Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Int J Geriatr Psychiatry. Author manuscript; available in PMC 2010 August 1.
Published in final edited form as:
Int J Geriatr Psychiatry. 2009 August; 24(8): 885–893.
doi:  10.1002/gps.2229
PMCID: PMC2743128

The neural correlates of naming and fluency deficits in Alzheimer’s disease: an FDG-PET study



To examine the neural processes associated with language deficits in Alzheimer’s disease (AD), and in particular to elucidate the correlates of confrontation naming and word retrieval impairments.


Sixty patients with probable AD were included. Confrontation naming was assessed using the number of words spontaneously named correctly on the Boston Naming Test. We recorded the number of additional words stated following phonemic cuing. We also assessed phonemic (FAS) and semantic (supermarket items) fluency. We then correlated performance on each measure with resting cortical metabolic activity using FDG-PET images.


We found that poorer ability to spontaneously name an object was associated with hypometabolism of bilateral inferior temporal lobes. In contrast, when a phonemic cue was provided, successful naming under this condition was associated with higher metabolic activity in bilateral inferior frontal gyrus (IFG), right superior frontal gyrus (SFG), left temporal, and occipital regions. Consistent with these findings, we found that poorer semantic fluency was associated with hypometabolism in regions including both IFG and temporal regions, and poorer phonemic fluency was associated with hypometabolism in only left IFG. Across analyses, measures that required cued retrieval were associated with metabolism in the left IFG, whereas measures taxing semantic knowledge were associated with metabolic rate of left temporal cortex.


Naming deficits in AD reflect compromise to temporal regions involved in the semantic knowledge network, and frontal regions involved in the controlled retrieval of information from that network.

Keywords: Alzheimer’s disease, language, naming, temporal lobe, inferior frontal gyrus, PET imaging

key points

  1. Naming deficits in Alzheimer’s disease reflect alterations to both temporal and frontal regions.
  2. Frontal lobe dysfunction may disrupt naming by impairing controlled retrieval processes.
  3. Temporal lobe dysfunction may impair naming by disrupting the processing of semantic information.


Alzheimer’s disease (AD) is the most frequent cause of dementia in older adults. Patients with AD experience a range of neurocognitive deficits, including problems with attention, learning and memory, visuospatial functioning, language, and executive/frontal lobe functioning (Lezak et al., 2004). Research has explored the neurobiological underpinnings of distinct neurocognitive deficits in AD, with much of the focus on memory and corresponding dysfunction of the medial temporal lobes (e.g. Desgranges et al., 2002; Eustache et al., 2004). Subtle deficits in language functioning can be detected in the earliest stages of the disease (Forbes-McKay and Venneri et al., 2005; Garrard et al., 2005). While neuroimaging studies have also examined language functioning in AD, our understanding remains limited. The purpose of the present investigation was to further explore the link between language difficulties and regional brain functioning in AD.

Language skills can be evaluated in clinical settings by asking the patient to name objects presented by the examiner. A quantitative assessment of confrontation naming can be obtained using the Boston Naming Test (BNT; Kaplan et al., 1983). On this measure, a patient is presented line drawings of objects and asked to spontaneously generate the name. In AD, poor performance on the BNT may be associated with difficulties recognizing the object (i.e. impairments in semantic knowledge), and/or problems retrieving the name for a recognized object (Chertkow and Bub, 1990; Garrard et al., 2005; Rogers and Friedman, 2008).

Previous investigators have explored the relation between confrontation naming impairments and brain functioning in AD. FDG-PET and structural MRI studies have found that naming deficits are associated predominantly with abnormalities in temporal cortex (Teipel et al., 2006, Hirono et al., 2001), although frontal and parieto-occipital cortex may also be involved (Apostolova et al., 2008). Researchers have proposed that dysfunction of temporal regions in particular results in damage to the semantic network.

These studies investigated the neural correlates of spontaneous naming. On the BNT, when a patient is unable to name an object, a phonemic cue is provided. This cue is comprised of the initial sounds of the object’s name (e.g. the phonemic cue for “harmonica” is “har”), and is thought to assist in retrieving the name if it is indeed recognized. Thus, successful naming following phonemic cuing can illustrate the extent to which naming impairments are due to poor retrieval. To date, there have been no studies exploring the neural correlates of phonemic cuing benefit in AD.

AD patients also demonstrate deficits with verbal fluency (Henry et al., 2004), which refers to the ability to quickly generate information according to a given rule. These tests require the retrieval of verbal information, and as such their neural correlates may provide some insight into the areas of the brain involved in word retrieval during naming. Semantic fluency (e.g., stating as many animals as possible) deficits are often more pronounced than phonemic fluency (e.g., stating as many words as possible that begin with the letter “F) deficits, presumably because semantic knowledge is dependent on temporal regions that are impacted earlier in the disease (Henry et al., 2004). Imaging studies have been mixed, and have found that semantic fluency deficits are associated with compromise to multiple cortical regions, including some combination of frontal, temporal and parietal regions (Apostolova et al., 2008; Desgranges et al., 1998; Hirono et al., 2001; Teipel et al., 2006; Venneri et al., 2008; Welsh et al., 1994). Studies of phonemic fluency have also been variable, finding that deficits correlate with dysfunction of some combination of frontal, temporal, parietal, and occipital regions (Apostolova et al., 2008; Bracco et al., 2007; Venneri et al., 2008). Taken together, the neuroimaging findings are mixed, but suggest that both semantic and phonemic fluency deficits in AD result from alterations to a diffuse cortical network.

We used FDG-PET to examine the neural correlates of language functioning in AD. We were interested in determining the neural underpinnings of confrontation naming, and more importantly, which regions support successful name retrieval following phonemic cuing. We also sought to compare and contrast the extent to which these regions overlap with areas that support verbal fluency. We predicted that poorer confrontation naming would be associated with hypometabolism in primarily temporal regions, and that both inability to benefit from phonemic cuing and poorer verbal fluency would be associated with hypometabolism in frontal, temporal, and parietal regions.



We included 60 patients who met criteria for probable AD established by the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Disease and Related Disorder Association (McKhann et al., 1984). Subjects were recruited from the Greater Los Angeles Healthcare System Geropsychiatry Outpatient Program and from the UCLA Alzheimer’s Disease clinics.

Each patient underwent a clinical evaluation that included complete history, cognitive assessment, psychiatric assessment, neurological examination, blood assays, and structural neuroimaging with magnetic resonance imaging (MRI) or computed tomography (CT). Final diagnosis was confirmed by a board certified geriatric psychiatrist (DLS). Patients were excluded from the study if they had a history of psychotic disorder unrelated to dementia, history of head trauma resulting in a loss of consciousness, a psychoactive substance use disorder, or a systemic illness or other neurological condition accounting for cognitive impairment. Patients taking antidementia or other psychotropic medications were excluded, except for those taking a stable dose of cholinesterase inhibitor (19/60 taking donepezil, 2/60 taking galantamine) or antidepressant (12/60) medication.

The study was reviewed and approved by the local IRB, and consent to participate was documented according to IRB guidelines.

Neuropsychological testing

The language measures were drawn from a comprehensive neuropsychological battery. The Folstein Mini-mental State Examination (MMSE; Folstein et al., 1975) was administered to assess global cognitive impairment. Naming was assessed using the Boston Naming Test (Kaplan et al., 1983). We recorded the number of items the patient correctly named spontaneously and used this score in our analysis (BNT). If a subject clearly misperceived an item or did not know its name, a semantic cue was provided. If the patient still did not correctly name the item, a phonemic cue was given (the first 2–3 letters of the word). We recorded the number of additional items the patient was able to name with phonemic cuing (PC). Because we were interested in examining brain regions that correlated with successful use of PC, we limited this analysis to patients with impaired naming in order to remove ceiling effects. We converted each subject’s BNT score to age adjusted percentiles (MOANS norms; Ivnik et al., 1996) and those patients with BNT total <10%ile were included in the analysis.

To assess phonemic fluency, we summed the number of words the patient generated within one minute beginning with the letters F, A, and S (FAS). To assess semantic fluency, we tabulated the number of supermarket items the patient named in one minute (SUP); this was extracted from the Mattis Dementia Rating Scale (Mattis, 1988). We used SPSS v15 to obtain descriptive information and correlations between the measures.

FDG-PET image acquisition

Subjects completed PET imaging on the same day as the neuropsychological assessments. Patients were scanned using three different tomographs with similar imaging characteristics over the course of the study. Forty patients were scanned on a Siemens 953/31 tomographic scanner, eighteen on a GE Advance PET-CT, and two on a Philips Gemini TF PET-CT. Each scanner has an inplane resolution of approximately 5 mm at full-width half maximum and axial slice thickness of 2–4 mm.

[18F] fluorodeoxyglucose (FDG) was synthesized at the Veterans Affairs Greater Los Angeles Healthcare System PET Imaging facility in accordance with the technique of Hamacher et al. (1986). Subject received 5–10 mCi of FDG intraveneously and rested with their eyes open during a 40-minute uptake phase. Subjects were then placed in the scanner with the imaging plane parallel to the canthomeatal plane. Metabolic data were acquired for 40 minutes.

PET image analysis

PET data were analyzed using SPM2 (Wellcome Trust Centre for Neuroimaging) in Matlab 6.5 (MathWorks). Images were normalized to MNI space using trilinear interpolation and resampled to 2×2×2 mm voxels. They were next smoothed using a 6 FWHM smoothing kernel.

To assess correlations, we used the simple correlation procedure in SPM. Separate correlations were run between cortical metabolic activity and each of the four measures: BNT, PC, FAS, and SUP. Images were normalized to the global mean using proportional scaling and threshold masking was used to remove signal from structures outside of grey matter (set to .8). The image threshold for viewing results was set to p≤.001 uncorrected, with an extent threshold of 10 voxels. Results were considered significant at the voxel level at p≤.001, uncorrected. We also assessed results at a more liberal threshold, although still in keeping with studies investigating the neural correlates of cognition in AD (Bracco et al., 2007; Kalpouzos et al., 2005). To this end, we raised the image threshold to p<.005, and increased the extent threshold to 20 voxels. The goal of this analytic strategy was to raise the sensitivity to detect associations with specific aspects of language function. Anatomical regions were determined using Human Brain Anatomy in Computerized Images (Damasio, 2005) and approximate Brodmann’s regions were determined using MRIcro (Rorden and Brett, 2000).

Because neuropsychological studies have demonstrated that age and education influence performance on cognitive measures, we planned separate analyses covarying for these factors. Image thresholding and significance determination were the same as described above. We also examined correlations between resting metabolism and disease severity as assessed by the MMSE.


Descriptive characteristics

Eleven women and forty-nine men were included in the study. Table 1 includes demographic variables.

Table 1
Demographic and cognitive information.

Neuropsychological testing

Patients were on average in the mild to moderate stages of AD. Language ability was moderately impaired (see Table 1). When we examined BNT score, 44 of the 60 patients were found to have poor naming (age corrected score <10%ile). The PC analysis was limited to these 44 patients. All of the language measures correlated with disease severity as assessed by the MMSE (BNT: r=.543, p<.001; PC: r=.345, p<.05; FAS: r=.585, p<.001; SUP: r=.578, p<.001). Age and BNT were inversely correlated (r=−.303, p<.05). We observed direct correlations between education and BNT (r=.427, p<.01), FAS (r=.523, p<.001), and SUP (r=.496, p<.001).

Association between language scores and metabolic activity

Significant associations between language scores and regional metabolic activity are shown in Table 2 and Figure 1. Associations in left lateral cortex are displayed in Figure 1b. There was a direct correlation (p≤.001, extent=10) between metabolic activity and BNT in right temporal pole, right inferior temporal gyrus (ITG), and left ITG/fusiform gyrus. When we relaxed the threshold (p<.005, extent=20), we also observed a direct correlation between BNT score and left temporal pole activity.

Figure 1
a) Areas of significant positive correlations between cortical metabolism and each language measure, separately. b) Areas of significant positive correlations in the left cortex. Areas of the left IFG are circled in green, and areas of left temporal cortex ...
Table 2
Coordinates for correlations (p≤.001 uncorrected, extent=10 voxels) for each test separately. Additional associations at a more liberal threshold are shown in italics (p<.005 uncorrected, extent=20).

In patients with poor naming (n=44), PC correlated with higher metabolic activity in bilateral inferior frontal gyrus (IFG), pars triangularis, right superior frontal gyrus (SFG)/premotor cortex, left middle temporal gyrus (MTG), and right occipital lobe (p≤.001, extent=10). When we relaxed the threshold (p<.005, extent=20), we additionally observed a correlation with activity of left frontal pole, right postcentral gyrus, bilateral middle occipital gyrus, and left occipital pole.

We observed a direct correlation between FAS and metabolic activity in the left IFG, pars opercularis (p≤.001, extent=10). When we relaxed the threshold, we observed a direct correlation with metabolic activity of the IFG, pars orbitalis (p<.005, extent=20).

We observed a direct correlation between SUP and metabolic activity of right IFG, pars triangularis, left SFG, right supplementary motor area (SMA), right IFG, pars opercularis, left middle cingulate, and left angular gyrus. When we relaxed the threshold, we additionally observed a direct correlation with metabolism in left IFG, pars triangularis, right MTG, bilateral ITG, and right angular gyrus.

Figure 1b highlights correlations within the left IFG and temporal lobe.

When we covaried for age and education, the pattern of results remained unchanged, although the strength of some associations changed. The associations in regions reported above (see also Table 2) all remained significant, except that the association between FAS and metabolism in the IFG pars orbitalis weakened (p=.006).

We observed significant correlations between the MMSE and resting metabolism of the right ITG (z=3.64, extent=715, p<.001), left ITG (z=3.33, extent=72, p<.001), and right angular gyrus (z=3.23, extent=86, p=.001).


This study aimed to understand the neural correlates of naming dysfunction in AD by exploring the association between several indices of language ability and cortical metabolism. We predicted that spontaneous naming would be associated with the integrity of the temporal lobes, and that both naming following phonemic cuing and verbal fluency performance would be associated with a more diffuse network of frontal, temporal and parietal regions. Our predictions were largely supported. We found that poorer ability to spontaneously name an object was associated with hypometabolism of bilateral temporal lobes. In contrast, when a phonemic cue was provided, successful naming under this condition was associated with greater metabolic activity in areas including bilateral inferior frontal gyrus (IFG) and left temporal cortex. Consistent with these findings, we found that poorer semantic fluency, which requires both semantic knowledge and verbal retrieval, was associated with hypometabolism in both IFG and temporal regions, whereas poorer phonemic fluency, which taxes verbal retrieval but is less reliant on semantic knowledge, was associated with hypometabolism in only left IFG. Our findings support the argument that naming deficits in AD reflect compromise to both a temporally-based semantic knowledge network and frontally-based retrieval processes, and in particular highlight the importance of frontal/executive retrieval processing during naming in AD.

During the BNT, successful naming requires the ability to recognize the object (visual processing), match the object to a previously encoded representation of that object (semantic knowledge), and match the recognized object to its word-form (phonological representation; Humphreys et al., 1999). If the patient is unable to provide the name, a phonemic cue (PC) is provided. The PC provides an external framework for word retrieval, suggesting that a patient who is able to use PC but unable to spontaneously name an object has difficulty with internally-guided retrieval processing (Jefferies et al., 2008). In this case, semantic knowledge about the object is intact, but the ability to retrieve the correct phonological representation is impaired. Borrowing from theories of semantic memory (Damasio et al., 2004; Badre and Wagner, 2007), spontaneous naming is a bottom-up “automatic” retrieval process that requires minimal attention or executive processing. If automatic retrieval is insufficient to elicit the name, top-down “controlled” retrieval processing is employed.

We observed that successful use of PC correlated with greater metabolic activity in bilateral frontal regions. We also found that all three measures that taxed verbal retrieval following a cue (PC, FAS, and SUP) showed a correlation between poorer performance and hypometabolism of the left IFG (see Figure 1b), suggesting that the left IFG in particular is important in cued verbal retrieval. A recent study contrasting benefit from phonemic cuing in patients with semantic dementia (prominent bilateral temporal damage) versus semantic amnesia (prefrontal and/or temporoparietal damage) demonstrated that the latter group was better able to benefit from cuing (Jefferies et al., 2008). The authors proposed that for this group, the phonemic cue reduced the executive demands of the task, which included defining the search parameters and inhibiting incorrect but semantically related responses. As such, naming deficits in AD patients with frontal dysfunction may stem in part from compromise to top-down controlled retrieval processes.

We further observed that different regions within the left IFG were associated with the different retrieval measures. All three tests showed a correlation with anterior LIFG: FAS correlated with LIFG, pars orbitalis (BA47) and PC and SUP correlated with LIFG pars triangularis (BA45). Only FAS showed a correlation with the posterior LIFG (pars opercularis, BA44). Consistent with this observation, anterior LIFG is implicated in processing semantic information, while the posterior LIFG, which is situated closer to the motor cortex and language production centers, is thought to be involved in phonological processing (for reviews see Bookheimer, 2002; Costafreda et al., 2006). It has also been proposed that regions within the anterior LIFG support different aspects of retrieval: BA47 is hypothesized to guide retrieval search, while BA45 acts on those retrieved representations to select the most appropriate response (Badre and Wagner, 2007). Consistent with the ideas proposed above, this suggests that frontal dysfunction leads to naming impairment by disrupting the executive aspects of controlled retrieval in AD.

We also found that successful use of PC correlated with metabolic activity in the left temporal cortex, specifically the posterior aspect of the middle temporal gyrus (MTG). This finding is consistent with the interpretation that controlled retrieval acts on the contents of temporally-stored semantic information. In further support of this, we found that performance on all three tests that required semantic processing (BNT, PC, and SUP) was associated with metabolism in the left temporal lobe (see Figure 1b).

Researchers have proposed that different regions within the left temporal lobe support different cognitive functions necessary for naming. In our study, the BNT correlated with metabolic activity in left fusiform and inferior temporal gyrus (ITG). The fusiform (BA37) is situated along the “what” visual pathway and is involved in object recognition, or perhaps more specifically linking a visual percept with meaningful semantic information (Murtha et al., 1999). The basal temporal area (which includes the fusiform and ITG: BA20) has been implicated in matching visually-based semantic information to a phonological representation (Miozzo et al., 1994; Usui et al., 2003), or put another way, this region may support bottom-up automatic retrieval during naming. We observed a correlation with the temporal basal area metabolic rate in both the BNT and SUP analyses. In the PC analysis, we observed a correlation between successful naming following cuing and greater cerebral metabolism in the posterior MTG. This region overlaps with the results from previous studies of spontaneous naming in AD (Hirono et al., 2001; Teipel et al., 2006). In our study, it is unclear if the variability between the regions observed in the PC (MTG, BA21/37) and SUP/BNT (ITG, BA20) analyses reflects an anatomical and/or cognitive distinction between these tests, or instead may just be areas within a single region involved in semantic processing. Taken together, our findings are consistent with the proposal that in AD, damage to the left temporal lobe impairs naming ability by disrupting the semantic knowledge network, including object recognition, linking semantic information to a visual percept, and automatic retrieval involved in linking a phonological representation to a recognized object.

Although we have focused on findings within the left cortex, we did observe associations between the language measures and metabolism in the right cortex. Poorer spontaneous naming (BNT) was associated with hypometabolism of the right temporal lobe, including the ITG and anterior pole. This is consistent with previous work in AD (Apostolova et al., 2008). This finding may reflect a functional role for this region in naming: in a heterogeneous group of dementia patients, poorer naming of living things was associated with reduced gray matter density of right anterior temporal pole (Brambati et al., 2006). Alternatively, because AD is primarily a bilateral degenerative disease, it may be an artifact of the association between BNT and left temporal regions. Likewise, PC was associated with metabolism in the right IFG, which may reflect task related processing (Kelley et al., 1998) or an artifact of the associations with the left frontal cortex.

We did not control for global cognitive impairment by using MMSE as a covariate in our analyses as the MMSE is heavily weighted towards language ability and shows high correlations with each of the language indices (r>.34 in all cases). Thus, we believed that controlling for the MMSE would unnecessarily remove variance associated with the variables of interest and would significantly underestimate the correlations. In support of this approach, we did observe relatively circumscribed regions for the majority of the analyses, suggesting that we were not measuring a global cognitive factor, but instead identifying areas specifically involved in the cognitive function under investigation (Desgranges et al., 1998). When we examined the correlation between MMSE and cerebral metabolism, we did observe an association with left ITG, suggesting that this region is sensitive to both global cognition and semantic processing, and is consistent with the observation that naming deficits are more severe in later stages of AD.

Several comments should be noted about the investigation. To increase our sample size, we included data from subjects scanned on 3 different PET tomographs. The normalization procedure used by SPM results in ratio-scaling for each subject. Ratio values have been shown to vary by less than 5% across scanners (Grady, 1991) and were not significantly different across 3 different tomographs in a study of AD patients (Herholz, et al., 1993). Including data across multiple scanners has likely increased the noise in our analyses, and we may not be detecting all of the associations between metabolism and language functioning. We used uncorrected p-values in conjunction with a spatial constraint to assess results; although this is common when examining correlations between metabolism and cognition in disease populations (Bracco et al., 2007; Eustache et al., 2004; Tiepel et al., 2006), this analysis is vulnerable to false positives. Moreover, the four independent analyses were not adjusted for multiple comparisons. We selected to present the results at two thresholds in order to identify the strongest correlations (p≤.001) while also raising the sensitivity of identifying weaker, yet spatially larger correlations (p<.005, extent=20 voxels). We also note that a subset of patients were on a stable of dose of cholinesterase inhibitor or antidepressant medication. While there is no clear evidence that these medications would modulate the observed correlations between metabolic rate and cognitive ability, this possibility remains. We also emphasize that we assessed naming following the presentation of visually-based stimuli, and our findings may not be representative of naming processes involved when stimuli are presented in other sensory modalities. Lastly, the present study did not address all of the language deficits that may be seen in AD.

It is important to note that we observed associations between resting cerebral metabolism and cognitive ability in the context of a disease that reflects diffuse brain damage. We have interpreted our findings as support that dysfunction of relatively precise cortical regions is responsible for specific cognitive deficits in AD. However, AD is not a focal disease, and this approach may be misleading. It is likely that each cognitive process under investigation requires the integrity of a much larger network of brain regions than we have observed. Due to the diffuse nature of the disease, we are only able to detect the strongest associations between cerebral metabolism and cognition. The correlations we report do support the hypothesis that specific brain areas are involved in the cognitive deficits observed in patients with AD, but there is likely disruption to a larger neural circuitry that underlies these deficits (Teipel et al., 2006).

In summary, naming deficits in AD appear to stem from damage to temporal regions involved in processing information within the semantic knowledge network, as well as frontal regions that are involved in the controlled retrieval of information from that network. These findings further our understanding of language impairment in AD, and highlight that for an individual patient, the type of language difficulty experienced may reflect compromise to different neuroanatomical regions.


Supported by the Dept. of Veterans Affairs (Merit Review, and Special Fellowship in Advanced Geriatrics Program) and the National Institute of Mental Health (#MH56301). We thank Amy Walston, M.D. and Natalya Bussel, M.D. for assistance with clinical care, and Natalie Achamallah for assistance with manuscript preparation. Dr. Sultzer has received research grant support from Forest Research Institute and Pfizer. Potential conflicts to study participants were disclosed. None of the other authors report conflicts of interest. This work was presented in poster form at the Cognitive Aging conference in Atlanta, GA in 2008. The supporting sources had no involvement in the study design, collection, analysis, interpretation, writing of this report, or the decision to submit this manuscript for publication.


  • Apostolova LG, Lu P, Rogers S, et al. 3D mapping of language networks in clinical and pre-clinical Alzheimer’s disease. Brain Lang. 2008;104:33–41. [PMC free article] [PubMed]
  • Badre D, Wagner AD. Left ventrolateral prefrontal cortex and the cognitive control of memory. Neuropsychologia. 2007;45:2883–901. [PubMed]
  • Bookheimer S. Functional MRI of language: new approaches to understanding the cortical organization of semantic processing. Annu Rev Neurosci. 2002;25:151–88. [PubMed]
  • Bracco L, Bessi V, Piccini C, et al. Metabolic correlates of executive dysfunction. Different patterns in mild and very mild Alzheimer’s disease. J Neurol. 2007;254:1052–65. [PubMed]
  • Brambati SM, Myers D, Wilson A, et al. The anatomy of category-specific object naming in neurodegenerative diseases. J Cogn Neurosci. 2006;18:1644–53. [PubMed]
  • Chertkow H, Bub D. Semantic memory loss in dementia of Alzheimer’s type. What do various measures measure? Brain. 1990;113 (Pt 2):397–417. [PubMed]
  • Costafreda SG, Fu CH, Lee L, et al. A systematic review and quantitative appraisal of fMRI studies of verbal fluency: role of the left inferior frontal gyrus. Hum Brain Mapp. 2006;27:799–810. [PubMed]
  • Damasio H. Human Brain Anatomy in Computerized Images. Oxford University Press; New York: 2005.
  • Damasio H, Tranel D, Grabowski T, Adolphs R, Damasio A. Neural systems behind word and concept retrieval. Cognition. 2004;92:179–229. [PubMed]
  • Desgranges B, Baron JC, de la Sayette V, et al. The neural substrates of memory systems impairment in Alzheimer’s disease. A PET study of resting brain glucose utilization. Brain. 1998;121 (Pt 4):611–31. [PubMed]
  • Desgranges B, Baron JC, Lalevee C, et al. The neural substrates of episodic memory impairment in Alzheimer’s disease as revealed by FDG-PET: relationship to degree of deterioration. Brain. 2002;125:1116–24. [PubMed]
  • Eustache F, Piolino P, Giffard B, et al. ‘In the course of time’: a PET study of the cerebral substrates of autobiographical amnesia in Alzheimer’s disease. Brain. 2004;127:1549–60. [PubMed]
  • Folstein M, Folstein S, McHugh P. “Mini-Mental State”: A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189–198. [PubMed]
  • Forbes-McKay KE, Venneri A. Detecting subtle spontaneous language decline in early Alzheimer’s disease with a picture description task. Neurol Sci. 2005;26:243–54. [PubMed]
  • Garrard P, Lambon Ralph MA, Patterson K, Pratt KH, Hodges JR. Semantic feature knowledge and picture naming in dementia of Alzheimer’s type: a new approach. Brain Lang. 2005;93:79–94. [PubMed]
  • Garrard P, Maloney LM, Hodges JR, Patterson K. The effects of very early Alzheimer’s disease on the characteristics of writing by a renowned author. Brain. 2005;128:250–60. [PubMed]
  • Grady CL. Quantitative comparison of measurements of cerebral glucose metabolic rate made with two positron cameras. J Cereb Blood Flow Metab. 1991;11:A57–A63. [PubMed]
  • Grossman M, McMillan C, Moore P, et al. What’s in a name: voxel-based morphometric analyses of MRI and naming difficulty in Alzheimer’s disease, frontotemporal dementia and corticobasal degeneration. Brain. 2004;127:628–49. [PubMed]
  • Hamacher K, Coenen HH, Stocklin G. Efficient stereospecific synthesis of no-carrier-added 2-18F-fluoro-2-deoxy-D-glucose using aminopolyether supported nucleophilic substitution. J Nucl Med. 1986;27:235–238. [PubMed]
  • Herholz K, Perani D, Salmon E, et al. Comparability of FDG PET studies in probable Alzheimer’s disease. J Nucl Med. 1993;34:1460–6. [PubMed]
  • Henry JD, Crawford JR, Phillips LH. Verbal fluency performance in dementia of the Alzheimer’s type: a meta-analysis. Neuropsychologia. 2004;42:1212–22. [PubMed]
  • Hirono N, Mori E, Ishii K, et al. Neuronal substrates for semantic memory: a positron emission tomography study in Alzheimer’s disease. Dement Geriatr Cogn Disord. 2001;12:15–21. [PubMed]
  • Humphreys GW, Price CJ, Riddoch MJ. From objects to names: a cognitive neuroscience approach. Psychol Res. 1999;62:118–30. [PubMed]
  • Ivnik RJ, Malee JF, Smith GE, Tangalos EG, Petersen RC. Neuropsychological Tests’ Norms Above Age 55: COWAT, BNT, MAE Token, WRAT-R Reading, AMNART, STROOP, TMT, and JLO. The Clinical Neuropsychologist. 1996;10:262–278.
  • Jefferies E, Patterson K, Ralph MA. Deficits of knowledge versus executive control in semantic cognition: insights from cued naming. Neuropsychologia. 2008;46:649–58. [PMC free article] [PubMed]
  • Kalpouzos G, Eustache F, de la Sayette V, et al. Working memory and FDG-PET dissociate early and late onset Alzheimer disease patients. J Neurol. 2005;252:548–58. [PubMed]
  • Kaplan E, Goodglass H, Weintraub S. The Boston Naming Test. 2. Lea & Febiger; Philadelphia: 1983.
  • Kelley WM, Miezin FM, McDermott KB, et al. Hemispheric specialization in human dorsal frontal cortex and medial temporal lobe for verbal and nonverbal memory encoding. Neuron. 1998;20:927–36. [PubMed]
  • Lezak MD, Howieson DB, Loring DW. Neuropsychological Assessment. Oxford; New York: 2004.
  • Mattis S. Dementia Rating Scale: Professional Manual. Psychological Assessment Resource, Inc; Odessa, FL: 1988.
  • McKhann G, Drachman D, Folstein M, et al. Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspices of the Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology. 1984;34:939–944. [PubMed]
  • Miozzo A, Soardi M, Cappa SF. Pure anomia with spared action naming due to a left temporal lesion. Neuropsychologia. 1994;32:1101–9. [PubMed]
  • Murtha S, Chertkow H, Beauregard M, Evans A. The neural substrate of picture naming. J Cogn Neurosci. 1999;11:399–423. [PubMed]
  • Rogers SL, Friedman RB. The underlying mechanisms of semantic memory loss in Alzheimer’s disease and semantic dementia. Neuropsychologia. 2008;46:12–21. [PMC free article] [PubMed]
  • Rorden C, Brett M. Stereotaxic display of brain lesions. Behav Neurol. 2000;12:191–200. [PubMed]
  • Teipel SJ, Willoch F, Ishii K, et al. Resting state glucose utilization and the CERAD cognitive battery in patients with Alzheimer’s disease. Neurobiol Aging. 2006;27:681–90. [PubMed]
  • Usui K, Ikeda A, Takayama M, et al. Conversion of semantic information into phonological representation: a function in left posterior basal temporal area. Brain. 2003;126:632–41. [PubMed]
  • Venneri A, McGeown WJ, Hietanen HM, et al. The anatomical bases of semantic retrieval deficits in early Alzheimer’s disease. Neuropsychologia. 2008;46:497–510. [PubMed]
  • Welsh KA, Butters N, Mohs RC, et al. The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part V. A normative study of the neuropsychological battery. Neurology. 1994;44:609–614. [PubMed]