Functional Neuroanatomy of Different Olfactory Judgments Jean P. Royet,* Julie Hudry,* David H. Zald,† Damien Godinot,* Marie C. Gre ´goire,‡ Franck Lavenne,‡ Nicolas Costes,‡ and Andre ´ Holley* *Neurosciences and Sensory Systems, CNRS UMR 5020, Claude-Bernard University Lyon 1, 69622 Villeurbanne, France; †Department of Psychology, Vanderbilt University, Nashville, Tennessee 37240; and ‡Neurological Hospital, CERMEP, 69003 Lyon, France Received June 21, 2000; published online January 19, 2001 Humans routinely make judgments about olfactory stimuli. However, few studies have examined the func- tional neuroanatomy underlying the cognitive opera- tions involved in such judgments. In order to delineate this functional anatomy, we asked 12 normal subjects to perform different judgments about olfactory stimuli while regional cerebral blood flow (rCBF) was mea- sured with PET. In separate conditions, subjects made judgments about the presence (odor detection), inten- sity, hedonicity, familiarity, or edibility of different odorants. An auditory task served as a control condi- tion. All five olfactory tasks induced rCBF increases in the right orbitofrontal cortex (OFC), but right OFC activity was highest during familiarity judgments and lowest during the detection task. Left OFC activity increased significantly during hedonic and familiarity judgments, but not during other odor judgments. Left OFC activity was significantly higher during hedonic- ity judgments than during familiarity or other olfac- tory judgments. These data demonstrate that aspects of odor processing in the OFC are lateralized depend- ing on the type of olfactory task. They support a model of parallel processing in the left and right OFC in which the relative level of activation depends on whether the judgment involves odor recognition or emotion. Primary visual areas also demonstrated a differential involvement in olfactory processing de- pending on the type of olfactory task: significant rCBF increases were observed in hedonic and edibility judg- ments, whereas no significant rCBF increases were found in the other three judgments. These data indi- cate that judgments of hedonicity and edibility engage circuits involved in visual processing, but detection, intensity, and familiarity judgments do not. © 2001 Academic Press Key Words: detection; intensity; hedonic; familiarity; edibility; orbitofrontal. INTRODUCTION In a previous positron emission tomography (PET) study, we examined the organization of cognitive oper- ations involved in the perception of odors (Royet et al., 1999). We assessed regional cerebral blood flow (rCBF) while subjects performed judgments of familiarity, ed- ibility, or attempted to detect odorants. We hypothe- sized that the familiarity and the edibility tasks, re- spectively, required either the activation of perceptual or the activation of both perceptual and semantic rep- resentations of odors, whereas the detection task re- quired a superficial judgment that did not involve stored representations of odors. Our results showed that familiarity judgments selectively activated the right medial orbitofrontal cortex (OFC). Edibility judg- ments significantly activated visual regions suggesting that such judgments require (or engage) visual repre- sentations of food. Several other neuroimaging studies of olfaction us- ing fMRI (Levy et al., 1997, 1998; Yang et al., 1997; Yousem et al., 1997; Fulbright et al., 1998; Sobel et al., 1997, 1998a,b, 2000) and PET (Zatorre et al., 1992; Small et al., 1997; Zald and Pardo, 1997, 2000; Dade et al., 1998; Zald et al., 1998a; Savic et al., 2000; Savic and Gulyas, 2000) have been reported. In most of these studies, significant activations localized to the OFC, especially in the right hemisphere. One of the few exceptions to this pattern is a series of studies by Zald and Pardo (1997, 2000) and ourselves (Royet et al., 2000) who observed greater left than right OFC (infe- rior frontal gyrus, pars orbitalis) activity during expo- sure to odorants with strong affective valences. In most of the above-mentioned studies, subjects were not asked to perform a specific task during the actual scanning period. However, as emphasized by De ´monet et al. (1993b), “The presence of a ‘passive task’ in an activation paradigm ignores the nature of the cognitive components of such a task and therefore obscures the interpretation of any between-task differ- ences in brain activity.” They suggested that subjects should perform tasks that require them to actively attend to sensory stimuli. They also proposed the use of active control conditions as opposed to so-called resting tasks in which subjects are not asked to engage in any particular activity. Such an uncontrolled resting state NeuroImage 13, 506 –519 (2001) doi:10.1006/nimg.2000.0704, available online at http://www.idealibrary.com on 506 1053-8119/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.