of color associations with concepts of daily life or objects; and 3) an ability to name colors. Brain color processing be- gins with the analysis of specific light ...
15658_pt.5
12/21/05
5:43 PM
Page 27
J Neurosurg 104:27–37, 2006
Category-specific cortical mapping: color-naming areas FRANCK-EMMANUEL ROUX, M.D., PH.D., VINCENT LUBRANO, M.D., VALÉRIE LAUWERS-CANCES, M.D., CHRISTOPHER R. MASCOTT, M.D., AND JEAN-FRANÇOIS DÉMONET, M.D., PH.D. Institut National de la Santé et de la Recherche Médicale, Unité 455; Fédération de Neurochirurgie; Service d’Epidémiologie; and Fédération de Neurologie, Centres Hospitalo-Universitaires, Toulouse, France Object. It has been hypothesized that a certain degree of specialization exists within language areas, depending on some specific lexical repertories or categories. To spare hypothetical category-specific cortical areas and to gain a better understanding of their organization, the authors studied patients who had undergone electrical stimulation mapping for brain tumors and they compared an object-naming task with a category-specific task (color naming). Methods. Thirty-six patients with no significant preoperative language deficit were prospectively studied during a 2-year period. Along with a reading task, both object- and color-naming tasks were used in brain mapping. During color naming, patients were asked to identify 11 visually presented basic colors. The modality specificity of the colornaming sites found was subsequently tested by asking patients to retrieve the color attributes of objects. High individual variability was observed in language organization among patients and in the tasks performed. Significant interferences in color naming were found in traditional language regions—that is, Broca (p , 0.003) and Wernicke centers (p = 0.05)—although some color-naming areas were occasionally situated outside of these regions. Color-naming interferences were exclusively localized in small cortical areas (, 1 cm2). Anatomical segregation of the different naming categories was apparent in 10 patients; in all, 13 color-specific naming areas (that is, sites evoking no object-naming interference) were detected in the dominant-hemisphere F3 and the supramarginal, angular, and posterior parts of the temporal gyri. Nevertheless, no specific brain region was found to be consistently involved in color naming (p . 0.05). At five sites, although visually presented color-naming tasks were impaired by stimulation, auditory color naming (for example, “What color is grass?”) was performed with no difficulty, showing that modality-specific areas can be found during naming. Conclusions. Within language areas, a relative specialization of cortical language areas for color naming can be found during electrical stimulation mapping.
KEY WORDS • brain mapping • cortical stimulation • color naming UR knowledge of colors is a puzzling human ability. Indeed, our ability to discriminate colors, to associate them with given concepts or objects, or to name them are important functions used in all activities from everyday life situations to the most refined artistic achievements. Since the seminal descriptions at the end of the nineteenth century by Verrey52 and Wilbrand56 of brain-damaged patients with impaired color processing, our knowledge of brain color processing has much improved.12,34,58 It has been hypothesized that brain color processing can be schematically divided into the following three hierarchically organized components, which may be specific and anatomically distinct: 1) an ability to discriminate colors; 2) a knowledge of color associations with concepts of daily life or objects; and 3) an ability to name colors. Brain color processing begins with the analysis of specific light wavelengths (that is, color discrimination) in the visual area V4 located in the dominant fusiform gyrus2,10, 12,58 and, more specifically, in the lateral part of the collateral sulcus.34 It has been further hypothesized that distinct areas could be involved in the
O
Abbreviations used in this paper: F1 = superior frontal gyrus; F2 = middle frontal gyrus; F3 = inferior frontal gyrus; T1 = superior temporal gyrus; T2 = middle temporal gyrus; T3 = inferior temporal gyrus.
J. Neurosurg. / Volume 104 / January, 2006
neurocognitive process of color knowledge in the basal temporal region.8,36,50 This region could be involved in the retrieval of color attributes of concepts or objects (for example, knowing that a banana is yellow or a letterbox is red). Patients with defects in color knowledge (color agnosia) can match colors but are unable to choose the color of a familiar object or are unable to answer questions like, “Could you name a few things that are red?” Areas involved in color knowledge could be located in the lateral inferior temporal lobe according to some authors32,51,58 or in the medial temporal structures according to others.24,36 Among the brain functions involved in color processing, color naming is the ability to attribute a color name to real objects or stimuli involving perceivable chromatic contrasts4,5 and can be viewed as the final verbal step in brain color processing. Data from previous studies on color naming in healthy volunteers9,30,46 demonstrated that the cortical areas involved in color naming were widely distributed throughout cortical territories9 and were related to language functions.30 In aphasic patients, in whom selective or preferential impairments of some semantic categories of object names have been reported,55 selective color-naming defects were often identified.11,15,19,20,27,42 Thus, it could be hypothesized that within language areas a certain degree of specialization occurs, depending on specific lexical repertories or 27
15658_pt.5
12/21/05
5:43 PM
Page 28
F. E. Roux, et al. categories such as color names. Moreover, some anatomical segregation of different naming categories has been observed in some neuroimaging studies.14,32,43 We routinely use direct brain mapping in patients with brain tumors to spare language-related cortical areas. To spare hypothetical category-specific cortical areas at surgery and to gain a better understanding of these category- and/or modality-specific naming areas within language areas, we studied patients who had undergone surgery for brain tumors. We used a standard object task and a category-specific task (that is, color naming from visually presented stimuli or from semantic information conveyed by auditorily presented sentences) during direct cortical mapping. We chose the color-naming category because it was the most simple; French words signifying a color usually consist of one or two syllables, are easy to pronounce, are learned early, and are individualized as a set from childhood;22 and specific impairment or preservation of this naming category has long been well described in the medical literature.26,42,47 The cortical organization of color-naming areas was analyzed in relation to the surgical data. Clinical Material and Methods Patient Population
Among the patients who had undergone surgery for brain tumors or other lesions (that is, cavernous angiomas) at our institution, 36 patients (18 men and 18 women) from 22 to 73 years old (mean 45 years of age) and demonstrating no language deficit were prospectively studied using cortical brain mapping between May 2002 and May 2004. All were French natives. Lesions were located in the left hemisphere in 26 patients and in the right in 10. The degree of handedness in the patients was assessed using the Edinburgh Handedness Inventory test.41 In this study, 35 patients were right handed (mean score 85 6 10 standard deviation), whereas one patient was left handed (score 250). During brain tumor surgery, we routinely use the awake surgery technique if any surgical mapping can be useful for the surgical procedure and if the case is truly appropriate for such a procedure. Although rare, the right hemisphere can occasionally be involved in some language processes in right-handed people.48 Accordingly, selected patients were offered awake surgery together with our standard brain mapping technique (for example, patients with right-hemisphere brain tumors whose removal was expected to require extreme delicacy because, for instance, mapping of the rolandic area was required or bilateral language representation was suspected).48 We think that dominant and nondominant rolandic areas are better studied during naming or reading tasks, especially for parts related to face or tongue movement. All patients and their families gave informed consent to undergo the study of their language areas using direct brain mapping. Language Testing and Color Naming
All patients underwent pre- and postoperative language examinations to assess their ability to perform the required tasks. This testing included the evaluation of written and oral understanding and naming, language fluency, reading, computation, dictation, repetition, copying, and object 28
handling. In this study, strict inclusion criteria were used; dysphasic patients (those with . 10% errors in the naming tests) were excluded. Visual field testing was performed to identify patients with visual field defects associated with achromatopsia; these patients were not included in this study. Preoperatively, the ability of the patients to identify and name colors was specifically tested. Patients were asked to perform the following tasks: 1) name different colors on a color data set and point to the color when a color name was given by the examiner or read by the patient; 2) match colors on an isoluminant color plate; 3) match different objects or concepts with an appropriate color (for example, banana = yellow or blood = red); 4) name the color of some objects shown on a data set of different colored objects; and 5) name black, gray, and white object figures. Only patients who completed these tests without error were included in the study. Brain Mapping Tasks
Our routine brain mapping procedure consists of two kinds of tasks: a naming task (using a data set of noncolored objects) to search for standard anomia and a task of reading various sentences and numbers aloud. Before concluding the direct brain mapping procedures, we routinely confirm essential naming sites by repeating an object-naming task (standard procedure: object naming, sentence and number reading, and object naming). Among the numerous teams using direct brain mapping in neurosurgical procedures, the choice of the applied tasks is variable. Naming and reading tasks (with various materials) are generally used, sometimes in conjunction with counting, short-term memory, or verb-generation tasks.39 Sentence comprehension and visual or auditory word generation have also been used by other teams.45 In this series, following the reading and object-naming tasks, we used not an object- but a color-naming task for the repeated naming procedure. We used separate sheets of paper painted with 11 different basic colors corresponding to the 11 basic color terms used in French (according to the criteria of Berlin and Kay3). The different colored sheets were randomly presented to the patients, and cortical mapping of color naming was performed while they were asked to name the colors. Once a color-naming site was found using the visually presented task, this site was stimulated while patients were asked to give the name of objects without visual input and to answer aurally presented questions (such as, “What is the usual color of a banana?” “Of snow?” “Of a letterbox?”). This process was completed in an attempt to distinguish color-naming sites related to visual inputs from those related to language input and retrieved from semantic knowledge and mental imagery. Indeed, it has been hypothesized that the modality of task presentation (visual compared with auditory) during object naming could also modify brain mapping data.31 For practical reasons, we did not perform mapping of auditory color naming in the entire cortical region studied but only at those sites where interference in visual color-naming had already occurred. The entire brain mapping procedure used in this study is summarized in Fig. 1. Because the color-naming procedure was substituted for the repeated standard objectnaming procedure, mapping of color naming did not exJ. Neurosurg. / Volume 104 / January, 2006
15658_pt.5
12/21/05
5:43 PM
Page 29
Color-naming areas
FIG. 1. Schematic of brain mapping procedure used in this study. Single asterisk represents car, wheelbarrow, plane, chair, bird, ball, lion, tomato, grapes, bomb, bottle, boat, key, strawberry, screwdriver, rose, saw, chicken, computer, slippers, bread, hands, book, rubber, clock, fridge, pencil, flute, camera, and door. Double asterisks refer to black, white, red, green, yellow, blue, brown, purple, pink, orange, and gray.
tend the duration of the brain mapping procedure, which was usually completed in less than 30 minutes. Our strategy was to spare areas of reading and object- or color-naming that were found during tumor removal by resecting no more than 1 cm from eloquent cortex (distance of the resection margin from the nearest functional site). Task Difficulty Comparison
The findings on different tasks may be affected or biased by the frequency or familiarity of the items presented during each task. Perhaps the object- and color-naming tasks were not equivalent in terms of complexity, which could entail different brain mapping results for both tasks. Indeed, some authors have demonstrated that the frequency or familiarity of items used for testing category effect in aphasic individuals can be strong factors influencing performance. Therefore, low-frequency or low-familiarity objects whose names are more difficult to retrieve than those of common items could lead to a difference in task performance that may not reflect a category effect but rather a nonspecific defect.6 Lexical frequency in the lists of items used in our two naming tasks (objects and colors) was compared using the French database Brulex (index table of word frequencies). No significant difference in lexical frequency was found between the two lists (Mann–Whitney U-test, p = 0.10). Therefore across-task differences for naming interferences cannot be accounted for by lexical-frequency effects. Cortical Mapping Procedures
In all patients we used the awake surgery technique, as J. Neurosurg. / Volume 104 / January, 2006
previously described.48 We used a neuronavigational system in all but four patients. Intraoperative cortical stimulation was applied to localize areas of functional cortex after determining the afterdischarge threshold with electrocorticography. All patients were tested at the same site for both reading and naming. When we started a direct cortical stimulation procedure, we chose a substantial number of sites on the brain surface. We stimulated the same areas during the entire procedure to test both naming and reading. The number of sites studied was variable and depended on the size of the craniotomy. (We usually used from 8–30 different stimulation sites in each patient.) The cortex was directly stimulated using a bipolar electrode consisting of 1-mm contacts placed 6 mm apart (Nimbus cortical stimulator; Newmedic, Toulouse, France). Cortical mapping procedures and all patient responses were recorded on video. Intraoperative photographs of the brain were obtained during cortical stimulation, indicating the positive (marked L) or negative (marked N) naming sites. Small color-naming tickets (marked CN) were used in the intraoperative pictures to distinguish the color-naming from object-naming procedure. Definitions for Determining Color-Naming Sites
The effect of stimulation had to be reproduced in a site at least three times for it to be validated as a language site. The sites effecting no reproducible anomia or speech arrest (in object or color mapping) were not included in this study. A site was defined as “color naming–specific” when no object-naming or reading interference occurred at that site. 29
15658_pt.5
12/21/05
5:43 PM
Page 30
F. E. Roux, et al.
FIG. 2. Illustration depicting the object- and color-naming sites found in the left hemisphere of 26 patients. To perform statistical analysis, we chose to define regions by using the gyral/sulcal anatomy. Cortices of the right and left hemispheres have been divided into several regions delimited by dotted lines. The majority of the interferences discovered during naming tasks were found in cortical areas common to object and color naming, although specific object or color interference also occurred.
Obviously, we cannot rule out the possibility that these sites induced stimulation interference for other functions (or other naming categories) not tested in this study. Furthermore, during direct brain mapping, single anomia sites not confirmed on repeated stimulation can occasionally be found38 and the interpretation of this phenomenon remains complex. In this study, sites effecting no reproducible language interference were excluded. In our work, we used the general term “color–naming interference” to refer to all color-naming difficulties our patients experienced while undergoing stimulation. “Color anomia” was used when a patient without overall language impairment had repeated difficulty in finding the exact name of the color being visually or aurally presented (as in, “This is. . . . Oh, I know this color. . . . This is. . . . I don’t remember the name.”). Other color-naming interferences were noted during brain mapping such as speech arrest (the patient said nothing after color presentation). Speech arrest can occur without visible face, mouth, or tongue contractions, following either frontal or temporoparietal stimulation. Speech arrest can also be noted during stimulation of pre- and postcentral gyri, generally with face, mouth, or tongue contractions. Hesitation, use of periphrases to describe colors, or paraphasias can also be observed during direct brain mapping. Finally, although rather imprecise, the terms “Broca center” and “Wernicke center” are still commonly used in the medical literature; for us the former term 30
referred to the posterior part of the dominant F3 (and only this gyrus) and the latter to the posterior part of the T1 as well as the supramarginal and angular gyri. A broad region including the dominant supramarginal, angular, and posterior parts of the T1, T2, and T3 was called the “posterior temporal associative cortex.” The posterior parts of temporal gyri were delineated posterior to an imaginary line extending the postcentral sulcus. Postoperative Tests
Postoperatively, patients were asked to repeat the preoperative tests to detect any language difficulties. The results were compared with the intraoperative findings. These postoperative tests can be considered important in determining either the degree of recovery or surgery-related deficits. In another study,48 we encountered difficulties and uncertainties in the analysis of surgery-related defects given that these defects were sometimes related to postsurgical stress or fatigue in the patients. Analysis of postoperative testing was further complicated by limited follow up of patients with high-grade brain tumors or metastases, nonsurgical factors such as environmental ones (for example, speech therapy), or other individual factors. Statistical Analysis
During this 2-year study, all data regarding brain mapJ. Neurosurg. / Volume 104 / January, 2006
15658_pt.5
12/21/05
5:43 PM
Page 31
Color-naming areas ping results were integrated into an Excel database. In presenting our data, we defined regions according to gyral/ sulcal anatomy. For instance, the supramarginal gyrus was considered a region, as was the angular gyrus. Large gyri, such as the temporal gyri, were arbitrarily divided into three segments by drawing an imaginary line extending to the pre- and postcentral sulci (Fig. 2). Fifteen regions were defined in the left hemisphere and 15 in the right hemisphere. We analyzed brain mapping data by separating color-naming sites from object-naming and reading sites (in each hemisphere). Secondarily, we analyzed the interdependence of color- and object-naming sites as a function of the location and the type of response produced. To avoid confusion in the presentation of our results, we localized all of the color-naming sites found in the left hemisphere of a standard brain, according to operative anatomical data, whereas the type of response and the variations in the responses obtained are presented elsewhere in the text. Right-hemisphere results are presented in the text. For each zone, the percentage of responses was analyzed according to the number of stimulations performed. Statistical analysis was performed using unbalanced repeated-measures analysis of variance after transforming the percentage values to obtain a normal distribution.57 The respective probability value was corrected for lack of sphericity using the Box conservative epsilon. Differences were estimated to be significant at a probability value less than 0.05. Analysis was performed using commercially available software (Stata, version 7.0; Stata Corp., College Station, TX). An independent statistician (V.L.) performed the statistical analysis. Results In 34 of the 36 patients who performed the reading, object-naming, and color-naming tasks, at least one language site was found (indicated by anomia, hesitation, and speech or reading arrest). In two of the right-handed patients, none of the language-related areas was found during stimulation of the nondominant upper and lower parietal regions. Some cortical regions were not studied such as the bilateral occipital lobes, basal temporal lobes (lingual and fusiform gyri), nondominant T2 and T3, and interhemispheric regions. Other cortical regions were studied at least once: right and left F1, F2, and F3; right and left pre- and postcentral gyri; parietal lobes; T1; and dominant T2 and T3. The maximal current that did not evoke afterdischarges ranged from 4.5 to 8 mA. None of the patients experienced generalized seizures intraoperatively. Language interferences found in the pre- and postcentral gyri (considered as nonspecific language interferences due to articulatory mechanisms) were not included in the final analysis. All of the found object- or color naming–interference sites were spared during surgery. Several features of the stimulation sites specific to color naming were studied: location and type of interference sites, number of sites found, category specificity, and modality specificity. Location of Color Naming–Interference Sites
Interferences in color naming were predominantly found in traditional language regions: dominant inferior frontal gyrus (p , 0.003) and supramarginal, angular, posterior J. Neurosurg. / Volume 104 / January, 2006
portions of the superior temporal gyri (p = 0.05). Note, however, that some color-naming areas were occasionally found outside these regions (Fig. 2). With regard to object-naming or reading sites, color-naming interferences were exclusively localized in small parts of cortical areas (, 1 cm2). Variability of Language Organization Based on Associative Region
High individual variability in language organization among patients was observed. The dominant left hemisphere was studied in 26 patients. In the 13 patients in whom the Broca center was studied, four (30%) had no language interference in this region regardless of whether they were performing an object, color, or reading task. Among the 13 patients in whom the posterior temporal associative region was studied, at least one interference area (object naming or reading) was found, although no color-naming area was found in two of them. The number of language interferences occurring among these patients varied from only one to eight different interferences (mean 2.8 interferences). Variability in Language Organization Based on Tasks Used
We found high individual variability in language organization according to the different tasks used. Color naming– interference sites overlapped standard object-naming areas in some but not all cases. Among the 26 patients whose dominant left hemisphere was studied, strict concordance between the results of object and color mappings was found in only nine (36%). The remaining 17 patients demonstrated different brain mapping results, according to the two different naming sets (object and color). Overall, fewer color naming–interference sites (mean 2.7 sites/patient) compared with object naming– (mean 3.3 sites/patient) or reading (mean 3.6 sites/patient) interference sites were found, although this difference did not reach statistical significance (p = 0.075). With regard to the Broca center and the posterior temporal associative areas, the different naming tasks had not only common interference sites on mapping but also different sites for the two tasks performed. Color-Specific Cortical Areas
Among the 26 patients whose dominant left hemisphere was studied, 13 color naming–specific interference sites were identified in 10 patients (that is, sites of interference in color naming only and not object naming or reading). Color naming-specific sites were found in the dominant F3 and supramarginal and angular gyri as well as the posterior part of the temporal gyrus (Figs. 3 and 4). Hypothetical Color-Naming Center
We determined that the same cortical site could be identified as a color naming–specific area or a center from one subject to another. Although color naming–specific sites were predominately found in the supramarginal and angular gyri, this incidence was not statistically significant. Among the cortical regions tested, no specific cortical region was found to be consistently involved in color naming (p . 0.05). 31
15658_pt.5
12/21/05
5:43 PM
Page 32
F. E. Roux, et al.
FIG. 3. Intraoperative photographs and drawing showing naming specificity (object compared with color naming) in the Broca center and T1. Direct brain mapping was performed in a 45-year-old right-handed man with a left perisylvian and temporal lobe astrocytoma. Preoperatively, no language deficit was revealed after language testing and no color discrimination nor classification disturbances were noted either. During brain mapping, five object-naming interference sites were found in the Broca center (L). An object naming hesitation (HE) site was also found, repeatedly, in the temporal lobe. During the color naming (CN) procedure we found only four sites of color-naming interference in the Broca center and no hesitation in color naming in the temporal lobe. To improve the understanding of the intraoperative pictures, cortical sites producing no language impairment were not systematically labeled with a sterile ticket. M = motor, rolandic area; N = negative areas.
Types of Color-Naming Interferences
Different types of interferences occurred, ranging from speech arrest to typical color anomia. Patients sometimes acknowledged that they perceived colors but were unable to name them (for example, “Oh, I know this color. Yes, I know but I can’t find the name.”). Repetitions, hesitations, wrong color naming, or typical paraphrasis also occurred. Periphrases were also given (for example, “It’s like the color of strawberries” for red). In most cases, the same pattern of interference occurred for object or color naming (that is, an object anomia site producing color anomia on stimulation). Most commonly, interference occurred in the form of speech arrest (51% of color-interference sites). Anomia sites were more rare (21% of color interference sites) and were not preferentially found in temporal areas as opposed to frontal areas (p . 0.05). Modality-Specific Color Naming
Once color naming–interference sites were detected, those precise sites were stimulated and patients were asked to attribute a color to a standard object while keeping their 32
eyes closed (for example, “What color is the sun?”). This task addressed visual knowledge of color as an attribute of objects or concepts. In these sites and in most patients, color naming from visual stimuli and from semantic knowledge (eyes closed) was equally impaired. Nevertheless, in five patients, differential impairment was noted between visually stimulated compared with semantic retrieval in the color-naming task. In these patients, although visually presented color-naming tasks were impaired during stimulation, semantic color naming (for example, “What color is grass?”) was performed without difficulty. This differential impairment in naming was observed at five sites—the angular gyrus and the supramarginal and the posterior parts of T2 and T3—illustrating that modality-specific areas can be found during color naming (Fig. 5). Note that the objectnaming task was performed with visually presented stimuli only in this study; thus, we could not demonstrate that object-naming impairment was also modality-specific at these sites. Right-Hemisphere Language Mapping
Excluding the pre- and postcentral gyri, language interJ. Neurosurg. / Volume 104 / January, 2006
15658_pt.5
12/21/05
5:43 PM
Page 33
Color-naming areas
FIG. 4. Intraoperative photographs and drawing showing that naming specificity was performed (object compared with color naming) in T1 and T2. Direct brain mapping was performed in a 55-year-old right-handed man with a left anterior temporal high-grade astrocytoma. During brain mapping, two sites of object anomia or speech arrest were found in the Broca center. Three object anomia sites (L) were also repeatedly found in T1. During the color-naming procedure we found two sites in the Broca center where color naming was impaired and only one color naming– interference site in T1. In T2, one specific color anomia site (only with the visual color task) was found (gray dot). To improve the understanding of the interoperative pictures, cortical sites producing no language impairment were not labeled.
ferences rarely occurred in the right hemisphere; among the 10 patients (nine right-handed and one left-handed) whose right hemisphere was studied, we found only six color-interference areas, although none were color-specific. In a left-handed patient whose right F2 and F3 were studied, we found a common interference area for all the tasks studied in the right F3. Postoperative Testing
Difficulties in both naming tasks were detected postoperatively in 11 patients (31%), probably due to brain retraction or postoperative edema. These difficulties were slow naming, naming hesitation or mistakes, and phonemic errors. No patient had a specific color-naming deficit. Among these 11 patients, six underwent repeated testing a few weeks after surgery and two experienced significant persistent object- and color-naming difficulties. For various reasons, the remaining five patients were lost to follow up and could not be retested. J. Neurosurg. / Volume 104 / January, 2006
Among the 10 patients in whom a color-specific site was found, we spared all of these areas, although in three patients the tumor margins of removal were very close (1– 2 cm) to color-specific areas. Each of these three patients experienced postoperative color-naming difficulties, which resolved in two of them within 2 months. The remaining patient (Fig. 5) had persistent object- and color-naming difficulties when he was examined after the 2-month follow up. He died shortly thereafter of his progressive high-grade brain tumor. Discussion Color Naming as a Category-Specific Task In this study, color naming–interference sites were found in traditional language areas, in areas identical to or distinct from common object naming– or reading-interference sites. Whereas anatomical segregation of different naming categories was sometimes observed, no brain region was specif33
15658_pt.5
12/21/05
5:43 PM
Page 34
F. E. Roux, et al.
FIG. 5. Example of brain mapping results by using three different tasks (reading and object and color naming) and two modalities (visual compared with auditory color naming) in the posterior temporal associative region. Language interferences are represented using different colors for each task: reading (green), object naming (orange), visual color naming (dark blue), and auditory color naming (sky blue). No language interference area was found during subcortical mapping. Although a color naming–specific site (for both modalities) was found in T1 (white arrow), no reading interference or auditory color-naming interference was noted (black arrow) in a cortical site in the T3. Partitioned circles represent areas of language testing; plain circles indicate that no interference site was found for the tasks in this study; dotted lines indicate the area of cortectomy for removal of a deep-seated T3 benign tumor.
ically and consistently involved in color naming. High variability among individuals and tasks was observed in this group of patients. We also found occasional differences in brain mapping results when we used semantic retrieval of color names from auditory input compared with visually stimulated color-naming tasks. Color naming is an ability acquired by training during childhood. Color-naming performance as well as other language functions can vary, and several factors must be considered, for instance, sex (girls perform better than boys in learning color names4) or the context of acquired language deficits.26,54 Variations in color-naming performance can also be noted according to different languages,3,11 the number of basic color terms used in any given language being variable.3 For instance, English has 11 basic color terms (black, white, red, green, yellow, blue, brown, purple, pink, orange, and gray), whereas the number of basic color terms varies 34
in other languages (for example, the Apache language has only five basic color terms; Russian, 12 basic color terms). The reasons for these variations are debatable but have been attributed to the cultural evolution of color nomenclature3 rather than differing abilities in discriminating color. The increase in the number of basic color names in different civilizations could be related to an increase in technical or scientific needs. Medicine and philosophy have played an important role in the development of color names, as evidenced by the significant increase in color terms in ancient Greek in the Hippocratic corpus from the fourth century BC or the works of Aristotle and Theophrastus compared with Homeric prose, which was very poor in color terms.53 Color-naming defects have long been described.56 Patients with such defects are able to match or sort colors, but cannot name them or point out the appropriate color once a color name is given.17,20 Color-naming deficits generally ocJ. Neurosurg. / Volume 104 / January, 2006
15658_pt.5
12/21/05
5:43 PM
Page 35
Color-naming areas cur in aphasic disorders.26 Cases of specific color-naming deficits or, conversely, selective preservation47 have been described.26,42 It is now acknowledged that there are at least two types of color-naming deficits: damage to language-related cortical areas26,42 and visuoverbal disconnection as a consequence of a lesion of the connecting pathways in the dominant occipitotemporal junction or lesions of the splenium.20,54,59 Specific color-naming deficits have been noted mainly in disconnection syndromes involving pure alexia— “alexia without agraphia”13,20—although the existence of cortical areas specifically involved in color naming has been postulated (for instance, in the inferior part of the left parietal lobe).17,42 By applying stimulation, we found that cortical color-naming or reading sites were exclusively localized in small parts of cortical areas (, 1 cm2). Some color naming–specific sites were also found. The extreme parcelling of these functions at the cortical level could explain why we found a dissociation between a color-naming defect and an ability to read; it is rare for visual color-naming defects to occur without alexia in clinical practice. Furthermore, in the patients suffering from alexia and color-naming defects, the cause is more often a color-discrimination problem (with a typical alexia without an agraphia syndrome) rather than a real color-naming problem. Excluding disconnection syndromes, the occurrence of a pure colornaming defect at the cortical level would require a lesion in the color-naming site while sparing other language sites; this uncommon condition might explain the rarity of pure color naming–defect syndromes. Color naming can be considered a category-specific task. Category-specific deficits have long been described in aphasic patients,23,55 as have cases of category-specific preservation.35 The most common dissociation has been found in naming animate as opposed to inanimate objects,23 although impaired comprehension of color or body part names has long been identified.19,21 A clear selective preservation of color naming in semantic dementia47 or in more general aphasic symptoms has been noted by some authors.16,18,37 These authors have reported cases of relative or strict selective preservation of color naming, suggesting that color naming could be a dissociable lexical category with dedicated functional and anatomical structures in some patients. The anatomical differentiation within the brain for some object categories is a puzzling phenomenon. The development of category-specific modules could be related to evolutionary factors or, in individuals, to a training effect for a specific topic.1,6 Authors demonstrating specific deficits in material acquired through education or special interests in individuals have favored the hypothesis that categorical knowledge could be organized, at least partially, in dedicated neuronal structures.1 Recent functional imaging studies have shown that, to a certain degree, category-dependent neural systems can be identified during naming.33,40 The basal temporal lobe has been suspected of supporting some category-specific naming areas,8,44 especially specific color-naming areas.32 Using positron emission tomodensitometry, the inferior temporal lobe has been shown to be important in animal discrimination, whereas the posterior middle temporal area could be activated in tool processing.14,32,43 Nevertheless, other discrete category-specific language areas have been situated elsewhere, such as in the frontal cortex.25 Recently, specific cortical areas related to different categories of objects were J. Neurosurg. / Volume 104 / January, 2006
found on direct stimulation, implying possible anatomical segregation of different naming categories.25,44 In this study, we found that color naming involved discrete color-naming areas in various language-related areas. Color Naming as a Modality-Specific Task
In a few patients we found that some color-naming areas could be modality-specific given that a deficit was observed on visual presentation, whereas they performed normally on auditory presention of semantic information (visual compared with verbal input). This phenomenon of differential naming was observed at five sites: the angular gyrus as well as the supramarginal and posterior parts of T2 and T3. This modality-specific effect suggests that in these patients, the stimulated portions of cortex localized in the inferior parietal region or the posterior temporal cortex depend on visual input to access lexical labels of colors that could be accessed normally from other (auditory) input. A similar pattern of dissociated performance depending on a specific input modality has been described in rare cases of aphasia.7 Differential impairments in naming have been reported in lateral temporal regions when a different modality (auditory as opposed to visual) of task presentation was used.31,55 To explain these modality-specific losses of semantic knowledge, the existence of multiple, partially independent, semantic systems has been hypothesized.55 Few cortical stimulation studies have been focused on the possibility of cortical areas specific to a task modality. In a study with eight patients, Malow, et al.,31 found that naming in response to auditory input was significantly more impaired than naming in response to visual input during stimulation of the dominant anterior and posterior lateral temporal cortices. On the other hand, auditory and visual naming were equally impaired in the dominant frontal and inferior temporal cortices. Those language mapping results may depend on not only naming categories, but also modalities of task presentation used in mapping, which reflects the extensive degree of specialization that can sometimes occur in language organization. Limits of the Study
We attempted to rule out other explanations for the category effects. For instance, perhaps apparent category-specific defects are really nonspecific difficulties amplified by low-frequency, low-familiarity objects (that is, differences in the presented tasks). We tried to exclude this factor by choosing two different naming categories with similar lexical frequency. The intensity of cortical stimulation may also influence the degree of language interference.29,45 In this study, we used the same stimulation intensity for object- and color-naming procedures. The apparently different anatomical localizations observed between color- and object-naming areas could be functional rather than anatomical in nature if color and object naming were disrupted by different levels of stimulation intensity. Although this explanation seems improbable when using two naming tasks in the same experimental design, the possibility of differential thresholds to stimulation must be acknowledged. Conclusions It has been shown that direct brain mapping results can be modified by the use of verb-generation tasks39 or reading 35
15658_pt.5
12/21/05
5:43 PM
Page 36
F. E. Roux, et al. tasks48 instead of a naming task, but they can also be modified by increasing task complexity39 or by testing the same task in different languages (in bilingual patients, for example).28,49 In this study, at sites where we considered color naming to be a category-specific task, relative specialization of the cortical areas involved in both object and color naming was found in all standard language areas, implying partial anatomical segregation of different naming categories. This relative degree of cortical specialization in naming could explain, at least partly, selective or preferential color-naming impairments sometimes observed in aphasic patients. Finally, we observed that language brain mapping results may depend on the modality of task presentation. Although these findings do not alter standard cortical mapping methodology, neurosurgeons using this technique must remember that naming categories could influence brain mapping results. As proposed in some recent functional imaging studies, the question of whether dedicated cortical areas exist for some categories must be raised. Consequently, the choice of the tasks that must be made in function of the brain region studied could arise. Acknowledgments We thank Denise Géblé for her help in supplying the references for this article, and Dr. Marc Vironneau and Marie-Claire Provost for assistance during brain mappings. We are also indebted to Rebecca Rinieri, Noëlle Ricard, and Christiane Lubrano for editorial assistance. References 1. Allison T, McCarthy G, Nobre A, Puce A, Belger A: Human extrastriate visual cortex and the perception of faces, words, numbers, and colors. Cereb Cortex 4:544–554, 1994 2. Barrett NA, Large MM, Smith GL, Michie PT, Karayanidis F, Kavanagh DJ, et al: Human cortical processing of colour and pattern. Hum Brain Mapp 13:213–225, 2001 3. Berlin B, Kay P: Basic Color Terms: Their Universality and Evolution. Berkeley: University of California Press, 1969 4. Bornstein MH: On the development of color naming in young children: data and theory. Brain Lang 26:72–93, 1985 5. Bornstein MH, Kessen W, Weiskopf S: The categories of hue in infancy. Science 191:201–202, 1976 6. Caramazza A, Shelton JR: Domain-specific knowledge systems in the brain the animate-inanimate distinction. J Cogn Neurosci 10: 1–34, 1998 7. Chanoine V, Ferreira CT, Demonet JF, Nespoulous JL, Poncet M: Optic aphasia with pure alexia: a mild form of visual associative agnosia? A case study. Cortex 34:437–448, 1998 8. Chao LL, Haxby JV, Martin A: Attribute-based neural substrates in temporal cortex for perceiving and knowing about objects. Nat Neurosci 2:913–919, 1999 9. Chao LL, Martin A: Cortical regions associated with perceiving, naming, and knowing about colors. J Cogn Neurosci 11:25–35, 1999 10. Corbetta M, Miezin FM, Dobmeyer S, Shulman GL, Petersen SE: Attentional modulation of neural processing of shape, color, and velocity in humans. Science 248:1556–1559, 1990 11. Critchley M: Acquired anomalies of colour perception of central origin. Brain 88:711–724, 1965 12. Damasio A, Yamada T, Damasio H, Corbett J, McKee J: Central achromatopsia: behavioral, anatomic, and physiologic aspects. Neurology 30:1064–1071, 1980 13. Damasio AR, Damasio H: The anatomic basis of pure alexia. Neurology 33:1573–1583, 1983
36
14. Damasio H, Grabowski TJ, Tranel D, Hichwa RD, Damasio AR: A neural basis for lexical retrieval. Nature 380:499–505, 1996 15. Davidoff J, De Bleser R: Impaired picture recognition with preserved object naming and reading. Brain Cogn 24:1–23, 1994 16. De Renzi E, di Pellegrino G: Sparing of verbs and preserved, but ineffectual reading in a patient with impaired word production. Cortex 31:619–636, 1995 17. De Vreese LP: Category-specific versus modality-specific aphasia for colors: a review of the pioneer case studies. Int J Neurosci 43: 195–206, 1988 18. Denes G, Meneghello F, Vallese F, Vanelli L: Task-dependent noun retrieval deficit: an experimental case study. Neurocase 2: 35–43, 1996 19. Gerstmann J: Syndrome of finger agnosia, disorientation for right and left, agraphia and acalculia. Arch Neurol Psych 44:398–408, 1940 20. Geschwind N, Fusillo M: Color-naming defects in association with alexia. Arch Neurol 15:137–146, 1966 21. Goodglass H, Budin C: Category and modality specific dissociations in word comprehension and concurrent phonological dyslexia. Neuropsychologia 26:67–78, 1988 22. Goodglass H, Wingfield A, Hyde MR, Theurkauf JC: Category specific dissociations in naming and recognition by aphasic patients. Cortex 22:87–102, 1986 23. Hillis AE, Caramazza A: Category-specific naming and comprehension impairment: a double dissociation. Brain 114 (Pt 5): 2081–2094, 1991 24. Howard RJ, ffytche DH, Barnes J, McKeefry D, Ha Y, Woodruff PW, et al: The functional anatomy of imagining and perceiving colour. Neuroreport 9:1019–1023, 1998 25. Ilmberger J, Rau S, Noachtar S, Arnold S, Winkler P: Naming tools and animals: asymmetries observed during direct electrical cortical stimulation. Neuropsychologia 40:695–700, 2002 26. Kinsbourne M, Warrington EK: Observations on Colour Agnosia. J Neurol Neurosurg Psychiatry 27:296–299, 1964 27. Lange J: Agnosien, in Foester BA (ed): Handbuch Neurologie. Berlin: Springer-Verlag, 1936, Vol 6, pp 807–885 (Reference unverified) 28. Lucas TH II, McKhann GM II, Ojemann GA: Functional separation of languages in the bilingual brain: a comparison of electrical stimulation language mapping in 25 bilingual patients and 117 monolingual control patients. J Neurosurg 101:449–457, 2004 29. Luders H, Lesser RP, Hahn J, Dinner DS, Morris HH, Wyllie E, et al: Basal temporal language area. Brain 114 (Pt 2):743–754, 1991 30. Luzzatti C, Davidoff J: Impaired retrieval of object-colour knowledge with preserved color naming. Neuropsychologia 32: 933–950, 1994 31. Malow BA, Blaxton TA, Sato S, Bookheimer SY, Kufta CV, Figlozzi CM, et al: Cortical stimulation elicits regional distinctions in auditory and visual naming. Epilepsia 37:245–252, 1996 32. Martin A, Haxby JV, Lalonde FM, Wiggs CL, Ungerleider LG: Discrete cortical regions associated with knowledge of color and knowledge of action. Science 270:102–105, 1995 33. Martin A, Wiggs CL, Ungerleider LG, Haxby JV: Neural correlates of category-specific knowledge. Nature 379:649–652, 1996 34. McKeefry DJ, Zeki S: The position and topography of the human colour centre as revealed by functional magnetic resonance imaging. Brain 120:2229–2242, 1997 35. McKenna P, Warrington EK: Category-specific naming preservation: a single case study. J Neurol Neurosurg Psychiatry 41: 571–574, 1978 36. Miceli G, Fouch E, Capasso R, Shelton JR, Tomaiuolo F, Caramazza A: The dissociation of color from form and function knowledge. Nat Neurosci 4:662–667, 2001 37. Miozzo A, Soardi M, Cappa SF: Pure anomia with spared action naming due to a left temporal lesion. Neuropsychologia 32: 1101–1109, 1994 38. Ojemann GA: Cortical organization of language. J Neurosci 11: 2281–2287, 1991
J. Neurosurg. / Volume 104 / January, 2006
15658_pt.5
12/21/05
5:43 PM
Page 37
Color-naming areas 39. Ojemann JG, Ojemann GA, Lettich E: Cortical stimulation mapping of language cortex by using a verb generation task: effects of learning and comparison to mapping based on object naming. J Neurosurg 97:33–38, 2002 40. Okada T, Tanaka S, Nakai T, Nishizawa S, Inui T, Sadato N, et al: Naming of animals and tools: a functional magnetic resonance imaging study of categorical differences in the human brain areas commonly used for naming visually presented objects. Neurosci Lett 296:33–36, 2000 41. Oldfield RC: The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 9:97–113, 1971 42. Oxbury JM, Oxbury SM, Humphrey NK: Varieties of colour anomia. Brain 92:847–860, 1969 43. Perani D, Cappa SF, Bettinardi V, Bressi S, Gorno-Tempini M, Matarrese M, et al: Different neural systems for the recognition of animals and man-made tools. Neuroreport 6:1637–1641, 1995 44. Pouratian N, Bookheimer SY, Rubino G, Martin NA, Toga AW: Category-specific naming deficit identified by intraoperative stimulation mapping and postoperative neuropsychological testing. Case report. J Neurosurg 99:170–176, 2003 45. Pouratian N, Cannestra AF, Bookheimer SY, Martin NA, Toga AW: Variability of intraoperative electrocortical stimulation mapping parameters across and within individuals. J Neurosurg 101: 458–466, 2004 46. Price CJ, Moore CJ, Humphreys GW, Frackowiak RS, Friston KJ: The neural regions sustaining object recognition and naming. Biol Sci 263:1501–1507, 1996 47. Robinson G, Cipolotti L: The selective preservation of colour naming in semantic dementia. Neurocase 7:65–75, 2001 48. Roux FE, Lubrano V, Lauwers-Cances V, Tremoulet M, Mascott CR, Demonet JF: Intra-operative mapping of cortical areas involved in reading in mono- and bilingual patients. Brain 127: 1796–1810, 2004 49. Roux FE, Tremoulet M: Organization of language areas in bi-
J. Neurosurg. / Volume 104 / January, 2006
50. 51. 52. 53. 54. 55. 56. 57. 58. 59.
lingual patients: a cortical stimulation study. J Neurosurg 97: 857–864, 2002 Shuren JE, Brott TG, Schefft BK, Houston W: Preserved color imagery in an achromatopsic. Neuropsychologia 34:485–489, 1996 Ungerleider LG: Functional brain imaging studies of cortical mechanisms for memory. Science 270:769–775, 1995 Verrey: Hémiachromatopsie droite absolue. Archives d’Ophtalmologie 8:289–301, 1888 (Reference unverified) Villard L: Couleurs et Vision dans l’Antiquité Classique. Rouen, France: Presses Universitaires de Rouen, 2002 Vincent FM, Sadowsky CH, Saunders RL, Reeves AG: Alexia without agraphia, hemianopia, or color-naming defect: a disconnection syndrome. Neurology 27:689–691, 1977 Warrington EK, Shallice T: Category specific semantic impairments. Brain 107:829–854, 1984 Wilbrand H: Ophtalmiatrische beiträge zur diagnostik der gehirn-krankenheit. Wiesbaden: JF Bergmann, 1884 (Reference unverified) Zar J: Biostatistical Analysis, ed 4. Englewood Cliffs, NJ: Prentice Hall, 1999 Zeki S, Marini L: Three cortical stages of colour processing in the human brain. Brain 121:1669–1685, 1998 Zihl J, von Cramon D: Colour anomia restricted to the left visual hemifield after splenial disconnexion. J Neurol Neurosurg Psychiatry 43:719–724, 1980
Manuscript received January 5, 2005. Accepted in final form September 29, 2005. Address reprint requests to: Franck-Emmanuel Roux, M.D., Ph.D., Service de Neurochirurgie et Institut National de la Santé 455, Hôpital Purpan, F-31059 Toulouse, France. email: rouxfran@ compuserve.com.
37
