could also explain why Savic and Berglund (2004) did not find differential activation in PC when subjects passively smelled familiar or unfamiliar odorants. It nevertheless stands to reason, however, that the task (when not explicitly required by the experiment) should be facilitated by the intrinsic properties of the odors themselves: That is, highly familiar odors would more easily facilitate familiarity judgments, and highly pleasant or unpleasant odors would more easily encourage emotional judgments. Involvement of piriform cortex in recognition memory Recognition memory is known to involve two different processes: familiarity and recollection (Bogacz et al., 2001; Mandler, 1980; Rajaram, 1998). According to the ddual process theoryT, processes underlying familiarity are perceptual in nature, and those subserving recollection include the retrieval of contextual information. Lehrner et al. (1999) demonstrated that these two forms of recognition memory processes also exist in olfaction. In other words, familiarity judgments are made on the basis of feelings devoid of specific information about the encoding episode, and thus relate to implicit memory. By contrast, recollection is more directly tied to specific events, and thus relates to explicit memory. To illustrate these concepts, Bogacz et al. (2001) note that, b. . .it is not uncommon to be able to recognize that a person is familiar to us even though we cannot immediately recollect anything more about the person or our previous encounters with themQ (Bogacz et al., 2001). Since familiarity judgments are inherent in recognition memory, our results relate to previous findings in humans indicating that PC is involved in long-term odor recognition memory (Dade et al., 2002). More recently, Gottfried et al. (2004) further reported that PC responds to non-olfactory stimuli with which odors were previously associated. Further experiments are needed to compare activation produced by different memory processes. Familiarity and recollection could involve slightly different neural networks, but the present experimental design cannot distinguish between them. Our findings nevertheless cohere with a large body of research using animal models, and lend support to the theory that the piriform cortex is involved in learning- and memory-related processes (e.g., Datiche et al., 2001; Schoenbaum and Eichenbaum, 1995). For instance, synaptic potentiation has been shown to occur in rat PC in vitro (e.g., Jung et al., 1990; Saar et al., 2002) and in vivo at the conclusion of learning (Litaudon et al.,

1997; Roman et al., 1993). These findings are thus consistent with models demonstrating that the primary olfactory cortex is a parallel-distributed architecture characteristic of associative memory systems (e.g., Bower, 1991; Haberly, 2001; Haberly and Bower, 1989). Lateralization of familiarity judgment process The current study shows that activation of the right PC was stronger than that of the left PC. This is in line with the results of a large body of other research. In a monorhinal odor recognition task, Savic et al. (2000) noted significant right, but not left, piriform activity. Although Dade et al. (2002) did not explicitly report hemispheric asymmetry for long-term olfactory memory, their results distinctly indicated strong activation in right OFC, and more activation in right than left PC. Interestingly, Gottfried et al.’s study of cross-modal visual–olfactory associations also showed unilateral right PC activation, and in this case without direct olfactory stimulation. Using behavioral measures, Broman et al. (2001) finally observed that odors presented to the right nostril were rated as more familiar than odors presented to the left nostril. They also reported that episodic recognition via the right nostril tended to nominally have more dknowT responses and fewer drememberT responses than did odors recognized via the left nostril, which is in keeping with the right-nostril advantage for familiarity. Taken together, these findings are consistent with the notion that left temporal lobe structures mediate processing of distinctiveness (i.e., what clearly distinguishes one percept from another), whereas right temporal lobe structures subserve processes underlying perceptual fluency (i.e., what involves perceptual analysis of the surface features of an item; Blaxton and Theodore, 1997; Rajaram, 1998). Such a perceptual analysis of surface features is especially observed for odors which are intrinsically difficult to name (Lawless and Engen, 1977). Findings in brain-damaged patients and results from neuroimaging studies converge with these data. For instance, epilepsy patients with left temporal lobe lesions who were asked to recognize previously seen abstract designs provided more dknowT than drememberT responses, whereas right temporal lesioned patients showed the opposite pattern (Blaxton and Theodore, 1997). Along the same line, Henson et al. (1999) explored word recognition with fMRI and showed a dissociation whereby a dknowT judgment induced right frontal activation, and a drememberT judgment induced left frontal activation. In conclusion, the present data, which are consistent with our previous work in PET, indicate a preferential involvement of the right hemisphere in familiarity judgments (Royet et al., 1999, 2001). An intriguing result in the present study is the lack of activation in right OFC. When specifically examining activation resulting from familiarity versus rest, and from detection versus rest, we nevertheless saw activation in the right OFC for both contrasts (44, 32, 16; Z = 3.62 and 46, 30, 12; Z = 2.79, respectively). This could explain the lack of activation when we compared images in the familiarity versus detection contrast. Participation of the hippocampal region, inferior frontal and mid-fusiform gyri in modality-independent mnemonic and semantic processing Since only a few authors have previously reported hippocampal activations in olfaction studies (e.g., Kareken et al., 2003; Suzuki

J. Plailly et al. / NeuroImage 24 (2005) 1032–1041

et al., 2001), the right hippocampal activation during the familiarity judgment task was not anticipated in the present study. Inconsistent activation of the hippocampal formation is not specific to olfaction, and has been characteristic of other studies of memory (Andreasen et al., 1995; Shallice et al., 1994; Tulving et al., 1996). Our data are nevertheless consistent with a recent finding showing hippocampal activation during the retrieval of olfactory episodic memories (Gottfried et al., 2004). Lesion studies recently examined whether the brain structures that comprise the medial temporal lobe memory system (i.e., the hippocampal and parahippocampal regions) differ in how they support recollective and familiarity components (Manns et al., 2003; Yonelinas et al., 2002). Our data do not permit distinguishing these aspects of memory, as the subjects may well have had consciously evoked memories from the odorants. They are, however, consistent with the idea that both regions probably contribute to olfactory recognition memory. Activation of the left inferior frontal gyrus, in the pars orbitalis, during familiarity judgments further supports the hypothesis that this region is involved in the selection and integration of semantic information in a modality-independent manner (Homae et al., 2002; Kareken et al., 2003). In a recent study, Savic and Berglund (2004) found that left frontal and right parahippocampal region activation positively correlated with familiarity ratings, showing the engagement of semantic circuits during passive smelling of familiar odorants. Along the same line, the mid-fusiform gyrus activation in the current study might lead to similar interpretations, since it has further been associated with visual, tactile and auditory recognition and categorization of objects (Adams and Janata, 2002; Joseph and Gathers, 2003; Stoeckel et al., 2003). Its involvement in olfactory object recognition therefore reinforces the idea of the polymodal nature of this area (Adams and Janata, 2002) and its role in semantic processing (Price, 2000; Wagner et al., 1998).

Conclusion Complementing previous PET studies that demonstrate right OFC involvement in odor familiarity judgments, the present fMRI study shows that right PC is also activated during this task, an activation that may be related to olfactory recognition memory. In previous fMRI and PET studies, we demonstrated that a neural network in the left hemisphere, involving the OFC and primary olfactory areas, mediated olfactory hedonic perception (Royet et al., 2000, 2003). It thus appears that odor processing activates a large neural network involving both hemispheres. Nevertheless, this network possesses hemispheric predominance depending on the type of olfactory task performed (see Royet and Plailly, 2004, for review). The present data provide further evidence that the right hippocampal region, left inferior frontal gyrus and mid-fusiform gyrus take part in recognition memory, likely cross-modally to assist in gathering relevant associations to enable identification of olfactory percepts.

Acknowledgments We thank the technical team (M. Vigouroux, B. Bertrand, and V. Farget) for designing and building the stimulation and recording materials and J.P. Lomberget and M.B. Sanglerat for medical

1039

examinations of subjects participating in the study. We are grateful to the companies Givaudan, International Flavors and Fragrances, Lenoir, Davenne, and Perlarom for supplying the odorants used in this study. This work was supported by research grants from the dRe´gion Rhoˆne-AlpesT and the dGIS Sciences de la Cognition,T the dCentre National de la Recherche Scientifique,T and the dUniversite´ Claude-Bernard de Lyon.T

References Adams, R.B., Janata, P., 2002. Comparison of neural circuits underlying auditory and visual object categorization. NeuroImage 16, 361 – 377. Anderson, A.K., Christoff, K., Stappen, I., Panitz, D., Ghahremani, D.G., Glover, G., Gabrieli, J.D., Sobel, N., 2003. Dissociated neural representations of intensity and valence in human olfaction. Nat. Neurosci. 6, 196 – 202. Andreasen, N.C., O’Leary, D.S., Cizadlo, T., Arndt, S., Rezai, K., Watkins, G.L., Ponto, L.L., Hichwa, R.D., 1995. Remembering the past: two facets of episodic memory explored with positron emission tomography. Am. J. Psychiatry 152, 1576 – 1585. Bensafi, M., Porter, J., Pouliot, S., Mainland, J., Johnson, B., Zelano, C., Young, N., Bremner, E., Aframian, D., Khan, R., Sobel, N., 2003. Olfactomotor activity during imagery mimics that during perception. Nat. Neurosci. 6, 1142 – 1144. Blaxton, T.A., Theodore, W.H., 1997. The role of the temporal lobes in recognizing visuospatial materials: remembering versus knowing. Brain Cogn. 35, 5 – 25. Bogacz, R., Brown, M.W., Giraud-Carrier, C., 2001. Model of familiarity discrimination in the perirhinal cortex. J. Comput. Neurosci. 10, 5 – 23. Bower, J.M., 1991. Piriform cortex and olfactory recognition. In: Davis, J.D., Eichenbaum, H. (Eds.), Olfaction: a Model System for Computational Neuroscience. MIT Press, Cambridge, MA, pp. 266 – 285. Brett, M., Anton, J.L., Valabregue, R., Poline, J.B., 2002. Region of interest analysis using an SPM toolbox [abstract]. The 8th International Conference on Functional Mapping of the Human Brain, June 2–6, 2002, Sendai, Japan. Available on CD-ROM in NeuroImage 16 (2). Broca, P., 1863. Localisation des fonctions ce´re´brales. Sie`ge de la faculte´ du langage articule´. Bull. Soc. Anthropol. Paris 4, 200 – 204. Broman, D.A., Olsson, M.J., Nordin, S., 2001. Lateralization of olfactory cognitive functions: effects of rhinal side of stimulation. Chem. Senses 26, 1187 – 1192. Dade, L.A., Jones-Gotman, M., Zatorre, R.J., Evans, A.C., 1998. Human brain function during odor encoding and recognition. A PET activation study. Ann. N. Y. Acad. Sci. 855, 572 – 574. Dade, L.A., Zatorre, R.J., Jones-Gotman, M., 2002. Olfactory learning: convergent findings from lesion and brain imaging studies in humans. Brain 125, 86 – 101. Dagnelie, P., 1975. The´ orie et Me´ thodes Statistiques, Les Presses Agronomiques de Gembloux, A.S.B.L 2. Datiche, F., Roullet, F., Cattarelli, M., 2001. Expression of Fos in the piriform cortex after acquisition of olfactory learning: an immunohistochemical study in the rat. Brain Res. Bull. 55, 95 – 99. De´ monet, J.F., Wise, R., Frackowiak, R.S.J., 1993. Les fonctions linguistiques explore´es en tomographie par e´mission de positons. Me´d./Sci. 9, 934 – 942. Duvernoy, H.M., 1991. The Human Brain-Surface, Three Dimensional Sectional Anatomy and MRI. Springer, Wien. Engen, T., 1982. The Perception of Odors. Academic Press, New York. Friston, K.J., Ashburner, J., Frith, C.D., Poline, J.B., Healther, J.D., Frackowiak, R.S.J., 1995a. Spatial registration and normalisation of images. Hum. Brain Mapp. 3, 165 – 189. Friston, K.J., Holmes, A.P., Worsley, K.J., Poline, J.B., Frith, C.D., Frackowiak, R.S.J., 1995b. Statistical parametric maps in functional imaging: a general linear approach. Hum. Brain Mapp. 2, 189 – 210.

1040

J. Plailly et al. / NeuroImage 24 (2005) 1032–1041

Gottfried, J.A., Deichmann, R., Winston, J.S., Dolan, R.J., 2002. Functional heterogeneity in human olfactory cortex: an event-related functional magnetic resonance imaging study. J. Neurosci. 22, 10819 – 10828. Gottfried, J.A., Smith, A.P.R., Rugg, M.D., Dolan, R.J., 2004. Remembrance of odors past: human olfactory cortex in cross-modal recognition memory. Neuron 42, 687 – 695. Haberly, L.B., 2001. Parallel-distributed processing in olfactory cortex: new insights from morphological and physiological analysis of neuronal circuitry. Chem. Senses 26, 551 – 576. Haberly, L.B., Bower, J.M., 1989. Olfactory cortex: model circuit for study of associative memory. TINS 12, 258 – 264. Henson, R.N.A., Rugg, M.D., Shallice, T., Josephs, O., Dolan, R.J., 1999. Recollection and familiarity in recognition memory: an eventrelated functional magnetic resonance imaging study. J. Neurosci. 19, 3962 – 3972. Homae, F., Hashimoto, R., Nakajima, K., Miyashita, Y., Sakai, K.L., 2002. From perception to sentence comprehension: the convergence of auditory and visual information of language in the left inferior frontal cortex. NeuroImage 16, 883 – 900. Joseph, J.E., Gathers, A.D., 2003. Effects of structural similarity on neural substrates for object recognition. Cogn. Affect. Behav. Neurosci. 3, 1 – 16. Jung, M.W., Larson, J., Lynch, G., 1990. Long-term potentiation of monosynaptic EPSPs in rat piriform cortex in vitro. Synapse 6, 279 – 283. Kareken, D.A., Doty, R.L., Moberg, P.J., Mosnik, D., Hsing Chen, S., Farlow, M.R., Hutchins, G.D., 2001. Olfactory-evoked regional cerebral blood flow in Alzheimer’s disease. Neuropsychology 15, 18 – 29. Kareken, D.A., Mosnik, D.M., Doty, R.L., Dzemidzic, M., Hutchins, G.D., 2003. Functional anatomy of human odor sensation, discrimination, and identification in health and aging. Neuropsychology 17, 482 – 495. Kareken, D.A., Sabri, M., Radnovich, A.J., Claus, E., Foresman, B., Hector, D., Hutchins, G.D., 2004. Functional anatomy of human odor sensation, discrimination, and identification in health and aging. NeuroImage 22, 456 – 465. Lawless, H., Engen, T., 1977. Associations to odors: interference, mnemonics, and verbal labeling. J. Exp. Psychol. Hum. Learn. Mem. 3, 52 – 59. Lehrner, J.P., Walla, P., Laska, M., Deecke, L., 1999. Different forms of human odor memory: a developmental study. Neurosci. Lett. 272, 17 – 20. Litaudon, P., Mouly, A.M., Sullivan, R., Gervais, R., Cattarelli, M., 1997. Learning-induced changes in rat piriform cortex activity mapped using multisite recording with voltage sensitive dye. Eur. J. Neurosci. 9, 1593 – 1602. Liu, G., Sobering, G., Duyn, J., Moonen, C., 1993. A functional MRI technique combining principles of echo-shifting with a train of observations (PRESTO). Magn. Reson. Med. 30, 764 – 768. Mai, J.K., Assheuer, J., Paxinos, G., 1997. Atlas of the Human Brain. Academic Press, San Diego. Mandler, G., 1980. Recognizing: the judgment of previous occurrence. Psychol. Rev. 87, 252 – 271. Manns, J.R., Hopkins, R.O., Reed, J.M., Kitchener, E.G., Squire, L.R., 2003. Recognition memory and the human hippocampus. Neuron 37, 171 – 180. Poellinger, A., Thomas, R., Lio, P., Lee, A., Makris, N., Rose, B.R., Kwong, K.K., 2001. Activation and habituation in olfaction—An fMRI study. NeuroImage 13, 547 – 560. Price, C.J., 2000. The anatomy of language: contributions from functional neuroimaging. J. Anat. 3, 335 – 359. Rajaram, S., 1998. The effects of conceptual salience and perceptual distinctiveness on conscious recollection. Psychon. Bull. Rev. 5, 71 – 78. Roman, F.S., Chaillan, F.A., Soumireu-Mourat, B., 1993. Long-term potentiation in rat piriform cortex following discrimination learning. Brain Res. 601, 265 – 272.

Royet, J.P., Koenig, O., Gregoire, M.C., Cinotti, L., Lavenne, F., Le Bars, D., Costes, N., Vigouroux, M., Farget, V., Sicard, G., Holley, A., Mauguie`re, F., Comar, D., Froment, J.C., 1999. Functional anatomy of perceptual and semantic processing for odours. J. Cogn. Neurosci. 11, 94 – 109. Royet, J.P., Zald, D., Versace, R., Costes, N., Lavenne, F., Koenig, O., Gervais, R., 2000. Emotional responses to pleasant and unpleasant olfactory, visual, and auditory stimuli: a positron emission tomography study. J. Neurosci. 20, 7752 – 7759. Royet, J.P., Hudry, J., Zald, D.H., Godinot, D., Gre´goire, M.C., Lavenne, F., Costes, N., Holley, A., 2001. Functional neuroanatomy of different olfactory judgments. NeuroImage 13, 506 – 519. Royet, J.P., Plailly, J., Delon-Martin, C., Kareken, D.A., Segebarth, C., 2003. Functional anatomy of the emotional responses to odors: Influence of hedonic valence, hedonic judgment, handedness, and gender. NeuroImage 20, 713 – 728. Royet, J.P., Plailly, J., 2004. Lateralization of olfactory processes. Review. Chem. Senses 29, 731 – 745. Saar, D., Grossman, Y., Barkai, E., 2002. Learning-induced enhancement of postsynaptic potentials in pyramidal neurons. J. Neurophysiol. 87, 2358 – 2363. Savic, I., Berglundd, H., 2004. Passive perception of odors and semantic circuits. Hum. Brain Mapp. 21, 271 – 278. Savic, I., Gulyas, B., Larsson, M., Roland, P., 2000. Olfactory functions are mediated by parallel and hierarchical processing. Neuron 26, 735 – 745. Schoenbaum, G., Eichenbaum, H., 1995. Information coding in the rodent prefrontal cortex. I. Single-neuron activity in orbitofrontal cortex compared with that in pyriform cortex. J. Neurophysiol. 74, 733 – 750. Shallice, T., Fletcher, P., Frith, C.D., Grasby, P., Frackowiak, R.S., Dolan, R.J., 1994. Brain regions associated with acquisition and retrieval of verbal episodic memory. Nature 368, 633 – 635. Small, D.N., Jones-Gotman, M., Zatorre, R.J., Petrides, M., Evan, A.C., 1997. Flavor processing: more than the sum of its parts. NeuroReport 8, 3913 – 3917. Sobel, N., Prabhakaran, V., Desmond, J.E., Glover, G.H., Goode, R.L., Sullivan, E.V., Gabrieli, J.D.E., 1998. Sniffing and smelling: separate subsystems in the human olfactory cortex. Nature 392, 282 – 286. Sobel, N., Prabhakaran, V., Zhao, Z., Desmond, J.E., Glover, G.H., Sullivan, E.V., Gabrieli, J.D.E., 2000. Time-course of odorant-induced activation in the human primary olfactory cortex. J. Neurophysiol. 82, 537 – 551. Soros, P., Knecht, S., Imai, T., Gurtler, S., Lutkenhoner, B., Ringelstein, E.B., Henningsen, H., 1999. Cortical asymmetries of the human somatosensory hand representation in right-and left-handers. Neurosci. Lett. 271, 89 – 92. Stoeckel, M.C., Weder, B., Binkofski, F., Buccino, G., Shah, N.J., Seitz, R.J., 2003. A fronto-parietal circuit for tactile object discrimination: an event-related fMRI study. NeuroImage 19, 1103 – 1114. Suzuki, Y., Critchley, H.D., Suckling, J., Fukuda, R., Williams, S.C.R., Andrew, C., Howard, R., Ouldred, E., Bryant, C., Chir, B., Swift, C.G., Jackson, S.H.D., 2001. Functional magnetic resonance imaging of odour identification: the effect of aging. J. Gerontol. 56, 756 – 760. Tulving, E., Markowitsch, H.J., Craik, F.I.M., Habib, R., Houle, S., 1996. Novelty and familiarity activations in PET studies of memory encoding and retrieval. Cereb. Cortex 6, 71 – 79. Van Gelderen, P., Ramsey, N.F., Liu, G., Duyn, J.H., Frank, J.A., Weinberger, D.R., Moonen, C.T., 1995. Three-dimensional functional magnetic resonance imaging of human brain on a clinical 1.5-T scanner. Proc. Natl. Acad. Sci. U. S. A. 92, 6906 – 6910. Vigouroux, M., Bertrand, B., Farget, V., Plailly, J., Royet, J.P., 2004. A stimulation method using odors suitable for PET and fMRI studies with recording of physiological and behavioral signals. J. Neurosci. Methods (in press). Wagner, A.D., Schacter, D.L., Rotte, M., Koutstaal, W., Maril, A., Dale, A.M., Rosen, B.R., Buckner, R.L., 1998. Building memories: remembering and forgetting of verbal experiences as predicted by brain activity. Science 281, 1188 – 1191.

J. Plailly et al. / NeuroImage 24 (2005) 1032–1041 Weintraub, S., Mesulam, M.M., 1987. Right cerebral dominance in spatial attention. Further evidence based on ipsilateral neglect. Arch. Neurol. 44, 621 – 625. Yonelinas, A.P., Kroll, N.E., Quamme, J.R., Lazzara, M.M., Sauve, M.J., Widaman, K.F., Knight, R.T., 2002. Effects of extensive temporal lobe damage or mild hypoxia on recollection and familiarity. Nat. Neurosci. 5, 1236 – 1241. Yousem, D.M., Williams, S.C.R., Howard, R.O., Andrew, C., Simmons, A., Allin, M., Geckle, R.J., Suskind, D., Bullmore, E.T., Brammer, M.J., Doty, R.L., 1997. Functional MR imaging during odour stimulation: preliminary data. Neuroradiology 204, 833 – 838. Yousem, D.M., Maldjian, J.A., Siddiqi, F., Hummel, T., Alsop, D.C., Geckle, R.J., Bilker, W.B., Doty, R.L., 1999. Gender effects on odorstimulated functional magnetic resonance imaging. Brain Res. 818, 480 – 487.

1041

Zald, D.H., Pardo, J.V., 1997. Emotion, olfaction, and the human amygdala: amygdala activation during aversive olfactory stimulation. Proc. Natl. Acad. Sci. U. S. A. 94, 4119 – 4124. Zald, D.H., Pardo, J.V., 2000. Functional neuroimaging of the olfactory system in humans. Int. J. Psychophysiol. 36, 165 – 181. Zald, D.H., Donndelinger, M.J., Pardo, J.V., 1998. Elucidating dynamic brain interactions with across-subjects correlational analyses of positron emission tomographic data: the functional connectivity of the amygdala and orbitofrontal cortex during olfactory tasks. J. Cereb. Blood Flow Metab. 18, 896 – 905. Zatorre, R.J., Jones-Gotman, M., Evans, A.C., Meyer, E., 1992. Functional localization and lateralization of human olfactory cortex. Nature 360, 339 – 340. Zatorre, R.J., Belin, P., Penhune, V.B., 2002. Structure and function of auditory cortex: music and speech. Trends Cogn. Sci. 6, 37 – 46.