Abstract

Mental visual synthesis is the capacity for experiencing, constructing, or manipulating ‘mental imagery’. To investigate brain networks involved in mental visual synthesis, brain activity was measured in right-handed healthy volunteers during mental imagery tasks, in which the subjects were instructed to imagine a novel object, that does not exist in the real world, by composing it from two visually presented words associated with a real object or two achromatic line drawings of a real object, using functional magnetic resonance imaging (fMRI). Both tasks activated the same areas in the inferior frontal and inferior temporal cortices of the left hemisphere. Our results indicate that the source of mental visual synthesis may be formed by activity of a brain network consisting of these areas, which are also involved in semantic operations and visual imagery.

Introduction

Mental visual synthesis is the act or power of forming a mental image of something not perceived by the senses or present in reality. It consists of taking parts of our various conceptions and combining them to give new forms and images more selective, more striking, more delightful and more terrible, among other things, than those existing in reality. There is no doubt that our brain generates mental visual synthesis. Therefore, brain networks must be activated and a population of neurons belonging to these brain networks must be activated when a person visualizes synthesized objects in their mind’s eye. Although neuroimaging techniques now have sufficient potential to detect changes in neuronal activity involved in higher cognitive or mental functions, to our knowledge no brain imaging study has been performed that directly investigated the relationship between brain activity and mental visual synthesis. Therefore, in the present study, we used functional magnetic resonance imaging (fMRI) to address the question of which brain areas are involved during mental visual synthesis that is the basis of imagination.

Mental visual imagery is one of the relevant counterparts of mental visual synthesis in the context of human cognitive functions. Mental visual synthesis is the mental capacity for experiencing, constructing, or manipulating mental imagery. Mental visual imagery is defined as the manipulation of visual information arising from memory, giving rise to the experience of seeing through the mind’s eye. In contrast to mental visual synthesis, there are many investigations on mental visual imagery using functional imaging techniques (for reviews, see Mellet et al., 1998; Kosslyn et al., 2001). One of the major controversies in previous functional imaging studies has been whether the primary visual cortex is involved in visual mental imagery. Although many of the previous studies addressed this issue, results of these studies also indicated that not only visual areas but also several brain areas are activated in relation to visual mental imagery and that the brain activation pattern varies according to the object that is mentally visualized. A significant activation of the parieto-occipital areas (often bilaterally) was reported during visual imagery tasks using mental rotation (Mellet et al., 1996), route finding (Roland and Friberg, 1985; Roland et al., 1987; Ghaem et al., 1997) and recall of prelearned three-dimensional scenes (Mellet et al., 2000), patterns (Knauff et al., 2000) and spatial locations (Kawashima et al., 1995). In contrast, temporo-occipital areas were significantly activated when subjects were instructed to imagine construction of concrete objects according to verbal instructions (D’Esposito et al., 1997; Mellet et al., 1998) and based on paired associates (Shallice et al., 1994; Fletcher et al., 1995) and mentally to recall prelearned patterns (Roland and Gulyas, 1995) and letters (Kosslyn et al., 1993). The findings of neuroimaging studies on mental visual imagery are consistent with those of previous neuroimaging studies of visual perception showing functional dichotomy between the dorsal and the ventral pathways according to the spatial and object form perceptions, respectively (for a review, see Ungerleider and Haxby, 1994) , which was first demonstrated in monkeys (Miskin et al., 1983). These findings indicate that mental visual imagery and visual perception share common and very specific mechanisms.

In the present study, we aimed to determine brain areas specifically involved in mental visual synthesis. To do this, we compared brain activity measured by fMRI during mental visual synthesis and mental visual imagery conditions. In this study, we assumed that at least the following cognitive processes have to be involved in mental visual imagery: retrieval of knowledge of two different objects from long-term semantic memory based on external cues (words or line drawings); bringing these pieces of information into working memory; holding it on-line; and seeing those objects through the mind’s eye. We also assumed that mental visual synthesis requires one more specific cognitive process in addition to the above-mentioned processes, that is to create a new object by combining and modifying these pieces of information in novel ways. Since we counterbalanced visual inputs and motor outputs between mental visual synthesis and mental visual imagery conditions in this study, a comparison of mental visual synthesis versus object imagery may indicate brain areas involved in the specific cognitive process for visual synthesis.

Methods

Subjects

Forty-three young right-handed healthy volunteers, of ages ranging from 18 to 31 years, participated in this study. Their handedness was assessed by the Edinburgh Handedness Inventory. Written informed consent was obtained from each subject in accordance with the guidelines approved by the Tohoku Fukushi University and the Helsinki Declaration of Human Rights, 1975. Brain activity was measured in 23 subjects using fMRI during a word task and the remaining 20 subjects performed a picture task during fMRI measurement.

Tasks

The word task consisted of two experimental conditions, that is, object imagery and synthesis conditions and a baseline condition. In the object imagery condition, two different nouns of a common object, e.g. ‘apple’ and ‘table’, were presented at the center of the screen simultaneously for 1 s, followed by a 4 s inter-trial interval. During the inter-trial interval, an eye-fixation point was presented at the center of the screen. The subjects were instructed to imagine the appearance of the real objects in turn during the inter-trial interval and were asked to press a button using the right hand when they succeed to imagine two objects. In the synthesis condition (Fig. 1A), a compound word, which consisted of two nouns but did not indicate a real object, e.g. ‘television–rock’, was presented for 1 s at the center of the screen, followed by a 4 s inter-trial interval. The subjects were instructed to imagine a novel object during the inter-trial interval and were asked to press a button when they succeed in visualizing an object. Actual verbal instruction for synthesis condition was as follows; ‘we ask you to create a vivid image of an object, which does not exist in the real word, in your mind from an artificial word consisting of two nouns of familiar objects’. In the base line condition, the subjects were asked to gaze at an eye fixation point. The visual inputs, that is, the presented words, were counterbalanced between the object imagery and synthesis conditions.

The picture task also consisted of two experimental conditions, that is, object imagery and synthesis conditions and a baseline condition. In the object imagery and synthesis conditions, two different achromatic line drawings of common objects were presented for 1 s at the center of the screen, followed by a 4 s inter-trial interval. During the inter-trial interval, an eye-fixation point was presented at the center of the screen. In the object imagery condition, the subjects were asked to imagine the appearance of real object in turn during the inter-trial interval and were asked to press a button when they succeed in visualizing two objects. In the synthesis condition (Fig. 1B), the subjects were instructed to create a novel object in their minds by compounding the two line drawings, to imagine this object during the inter-trial interval and to press a button when they succeed in visualizing this object. Actual verbal instruction for the synthesis condition was the same as for the word task. The presented line drawings were generated from the same words used in the word task. the baseline condition was the same in the word task. The visual inputs and motor outputs were counterbalanced between the object imagery and synthesis conditions.

Prior to the fMRI scan, subjects were trained with both object imagery and imagination conditions. The reaction time, that is, duration between the start of the inter-trial interval and pressing the button, was measured during fMRI measurements using a personal computer. A debriefing of the subject was held immediately after each fMRI scan.

fMRI Measurements

T1-weighted structural images were acquired for each volunteer, using a 1.5 T Siemens Vision plus scanner (TR = 9.7 ms, echo time = 4 ms, FA = 12°, FOV = 250 mm, pixel size = 1.25 × 0.98 mm, matrix = 175 × 256; Siemens Magnetron Vision, Erlanger, Germany). To measure neural responses, gradient echo, echoplanar T2*-weighted images with blood oxygenation level-dependent (BOLD) contrast were acquired by GE–EPI (TE = 60 ms, FA = 90°, matrix = 64 × 64, FOV = 256 × 256 mm, slice thickness 3 mm, gap = 2.25 mm) covering the entire brain at a TR of 5 s. Forty-two axial slices were oriented over the whole brain. To eliminate T1 equilibration effects, four preliminary scans were acquired and the others subsequently discarded. Thereafter, scans were acquired continuously every 5.0 s.

In the present study, a conventional block design consisting of ten alterations of a 60 s ON block and a 20 s baseline condition block was used for each task. Each of the two experimental conditions was assigned to the ON block five times in the same order for each subject (e.g. ABCABCABCABCABC) and the order of the experimental conditions was counterbalanced across the subjects. Therefore, each ON block consisted of 10 trials for each task.

Image Data Processing and Statistical Analyses

All preprocessing procedures and data analyses were performed using Statistical Parametric Mapping 99 (SPM99) software (Wellcome Department of Cognitive Neurology, London, UK) implemented on MATLAB (Mathworks Inc., Natick, MA). First, slice time adjustment was performed to correct for differences in acquisition time among 34 slices of each scan. Secondly, to correct for artifacts caused by small head movements, images from each subject were realigned to the first image and resliced using a sinc interpolation. Thirdly, a mean image created from the realigned volumes was coregistered with the structural T1 volume and the structural volumes were spatially normalized to a standard template in the space of Talairach and Tournoux (1988) using nonlinear basis functions. Fourthly, the derived spatial transformation was applied to the realigned T2* volumes, which were then spatially smoothed with a three-dimensional isotropic Gaussian Kernel (12 mm full width half-maximum). This improves the signal-to-noise ratio and accommodates for residual variations in functional neuroanatomy that usually persist between subjects after spatial normalization.

Condition effects at each voxel were estimated according to this general linear model and regionally specific effects were compared using linear contrasts. Global changes were adjusted by a proportional scaling and low-frequency confounding effects were removed using an appropriate high-pass filter. Voxel values for each contrast yielded a statistical parametric map of the t-statistic (SPM t) and were subsequently transformed to the unit normal distribution (SPM Z). Then, inter-subject maps were produced by performing one-sample t-tests to identify voxels that survived the statistical threshold of P < 0.05 after correction for multiple comparisons. Finally, the resulting activation maps were constructed and superimposed onto stereotactically standardized T1-weighted MR images.

A set of ROIs was anatomically defined for each subject on the regions that were significantly activated in the synthesis versus object imagery comparisons. For each subject and each ROI, a mean time series, averaged across activated voxels in the region and across all runs, was calculated. We then compared the mean amplitude of fMRI signal for each condition.

Results

Task Performance

Immediately after the fMRI measurements, the subjects were asked to sketch the appearance of objects they imagined during the synthesis conditions. Examples of these sketches are shown in Figure 1.

The mean (SEM) reaction times during object imagery and synthesis conditions of the word task were 1.3 s (0.1) and 1.4 s (0.1), respectively, with no statistically significant differences (paired t-test). Those during object imagery and synthesis conditions of the picture task were 2.5 s (0.2) and 2.7 s (0.2), respectively, with no statistically significant differences (paired t-test). The reaction time in the picture task was statistically significantly longer than that in the word task (ANOVA, P < 0.001).

The mean (SEM) percentages of successful trials of the word task were 98.0% (2.8) and 95.0% (2.7) during object imagery and synthesis conditions with no statistically significant differences (paired t-test). Those during object imagery and synthesis conditions of the picture task were 93.4% (4.0) and 90.0% (5.0), respectively, with no statistically significant differences (paired t-test). The mean percentage of successful trials was not statistically significant between the word and picture tasks (ANOVA).

Brain Activation

Table 1 and Figure 2 summarize brain areas activated during the word task. The object imagery versus baseline comparison revealed the significant activation of the inferior frontal and inferior temporal cortices of the left hemisphere, anterior cingulate cortex, as well as bilateral cerebellum. The synthesis versus baseline comparison showed the significant activation of the inferior frontal, inferior temporal and intraparietal cortices of the left hemisphere, anterior cingulate cortex, bilateral parahippocampal cortices, right insula and the right cerebellum. The synthesis condition more significantly activated the left inferior prefrontal cortex, left inferior temporal cortex, bilateral striatum and the right cerebellum than the object imagery condition.

Table 2 and Figure 2 summarize brain areas activated during the picture task. The object imagery versus baseline comparison showed significant activation of the left inferior frontal cortex, anterior cingulate cortex and bilateral cerebellum. Although subjects looked at the line drawing pictures during the object imagery condition, activation of the primary visual cortex did not reach a statistically significant level. The synthesis versus baseline comparison revealed the significant activation of the bilateral inferior frontal, left inferior temporal, left intraparietal, left fusiform and anterior cingulate cortices, as well as the right striatum and right cerebellum. The synthesis condition more significantly activated the left inferior frontal, bilateral inferior temporal, bilateral intraparietal and left lateral occipital cortices than the object imagery condition. A larger extent of activation was evoked by the picture task compared with the word task in each comparison.

It should be noted that the same two areas in the left inferior frontal cortex, one in the anterior region, the other in the posterior region and an area in the posterior part of the left inferior temporal cortex were significantly activated during the synthesis condition compared with the object imagery condition during both tasks (Fig. 2). The anterior and posterior regions of the inferior frontal gyrus are probably located in the orbital and opercular parts of the inferior frontal gyrus, respectively. The magnitude of activation in these areas during each condition was not statistically significant (Student’s t-test) between the picture and word tasks (Fig. 3).

Discussion

Our results indicate that the neuronal network consisting of the inferior frontal and the inferior temporal cortices of the left hemisphere plays an important role in creating an entirely new image in one’s mind.

In this study, we assumed that in order to perform the synthesis condition of each task, at least following the three cognitive processes are involved: (i) retrieval of information, that is, knowledge of two different objects, from long term semantic memory provided by external cues (words or line drawings) and bringing this information into working memory and holding it on-line; (ii) creation of a new object by combining and modifying this information in novel ways; and (iii) visualize that object through the mind’s eye. We also assumed that processes (i) and (iii) are necessary to perform the object imagery condition as well. Therefore, visual inputs and motor outputs were counterbalanced between the object imagery and synthesis conditions during each task; a comparison of synthesis versus object imagery may indicate brain areas involved in the creation of entirely new objects by combining two different representations of the objects.

Hemispheric Asymmetry

We found that both object imagery and synthesis evoked the left-lateralized activation of the fronto-temporal network. The similar hemispheric asymmetry was suggested by brain imaging studies of visual imagery of familiar objects (D’Esposito et al., 1997; Ishai et al., 2000, 2002). On the other hand, the activation of the right inferior temporal cortex has also been reported during the mental imagery of complex objects or schematic objects (Mellet et al., 1996) and letters (Kosslyn et al., 1993). Mellet et al. (1998) reviewed previous brain imaging studies and concluded that functional lateralization during mental imagery can depend on two different characteristics of mental images to be generated: complexity, which would modulate the degree of involvement of the right hemisphere; and lexicality, which would drive the left hemisphere participation.

Farah (1986) found that left hemisphere was better at using imagery to prime perceptual recognition by a psychophysical study of normal subjects, although Kosslyn et al. (1985) found that in split-brain patients this left-hemisphere advantage was selective, occurring only for multipart images and that both hemispheres were equally adept at generating single-unit images of an overall shape. Kosslyn et al. (1995) suggested that left hemisphere more effectively generates images by arranging parts according to descriptions (using categorical spatial relations) whereas the right hemisphere more effectively generates images by positioning parts in precise locations in space (using coordinate spatial relations) from the results of their psychophysical experiments. Although there is no doubt that both hemispheres can generate visual mental images, the specific functional role for each hemisphere in visual mental imagery still remains unclear. The variability in these findings might reflect differences in the specific combinations of processes engendered by the instructions, nature of the stimuli and subject’s previous experience (Kosslyn et al., 1995).

The results of the present study indicate that mental visual synthesis shares common neural substrates with object imagery, that these substrates, that is the inferior frontal and inferior temporal cortices of the left hemisphere, play roles in creating mental images from long-term memory and manipulating these in the mind’s eye, that the synthesis condition differs from the object imagery condition in a quantitative rather than a qualitative way and that the left-lateralized activation may be resulted from intense lexical and categorical processing required for mental visual synthesis.

Inferior Frontal Cortex

We defined two distinct areas in the inferior frontal cortex of the left hemisphere involved in mental visual synthesis. These two areas were also activated during the object imagery condition, although they were more significantly activated during the synthesis condition.

The previous neuroimaging investigations suggested that there is no doubt the inferior frontal cortex is involved in semantic processing. One of the functional roles of the inferior frontal cortex derived from the results of previous neuroimaging studies is semantic generation. Previous positron emission tomography (PET) studies of verb generation tasks, that is, to generate semantically appropriate verbs from the presented nouns, reported activation of the left inferior frontal cortex (Petersen et al., 1988; Raichle et al., 1994; Tatsumi et al., 1999). This area is also activated when subjects name colors (Wiggs et al., 1999) or generate action words (Martin et al., 1995) from visually presented line drawings of objects. Vandenberghe et al. (1996) also suggested that the left inferior frontal cortex is commonly activated when subjects make semantically based decisions on both presented words and pictures. Another functional role of the inferior frontal cortex determined from brain imaging studies is retrieval of semantic knowledge. Previous PET studies using picture naming paradigms indicate involvement of the left inferior frontal cortex in category-specific retrieval (Damasio et al., 1996; Martin et al., 1996; Grabowski et al., 1997). fMRI studies of semantic category generation and semantic memory tasks also activated this area (Shaywitz et al., 1995; Gabrieli et al., 1998; Crosson et al., 1999). In addition, some studies indicated that the inferior frontal cortex is involved in not only for retrieval but also semantic encoding (Demb et al., 1995; Owen et al., 1996; Nyberg et al., 1996). Demb et al. (1995) suggested that the inferior frontal cortex is part of a semantic executive system that contributes to the on-line retireval of semantic information, similar to executive systems in a model or working memory (Baddeley, 1986; Goldman-Rakic, 1987). However, the inferior frontal cortex was not activated during the Tower of London task, which required motor planning and working memory (Baker et al., 1996), or during non-verbal encoding tasks (Golby et al., 2001). The results of previous studies suggest that the inferior frontal cortex is involved in semantic encoding and retrieval. The other function is selection of semantic knowledge. The prefrontal cortex has been argued to produce flexible and context-sensitive responses (Cohen and Servan-Schreiber, 1992) and to mediate the selection of action by the weighing of information active in working memory (Kimberg and Farah, 1993). From the results of an fMRI study, Thompson-Schill et al. (1997) found that demands on the inferior frontal cortex are high in any task that requires selection among competing sources of information in working memory to guide a response. Since semantic generation, retrieval of semantic knowledge from presented words or pictures and selection of semantic knowledge are important cognitive process for mental visual synthesis, we assumed that the results of our study are in agreement with those of previous imaging studies and that several different semantic modules in the inferior frontal cortex are responsible to mental visual synthesis.

Although, in this study, we were unable to determine functional dissociations of the anterior and posterior regions of the left inferior frontal cortex, some previous neuroimaging studies indicated interesting functional differences between the two regions. Poldrak et al. (1999) and Heim et al. (2003) suggested from the results of their fMRI studies that the anterior region subserves semantic processing and the posterior region subserves phonological processing. Bokde et al. (2001) obtained consistent results from functional connectivity analysis using fMRI. Fletcher and Henson (2001) summarized from the literature review that the anterior region is involved in the semantic contribution to successful encoding and the posterior region is involved in selection from various semantic attributes. At the moment, we cannot explain for the differences in functional organization of these two regions, although we can argue that the two regions may have different functional roles and that the both areas are involved in mental visual synthesis.

In summary, since inferior frontal cortex was activated during the object imagery as well as synthesis conditions and was more significantly activated during the synthesis condition, the difference in the inferior frontal activation between these two conditions may well be resulted from the difference in the semantic demands of the tasks.

Inferior Temporal Cortex

In this study, the posterior part of the inferior temporal cortex was more activated during the synthesis condition than during the imagery condition in both tasks. Activation of the same area was reported in the previous brain imaging studies during retrieval of object features (Haxby et al., 1994; Kohler et al., 1995; Moscovitch et al., 1995; Martin et al., 1996; Owen et al., 1996) and encoding of object features (Owen et al., 1996). This area has been considered as a target of the ventral stream (Miskin et al., 1983) originating from the visual cortex, which is specialized for processing object forms (for reviews, see Goodale, 1993; Ungerleider and Haxby, 1994). Previous lesion studies of humans revealed that the posterior temporo-occipital damage impairs mental-image generation (Sergent, 1989; Tippet, 1992; Trojano and Grossi, 1994; Farah, 1995). Therefore, it is argued that the inferior temporal cortex is involved in object mental imagery and perception (Mellet et al., 1998). The involvement of the left inferior temporal cortex in semantic processing was also reported in previous brain imaging studies (Tranel et al., 1997; Martin et al., 2000). Since one can easily build a mental image based on a verbal description (Denis and Cocude, 1992), or provide an oral description of a mental image, we may argue that strong interactions exist between language and mental imagery that permit an easy transfer from one representation mode to the other and that the interactions may take place in the inferior temporal cortex. It should be noted that a recent fMRI study by Bokde et al. (2001) indicated strong functional connectivity between anterior or posterior regions of the left inferior frontal cortex and the posterior part of the inferior temporal cortex during phonological processing of words. Their findings support our results, since we noted that the inferior temporal cortex was co-activated with the left inferior frontal cortex during the synthesis condition.

This study has been supported by JST/RISTEX, R&D promotion scheme for regional proposals promoted by TAO, a Grant-Aid for Scientific Research on Priority areas (C) — Advanced Brain Science Project — from MEXT and the 21st Century Center of Excellence (COE) Program (Ministry of Education, Culture, Sports, Science and Technology) entitled ‘A Strategic Research and Education Center for an Integrated Approach to Language and Cognition’ (Tohoku University).

Figure 1. Diagrams of synthesis conditions. (A) Word task. (B) Picture task.

Figure 1. Diagrams of synthesis conditions. (A) Word task. (B) Picture task.

Figure 2. Surface projections of color-coded statistical parametric maps (SPMs) showing statistically significant activations. The left sides of the figures correspond to the left hemisphere. The SPMs are group maps from 23 and 20 subjects for word task and picture task, respectively. Thresholds were set at P < 0.05 after correction for multiple comparisons.

Figure 2. Surface projections of color-coded statistical parametric maps (SPMs) showing statistically significant activations. The left sides of the figures correspond to the left hemisphere. The SPMs are group maps from 23 and 20 subjects for word task and picture task, respectively. Thresholds were set at P < 0.05 after correction for multiple comparisons.

Figure 3. Mean amplitude of fMRI signal in significantly activated areas in synthesis versus object imagery comparisons. Black and white bars indicate mean amplitude during the synthesis and object imagery conditions, respectively. Error bars indicate standard error of the mean (SEM). MFC, middle frontal cortex; IFC, inferior frontal cortex; IPC, intraparietal cortex; ITC, inferior temporal cortex; Str, striatum; CBp, posterior lobule of the cerebellum; L, left hemisphere; R, right hemisphere.

Figure 3. Mean amplitude of fMRI signal in significantly activated areas in synthesis versus object imagery comparisons. Black and white bars indicate mean amplitude during the synthesis and object imagery conditions, respectively. Error bars indicate standard error of the mean (SEM). MFC, middle frontal cortex; IFC, inferior frontal cortex; IPC, intraparietal cortex; ITC, inferior temporal cortex; Str, striatum; CBp, posterior lobule of the cerebellum; L, left hemisphere; R, right hemisphere.

Table 1


 Talairach coordinates and T-value of peak activation during the word task

Anatomical name Talairach coordinates  T Volume 
 x y z    
Object imagery versus baseline       
Lt inferior frontal –42   6  30   8.2    1192 
Lt inferior frontal –42  10   2   8.1     856 
Lt inferior frontal –46  24  –4   7.3    1008 
Lt inferior temporal –50 –64 –32   8.6    3848 
Lt cerebellum ant lobule –16 –88 –12   5.3     192 
Rt anterior cingulate   8  14  46   6.6     392 
Rt cerebellum anterior lobule  40 –42 –30   8.8     848 
Rt cerebellum posterior lobule  44 –64 –26   6.4     424 
Rt cerebellum posterior lobule  10 –82 –20   9.2    2904 
Synthesis versus baseline       
Lt inferior frontal –40   6  32  10.9    4192 
Lt inferior frontal –48  30  10  10.7    4320 
Lt inferior temporal –52 –60 –10  10.8 18  376 
Lt intraparietal –26 –60  40   9.1    2000 
Lt parahipocampal –16 –24 –12   7.0    2112 
Rt parahipocampal  22 –26 –16   5.9     200 
Rt striatum  16  –4  12   7.2    2336 
Anterior cingulate   8  14  48   7.2     592 
Rt cerebellum anterior lobule  30 –34 –34  11.9    5016 
Synthesis versus object imagery       
Lt inferior frontal –48   6  22   8.2    1856 
Lt inferior frontal –48  30  10  10.6    4032 
Lt inferior temporal –54 –60  –8   6.9    1032 
Rt striatum  16   0  14   7.6     744 
Lt striatum –12   0  12   6.0     176 
Rt cerebellum posterior lobule  16 –82 –26   7.8    1680 
Anatomical name Talairach coordinates  T Volume 
 x y z    
Object imagery versus baseline       
Lt inferior frontal –42   6  30   8.2    1192 
Lt inferior frontal –42  10   2   8.1     856 
Lt inferior frontal –46  24  –4   7.3    1008 
Lt inferior temporal –50 –64 –32   8.6    3848 
Lt cerebellum ant lobule –16 –88 –12   5.3     192 
Rt anterior cingulate   8  14  46   6.6     392 
Rt cerebellum anterior lobule  40 –42 –30   8.8     848 
Rt cerebellum posterior lobule  44 –64 –26   6.4     424 
Rt cerebellum posterior lobule  10 –82 –20   9.2    2904 
Synthesis versus baseline       
Lt inferior frontal –40   6  32  10.9    4192 
Lt inferior frontal –48  30  10  10.7    4320 
Lt inferior temporal –52 –60 –10  10.8 18  376 
Lt intraparietal –26 –60  40   9.1    2000 
Lt parahipocampal –16 –24 –12   7.0    2112 
Rt parahipocampal  22 –26 –16   5.9     200 
Rt striatum  16  –4  12   7.2    2336 
Anterior cingulate   8  14  48   7.2     592 
Rt cerebellum anterior lobule  30 –34 –34  11.9    5016 
Synthesis versus object imagery       
Lt inferior frontal –48   6  22   8.2    1856 
Lt inferior frontal –48  30  10  10.6    4032 
Lt inferior temporal –54 –60  –8   6.9    1032 
Rt striatum  16   0  14   7.6     744 
Lt striatum –12   0  12   6.0     176 
Rt cerebellum posterior lobule  16 –82 –26   7.8    1680 

Stereotaxic coordinates (in millimeters) identify the location of the maxima of haemodynamic responses corresponding to the atlas of Talairach and Tournoux (1988). Lt and Rt indicate left and right hemisphere, respectively. Volumes are in mm3.

Table 2


 Talairach coordinates and T-value of peak activation during the picture task

Anatomical name Talairach coordinates  T Volume 
 x y z    
Object imagery versus baseline       
Lt inferior frontal –38  24  24  10.2    4808 
Lt inferior frontal –32  52  16   8.9     392 
Lt inferior frontal –32  26  –4   8.8    2176 
Anterior cingulate –10   6  52   9.3    5656 
Lt cerebellum anterior lobule –40 –48 –24   7.9    1464 
Rt cerebellum anterior lobule  38 –44 –32  15.4    3936 
Synthesis vers Lt middle frontal us baseline –22   0  60  11.2 11  944* 
Lt inferior frontal –48  16  24  15.0 29  552** 
Lt inferior frontal –50  30  28  15.9 29  552** 
Lt inferior frontal –46  46  –4  11.4 29  552** 
Rt inferior frontal  48  36  12   7.1     248 
Lt intraparietal –32 –50  44  11.1 13  752 
Lt inferior temporal –50 –56 –12  10.4    3776 
Rt striatum  16  –2  12   7.1     168 
Anterior cingulate   6  20  44  14.2 11  944* 
Rt cerebellum anterior lobule  36 –48 –36  10.9    7608 
Synthesis versus object imagery       
Lt middle frontal –26   6  66  10.1    1352 
Lt inferior frontal –46  10  26  11.4    7448 
Lt inferior frontal –48  48  –4   9.7    1392 
Lt intraparietal –36 –48  46  11.0 10  552 
Rt intraparietal  26 –66  52   7.7     320 
Lt inferior temporal –52 –62 –10  13.7    2280 
RT inferior temporal  56 –60 –12   7.9     200 
Lt lateral occipital –30 –74  32   9.4    1568 
Anatomical name Talairach coordinates  T Volume 
 x y z    
Object imagery versus baseline       
Lt inferior frontal –38  24  24  10.2    4808 
Lt inferior frontal –32  52  16   8.9     392 
Lt inferior frontal –32  26  –4   8.8    2176 
Anterior cingulate –10   6  52   9.3    5656 
Lt cerebellum anterior lobule –40 –48 –24   7.9    1464 
Rt cerebellum anterior lobule  38 –44 –32  15.4    3936 
Synthesis vers Lt middle frontal us baseline –22   0  60  11.2 11  944* 
Lt inferior frontal –48  16  24  15.0 29  552** 
Lt inferior frontal –50  30  28  15.9 29  552** 
Lt inferior frontal –46  46  –4  11.4 29  552** 
Rt inferior frontal  48  36  12   7.1     248 
Lt intraparietal –32 –50  44  11.1 13  752 
Lt inferior temporal –50 –56 –12  10.4    3776 
Rt striatum  16  –2  12   7.1     168 
Anterior cingulate   6  20  44  14.2 11  944* 
Rt cerebellum anterior lobule  36 –48 –36  10.9    7608 
Synthesis versus object imagery       
Lt middle frontal –26   6  66  10.1    1352 
Lt inferior frontal –46  10  26  11.4    7448 
Lt inferior frontal –48  48  –4   9.7    1392 
Lt intraparietal –36 –48  46  11.0 10  552 
Rt intraparietal  26 –66  52   7.7     320 
Lt inferior temporal –52 –62 –10  13.7    2280 
RT inferior temporal  56 –60 –12   7.9     200 
Lt lateral occipital –30 –74  32   9.4    1568 

Details are as in Table 1; * and ** indicate that these activation peaks were observed in one activation cluster.

References

Baddeley A (
1986
) Working memory. Oxford: Clarendon Press.
Baker SC, Rogers RD, Owen AM, Frith CD, Dolan RJ, Frackwiak RSJ, Robbins TW (
1996
) Neural systems engaged by planning: a PET study of the Tower of London task.
Neuropsychologia
 
34
:
515
–526.
Bokde ALW, Tagamets MA, Friedman RB, Horwitz B (
2001
) Functional interactions of the inferior frontal cortex during the processing of words and word-like stumuli.
Neuron
 
30
:
609
–617.
Cohen JD, Servan-Schreiber D (
1992
) Context, cortex and dopamine: a connectionist approach to behavior and biology in schizophrenia.
Psychol Rev
 
99
:
45
–77.
Crosson B, Rao SM, Woodley SJ, Rosen AC, Bobholz JA, Mayer A, Cunningham JM, Hammeke TA, Fuller SA, Binder JR, Cox RW, Stein EA (
1999
) Mapping of semantic, phonological and orthographic verbal working memory in normal adults with sunctional magnetic resonance imaging.
Neuropsychology
 
13
:
171
–187.
Damasio H, Grabowski TJ, Tranel D, Hichwa RD, Damasio AR (
1996
) A neural basis for lexical retrieval.
Nature
 
380
:
499
–505.
Demb JB, Desmond JE, Wagner AD, Vaidya CJ, Glover GH, Gabrieli DE (
1995
) Semantic encoding and retrieval in the left inferior prefrontal cortex: a functional MRI study of task difficulty and process specificity.
J Neurosci
 
15
:
5870
–5878.
Denis M, Cocude M (
1992
) Structual properties of visual images constructed from poorly, or well-structured, verbal descriptions.
Mem Cognit
 
20
:
497
–505.
D’Esposito M, Detre JA, Aguirre GK, Stallcup M, Alsop DC, Tippet LJ, Farah MJ (
1997
) A functional MRI study of mental image generation.
Neuropsychologia
 
35
:
725
–730.
Farah MJ (
1986
) The laterality of mental imagery generation: a test with normal subjects.
Neuropsychologia
 
24
:
541
–551.
Farah MJ (
1995
) Current issues in neuropsychology of image generation.
Neuropsychologia
 
33
:
1455
–1471.
Fletcher PC, Henson RNA (
2001
) Frontal lobes and human memory. Insights from functional neuroimaging.
Brain
 
124
:
849
–881.
Fletcher PC, Frith CD, Baker SC, Shallice T, Frackowiak RSK, Dolan RJ (
1995
) The mind’s eye — precuneus activation in memory-related imagery.
Neuroimage
 
2
:
195
–200.
Gabrieli JDE, Poldrack RA, Desmond JE (
1998
) The role of left prefrontal cortex in language and memory.
Proc Natl Acad Sci USA
 
95
:
906
–913.
Ghaem O, Mellet E, Crivello F, Tzourio N, Mazoyer B, Berthoz A, Denis M (
1997
) Mental navigation along memorized routes activates the hippocampus, precuneus and insula.
Neuroreport
 
8
:
739
–744.
Golby AJ, Poldrack RA, Brewer JB, Spencer D, Desmond JE, Aron AP, Gabrieli JDE (
2001
) Material-specific lateralization in the medial temporal lobe and prefrontal cortex during memory encoding.
Brain
 
124
:
1841
–1854.
Goldman-Rakic PS (
1987
) Circuitry of primate prefrontal cortex and regulation of behavior by representational memory. In Handbook of physiology (Mountcastle VB, Plum F, eds), pp.
373
–417. Bethesda, MD: American Physiological Society.
Goodale MA (
1993
) Visual pathways supporting perception and action in the primate cerebral cortex.
Curr Opin Neurobiol
 
3
:
578
–585.
Grabowski TJ, Damasio H, Damasio AR (
1997
) Premotor and prefrontal correlates of category-related lexical retrieval.
Neuroimage
 
7
:
232
–243.
Haxby JV, Horwitz B, Ungerleider LG, Maisog JM, Pietrini P, Grady CL (
1994
) The functional organization of human extrastriate cortex: a PET–rCBF study of selective attention to faces and locations
J Neurosci
 
14
:
6336
–6353.
Heim S, Opitz B, Friederici AD (
2003
) Distributed cortical networks for syntax processing: Broca’s area as the common denominator.
Brain Lang
 
85
:
402
–408.
Ishai A, Ungerleider LG, Haxby JV (
2000
) Distributed neural systems for the generation of visual images.
Neuron
 
28
:
979
–990.
Ishai A, Haxby JV, Ungerleider LG (
2002
) Visual imagery of famous faces: effects of memory and attention related by fMRI.
Neuroimage
 
17
:
1729
–1741.
Kawashima R, Roland PE, O’Sullivan BT (
1995
) Functional anatomy of reaching and visuo-motor learning: a positron emission tomography study.
Cereb Cortex
 
2
:
111
–122.
Kimberg DY, Farah MJ (
1993
) A unified account of cognitive impairments following frontal lobe damage: the role of working memory in complex, organized behavior.
J Exp Psychol Gen
 
122
:
411
–428.
Knauff M, Kassubek J, Mulack T, Greenlee MW (
2000
) Cortical activation evoked by visual mental imagery as measured by fMRI.
Neuroreport
 
11
:
3957
–3962.
Kohler S, Kapur S, Moscovitch M, Winocur G, Houle S (
1995
) Dissociation of pathways for object and spatial vision: a PET study in humans.
Neuroreport
 
6
:
1865
–1868.
Kosslyn SM, Holtzman JD, Farah MJ, Gazzaniga MS (
1985
) A computational analysis of mental image generation: evidence from functional dissociations in split-brain patients.
J Exp Psychol Gen
 
114
:
311
–341.
Kosslyn SM, Alpert NM, Thompson WL, Maljkovic V, Weise SB, Chabris CF, Hamilton SE, Rauch SL, Buonanno FS (
1993
) Visual mental imagery activates topographically organized visual cortex: PET investigations.
J Cogn Neurosci
 
5
:
263
–287.
Kosslyn SM, Maljkovic V, Hamilton SE, Horwitz G, Thompson WL (
1995
) Two types of image generation: evidence for left and right hemisphere processes.
Neuropsychologia
 
33
:
1485
–1510.
Kosslyn SM, Ganis G, Thompson WL (
2001
) Neural foundations of imagery.
Nat Rev Neurosci
 
2
:
635
–642.
Martin A, Haxby JV, Lalonde M, Wiggs CL, Ungerleider LG (
1995
) Discrete cortical regions associated with knowledge of color and knowledge of action.
Science
 
270
:
102
–105.
Martin A, Wiggs CL, Ungerleider LG, Haxby JV (
1996
) Neural correlates of category-specific knowledge.
Nature
 
379
:
649
–652.
Martin A, Ungerleider LG, Haxby JV (
2000
) Category specificity and the brain: the sensory/motor model of semantic representations. In: The new cognitive neurosciences (Gazzaniga MS, ed.), pp.
1023
–1036. Cambridge, MA: MIT Press.
Mellet E, Tzourio N, Crivello F, Joliot M, Denis M, Mazoyer B (
1996
) Functional anatomy of spatial mental imagery gerenated from verbal instructions.
J Neurosci
 
16
:
6504
–6512.
Mellet E, Petit L, Mazoyer B, Denis M, Tzourio N (
1998
) Reopening the mental imagery debate: lessons from functional anatomy.
Neuroimage
 
8
:
129
–139.
Mellet E, Kosslyn SM, Denis M (
2000
) Functional anatomy of high-resolution visual mental imagery.
J Cogn Neurosci
 
12
:
98
–109.
Miskin M, Ungerleider LG, Macko KA (
1983
) Object vision and spatial vision: two cortical pathways.
Trends Neurosci
 
6
:
414
–417.
Moscovitch M, Kapur S, Kohler S, Houle S (
1995
) Distinct neural correlates of visual long-term memory for spatial location and object identity: a positron emission tomography study in humans.
Proc Natl Acad Sci USA
 
92
:
3721
–3725.
Nyberg L, McIntosh AR, Cabeza R, Habib R, Houle S (
1996
) General and specific brain regions involved in encoding and retrieval of events: what, where and when.
Proc Natl Acad Sci USA
 
93
:
11280
–11285.
Owen AM, Milner B, Petrides M, Evans AC (
1996
) Memory for object features versus memory for object location: a positron-emission tomography study of encoding and retrieval processes.
Proc Natl Acad Sci USA
 
93
:
9212
–9217.
Petersen SE, Fox PT, Posner MI, Mintun M, Raichle ME (
1988
) Positron emission tomographic studies of the processing of single words.
J Cogn Neurosci
 
1
:
153
–170.
Poldrak RA, Wagner AD, Prull MW, Desmond JE, Glover GH, Gabrieli JD (
1999
) Functional specialization for semantic and phonological processing in the left inferior prefrontal cortex.
Neuroimage
 
10
:
15
–35.
Raichle ME, Fiez JA, Videen TO, MacLeod AM, Pardo JV, Fox PT (
1994
) Practice-related changes in human brain functional anatomy during nonmotor learning.
Cereb Cortex
 
4
:
8
–26.
Roland PE, Friberg L (
1985
) Localization of cortical areas activated by thinking.
J Neurophysiol
 
53
:
1219
–1243.
Roland PE, Gulyas B (
1995
) Visual memory, visual imagery and visual recognition of large field patterns by the human brain: functional anatomy by positron emission tomography.
Cereb Cortex
 
1
:
79
–93.
Roland PE, Erikson L, Stone-Elander S, Widen L (
1987
) Does mental activity change the oxidative metabolism of the brain?
J Neurosci
 
7
:
2373
–2389.
Sergent J (
1989
) Image generation and processing of generated images in the cerebral hemispheres.
J Exp Psychol Hum Percept Perform
 
15
:
170
–178.
Shallice T, Fletcher P, Frith CD, Grasby P, Frackowiak RS, Dolan RJ (
1994
) Brain regions associated with acquisition and retrieval of verbal episodic memory.
Nature
 
368
:
633
–635.
Shaywitz BA, Shaywitz SE, Pugh KR, Constable RT, Skudlarski P, Fulbright RK, Bronen RA, Fletcher JM, Shankweller DP, Katz L, Gore JC (
1995
) Sex difference in the functional organization of the brain for language.
Nature
 
373
:
607
–609.
Talairach J, Tournoux P (
1988
) Co-planar sterotaxic atlas of the human brain. New York: Thieme.
Tatsumi IF, Fushimi T, Sadato N, Kawashima R, Yokoyama E, Kanno I, Senda M (
1999
) PET activation studies for verb generation in Japanese — a multi-center study.
Neuroimage
 
9
:
154
–164.
Thompson-Schill S, D’Esposito M, Aguirre GK, Farah MJ (
1997
) Role of left inferior prefrontal cortex in retrieval of semantic knowledge: a reevaluation.
Proc Natl Acad Sci USA
 
94
:
14792
–14797.
Tippet LJ (
1992
) The generation of mental images: a review of neuropsychological research and theory.
Psychol Bull
 
112
:
415
–432.
Tranel D, Damasio H, Damasio AR (
1997
) A neural basis for the retrieval of conceptual knowledge.
Neuropsychologia
 
35
:
1319
–1328.
Trojano L, Grossi D (
1994
) A critical review of mental imagery defects.
Brain Cogn
 
24
:
213
–243.
Ungerleider LG, Haxby JV (
1994
) ‘What’ and ‘where’ in the human brain.
Curr Opin Neurobiol
 
4
:
157
–165.
Vandenberghe R, Price C, Wise R, Josephs O, Frackowiak RSJ (
1996
) Functional anatomy of a common semantic system for words and pictures.
Nature
 
383
:
254
–256.
Wiggs CL, Weisberg J, Martin A (
1999
) Neural correlates of semantic and episodic memory retrieval.
Neuropsychologia
 
37
:
103
–118.