Abstract

The purpose of the present study was to explore the brain regions involved in human episodic memory by correlating unilateral memory performance estimated by the intracarotid amobarbital test (IAT) and interictal cerebral metabolism measured by [18F]fluorodeoxyglucose positron emission tomography ([18F]FDG-PET). Using this method, regional alterations of cerebral metabolism associated with epilepsy pathophysiology are used to predict hemisphere-specific episodic memory function, hence, investigate the differential distribution of memory in each hemisphere. Sixty-two patients with unilateral temporal lobe epilepsy (35 left and 27 right) were studied using [18F]FDG-PET with complementary voxel-based statistical parametric mapping (SPM) and region-of-interest (ROI) methods of analysis. Positive regression was analyzed in SPM with a series of different thresholds (p = .001, .01 or .05) with a correction to 100 voxels. IAT memory performance in which left hemisphere was tested by right-sided injection of amobarbital correlated with [18F]FDG uptake in left lateral and medial temporal regions, and in the left ventrolateral frontal cortex. Right IAT memory performance correlated with [18F]FDG uptake in the right inferior parietal lobule, right dorsolateral frontal cortex, right precentral gyrus, and caudal portion of the right anterior cingulate cortex. ROI analysis corroborated these results. Analyses carried out separately in patients with left (n = 50) and nonleft (n = 12) dominance for language showed that in the nonleft dominant group, right IAT scores correlated with right fronto-temporal regions, whereas left total memory scores correlated with left lateral and medial temporal regions. The findings indicate that (i) episodic memory is subserved by more widespread cortical regions beyond the core mesiotemporal lobe memory structures; (ii) there are different networks functional in the two hemispheres; and (iii) areas involved in memory may be different between patients with left and nonleft dominance for language, particularly in the right hemisphere.

INTRODUCTION

The areas responsible for memory function in the human brain are still debated. Following the report of Patient H. M., who developed severe anterograde amnesia after bilateral medial temporal lobectomy (Scoville & Milner, 1957), the role of medial temporal structures in learning and memory has been studied extensively in animal models and in patients with temporal lobe epilepsy (TLE). Clinical and experimental studies have shown that medial temporal structures play a major role in episodic memory function (Eichenbaum, 2000; Gabrieli, Brewer, Desmond, & Glover, 1997; Squire, 1992). Modern imaging studies have provided further evidence that the extrahippocampal temporal cortex, including peri- and entorhinal cortices, supports episodic memory function in combination with the hippocampus in normal subjects and patients with TLE (Weintrob, Saling, Berkovic, Berlangieri, & Reutens, 2002; Ploner et al., 1999).

The involvement of the extrahippocampal cortex is suggested by the fact that cognitive deficits following extensive unilateral anterior temporal lobectomy are more pronounced than those seen after resections restricted to unilateral medial temporal structures (Lacruz et al., 2004; Pauli, Pickel, Schulemann, Buchfelder, & Stefan, 1999; Goldstein & Polkey, 1993; Wieser & Yasargil, 1982). Neuroimaging studies in organic amnesia of diverse etiology have shown reduced metabolism throughout limbic–diencephalic circuits (e.g., Fazio et al., 1992) and the retrosplenial cortex (Reed et al., 1999, 2005; Aupee et al., 2001; Signorini et al., 1999). In patients with anoxic amnesia, known to damage preferentially hippocampal CA1 fields, glucose metabolism was decreased bilaterally in the thalamus and retrosplenial cortex in the absence of hippocampal hypometabolism (Reed et al., 1999). This observation was interpreted as distal effects of hippocampal damage within the network supporting episodic memory function. Extratemporal cortical involvement in episodic memory is also found in relation to circuitary involving the frontal cortex. In nonhuman primates, systematic lesion techniques suggest that object memory is served by a circuit involving the orbito-frontal cortex, the rhinal cortex, and the medial dorsal thalamus (Meunier, Bachevalier, & Mishkin, 1997). Although in humans, localized frontal lesions do not produce severe amnesia, nevertheless, memory impairment is found across a range of tasks (Turner, Cipolotti, Yousry, & Shallice, 2007; Wheeler, Stuss, & Tulving, 1995; Incisa della Rocchetta & Milner, 1993; Janowsky, Shimamura, Kritchevsky, & Squire, 1989; Schacter, 1987) and functional neuroimaging studies suggest that the frontal lobes contribute to episodic memory (Fletcher & Henson, 2001; Frey & Petrides, 2000; Buckner, Kelley, & Petersen, 1999; Petrides, Alivisatos, Evans, & Meyer, 1993).

A key issue relating to episodic memory is functional hemispheric asymmetry in humans. There is evidence in relation to the temporal lobe involvement in memory for domain specificity, with verbal memory more likely to be impaired by lesions in the left temporal lobe and visual or spatial memory impairments associated with right temporal lobe lesions (Morris & Mayes, 2004; Morris, Nunn, Abrahams, Feigenbaum, & Recce, 1999; Smith, 1989). This pattern tends to be found in relation to the frontal cortex, where results of functional neuroimaging studies show material-specific effects, with verbal memory activating the left hemisphere (e.g., Kelley et al., 1998; Wagner et al., 1998) and “nonverbal” memory activating the right frontal cortex (McDermott, Buckner, Petersen, Kelley, & Sanders, 1999; Kelley et al., 1998). Nevertheless, other distinctions have been made in relation to hemispheric regions, principally the notion that the left frontal cortex is activated during encoding and the right during retrieval, the so-called hemispheric encoding/retrieval asymmetry (HERA) model (Owen, 2003; Tulving, Kapur, Craik, Moscovitch, & Houle, 1994) although this model may not hold up in all circumstances and may be material specific (Habib, Nyberg, & Tulving, 2003; Fletcher & Henson, 2001).

Because both cerebral hemispheres can process information in parallel or independently, determination of hemisphere laterality for specific memory tasks is a major difficulty for in vivo localization of episodic memory. Additionally, findings from functional neuroimaging, such as functional magnetic resonance imaging (fMRI), cannot always be used conclusively to establish what brain regions are critical for function. In this context, it is arguable that lateralization effects can be explored fruitfully by selective and temporary deactivation of each hemisphere. One technique for doing this is transcranial magnetic stimulation (TMS), which has been used to implicate specific cortical regions in episodic memory (e.g., Rossi et al., 2001). To obtain more widespread deactivation of one hemisphere, currently, the intracarotid sodium amobarbital test (IAT or Wada test) can be used (Milner, Branch, & Rasmussen, 1962; Wada & Rasmussen, 1960). The purpose of IAT is to estimate and avoid potential neuropsychological deficits associated with temporal lobe surgery. The effects of unilateral injections of sodium amobarbital on memory can be evaluated by presenting memory material for encoding just after one hemisphere is anesthetized. Encoding of memory material permits estimation of unilateral memory function, for example, allowing the study of the relationship between memory and ipsilateral in vivo brain measurements that can be obtained with modern neuroimaging techniques (Akanuma, Koutroumanidis, Adachi, Alarcon, & Binnie, 2003; Ferrier et al., 2000). Kelley et al. (2002) have suggested that the participation of the frontal cortex is related to memory performance during IAT. Encoding tasks produce dorsal frontal activation on fMRI with various degrees of laterality according to the material used (Kelley et al., 1998). The laterality of frontal activation was highly concordant with the laterality of IAT memory performance. These results provide further support that, in addition to temporal structures, frontal regions also contribute to memory formation. In the study by Kelley et al., findings from neuroimaging and the IAT were obtained from different subject populations: Neuroimaging findings were obtained from normal subjects, whereas IAT memory performance was measured in patients with intractable epilepsy.

In the current study, IAT hemispheric-specific memory data were combined with [18F]fluorodeoxyglucose positron emission tomography ([18F]FDG-PET) to measure regional integrity of brain tissue in patients undergoing surgery for treatment of epilepsy. In patients with suspected TLE, [18F]FDG-PET and the IAT are carried out as part of presurgical assessment in several centres. By anesthetizing one hemisphere using the IAT and estimating unilateral memory function in the other hemisphere, it was then possible to cross-refer this information. Consequently, in brain areas involved in processing episodic memory, regional metabolic rates shown by [18F]FDG-PET would be expected to correlate with unilateral memory scores estimated by the IAT. We hypothesize that subjects with less memory function would be expected to show lower metabolic rate in the areas involved in memory processing. In contrast, in areas exclusively involved in sensory processing, no such correlation between metabolic rate and unilateral memory scores would be expected. Several pathological and neuroimaging studies have revealed a relationship between unilateral memory performance assessed by IAT and several measures of hippocampal integrity, including hippocampal damage (Sass et al., 1990; Rausch, Babb, Engel, & Crandall, 1989), hippocampal atrophy shown on magnetic resonance imaging (MRI) (Baxendale et al., 1997; Davies, Hermann, & Foley, 1996; Loring et al., 1993), and hippocampal neuronal density as estimated by magnetic resonance spectroscopy (Ferrier et al., 2000). Studies comparing IAT with [18F]FDG-PET have revealed a high concordance between unilateral memory and temporal hypometabolism (Salanova, Markand, & Worth, 2001; Salanova et al., 1992). Using a semiquantitative approach based on regions of interest (ROIs), Hong, Roh, Kim, and Seo (2000) found a correlation between IAT memory and [18F]FDG-PET metabolism for temporal regions. However, no correlation was studied between IAT memory and [18F]FDG-PET metabolism in extratemporal areas.

To summarize, the purpose of the present study is to identify the brain regions involved in human episodic memory by correlating unilateral memory performance estimated by the IAT with interictal regional cerebral metabolism measured by [18F]FDG-PET. This was studied in patients with medically intractable unilateral epilepsy, as IAT can only be carried out in this population for ethical reasons. The novelty of this approach is that it enables variations in cerebral metabolism within each hemisphere to be associated with hemisphere-specific memory performance. Hence, it provides a way of exploring how functional distribution of episodic memory may differ between the two hemispheres of the brain. We explored this issue in relation to the regions involved in material-specific memory (words, pictures, faces). For processing PET images, we have used two complementary methods: (a) whole-brain voxel-based mapping approach using statistical parametric mapping (SPM) as an objective method for analysis of images of glucose metabolism; and (b) ROI analysis to account for any potential spatial normalization errors. For each voxel or ROI, intersubject correlations were carried out between metabolism in the voxel or region, and IAT scores for the ipsilateral hemisphere.

METHODS

Subjects

The study participants were selected from the cohort of consecutive patients with medically intractable focal seizures compatible with temporal lobe origin (Wieser, Engel, Williamson, Babb, & Goldman, 1993), assessed in the King's College Hospital epilepsy surgery program between 1993 and 1998. The subjects gave written informed consent to the investigations used in this program. This study was approved by the relevant ethics committees. The inclusion criteria were: (i) unilateral temporal seizure onset documented by scalp or intracranial video-EEG telemetry; (ii) patients had IAT; and (iii) patients had an [18F]FDG-PET scan. The exclusion criteria were: (i) past history of significant head injury or any neurological condition other than epilepsy; (ii) any acute or chronic medical illness at the time of the [18F]FDG-PET study; (iii) age younger than 16 years; (iv) full-scale intelligence quotient (FSIQ) less than 70; (v) intra-axial space-occupying lesions on MRI; (vi) cross-filling of the vascular tree of the hemisphere contralateral to the injection on the angiogram prior to the IAT; and (vii) peri-ictal [18F]FDG-PET scans (Barrington et al., 1998). As a result, 62 patients (40 men and 22 women) were enrolled in the study.

MRI was performed in all patients using a prespecified “temporal lobe” protocol described elsewhere (Alarcon et al., 2006; Koutroumanidis et al., 1998). When no abnormality could be ascertained on visual analysis, volume differences between right and left hippocampi were calculated. All patients had at least one prolonged awake and sleep EEG and scalp telemetry. If the latter was inconclusive, ictal recordings with intracranial electrodes were carried out, usually during the same admission and always after the [18F]FDG-PET scan. If seizures were not recorded by the third or the fourth day of telemetry, antiepileptic medication would be reduced by tapering a single drug at a time, usually either carbamazepine or phenytoin. No patient was withdrawn from phenobarbital, primidone, or benzodiazepines. The interictal EEG and the electroclinical features of all seizures were reviewed, and patients were selected only if they had habitual focal seizures, as suggested by the available history and confirmed by their relatives who were shown the videotapes at the time of the telemetry. All patients had full neuropsychological assessment, the methodology of which has been described elsewhere (Akanuma, Alarcon, et al., 2003).

Intracarotid Amobarbital Test

Our standard IAT protocol has been detailed elsewhere (Ferrier et al., 2000; Morton, Polkey, Cox, & Morris, 1996). Briefly, sodium amobarbital (75 to 125 mg) was injected unilaterally at a rate of 25 mg per 5 sec in the internal carotid artery until hemiparesis of the contralateral arm appeared. Immediately after appearance of the contralateral paresis, patients were shown 18 consecutive items (six words, six pictures of common objects and six male faces, mixed in a pseudorandom order) for approximately 5 sec each and were asked to remember these items. Language dominance was assessed clinically on the basis of observation of signs of dysphasia and inability to read the words or name the line drawings presented. In this article, patients with left-sided language dominance were defined as “left dominant for language” and those with right-sided or mixed language dominance as “nonleft dominant for language.” Memory was tested 10 min after neurological deficits and scalp EEG changes returned to baseline. Eighteen target items were presented individually, mixed in pseudorandom order with the same number of equivalent distracter items using a “yes–no” recognition memory paradigm.

Raw (nonstandardized) subscores for each material category (words, pictures, or faces) were calculated for each hemisphere by subtracting false from correct recognitions. Raw total IAT scores were calculated by adding the three material-specific memory raw subscores for each hemisphere. Throughout this article, right and left IAT scores refer to the side of the tested hemisphere (i.e., contralateral to the injection side). In order to adjust unilateral IAT scores to the overall patient performance to reduce individual variability unrelated to unilateral performance, the total IAT memory score and each material-specific subscore (the subscore for words, pictures, or faces) were standardized using the formulae:
formula
formula
formula
formula

Analyses with standardized scores were not carried out where the sum of the left and right raw scores (or material-specific subscores) was zero. On this basis, 1 patient (1.6%) was excluded in the analysis of standardized total scores, 3 (4.8%) in the analysis of standardized subscores for words, and 10 (16%) in the analysis of standardized subscores for faces. No one was excluded in the analysis of standardized subscores for pictures. Note that the standardized total score is not necessarily the addition of all three standardized subscores for words, pictures, and faces. For instance, a standardized total score can be calculated in the absence of one or more standardized subscores. Because standardized IAT scores reflect the degree of memory lateralization rather than the absolute memory performance of one hemisphere, we have also carried out SPM analysis using raw (nonstandardized) IAT scores in the whole population.

[18F]FDG-PET

Our standard imaging protocol has been described elsewhere (Koutroumanidis et al., 2000; Barrington et al., 1998). In brief, all PET scans were performed using an ECAT 951R whole-body scanner (Siemens CTI, Knoxville, TN, USA) with a 10.8-cm axial field of view. Following an uptake period of 30 min after intravenous injection of 250 MBq [18F]FDG, images were acquired from the skull base to the vertex over a 30-min period. The field of view was centered upon the cerebral areas of interest comprising frontal and temporal lobes and the diencephalon. In most participants, this resulted in the loss of the superior parietal and inferior cerebellar cortex. The participant's head was placed in the scanner with a supportive headrest to minimize head movement, aligned axially and to the orbitomeatal line (approximating to the AC–PC line), which was checked at intervals during image acquisition and corrected if drift was identified. Each reconstructed image was displayed in a matrix of 128 × 128 × 31 voxel format; image resolution was 8.5 × 8.5 × 5.5 mm full width at half maximum. Interictal [18F]FDG-PET studies were carried out with the patients on treatment with antiepileptic drugs. For clinical purposes, scans were interpreted visually by two nuclear medicine physicians blind to patient identity, with any differences resolved by consensus.

Image Analysis

Two complementary approaches were adopted for analysis of the [18F]FDG-PET images: SPM and ROI analysis. Due to the asymmetrical nature and the high degree of hypometabolism that frequently characterize [18F]FDG-PET images in patients with TLE, spatial normalization routines (such as SPM) may be misleading and we have cross-validated SPM findings with ROI analysis (Reed et al., 1999, 2003; Desgranges et al., 1998; Kennedy et al., 1995).

Statistical Parametric Mapping Analysis

Spatial preprocessing and statistical analysis were performed using SPM2 software (Wellcome Department of Imaging Neuroscience, London, UK) implemented in Matlab 6.0 (Mathworks, Sherborn, MA, USA). All reconstructed [18F]FDG-PET images were spatially normalized into Montreal Neurological Institute standard templates by affine transformation with 16 parameters and nonlinear transformations. The normalized images were smoothed using an isotropic Gaussian kernel with 16-mm full width at half maximum (Swartz, Thomas, Simpkins, Kovalik, & Mandelkern, 1999). Resliced voxel size was 2 × 2 × 2 mm. Proportional scaling was used to adjust for differences in global activity between scans. The voxels with values >40% of the mean for the whole brain were selected for the statistical analysis to minimize “edge effects” without excluding hypometabolic areas (Desgranges et al., 1998).

Correlations between ipsilateral memory and cerebral metabolism estimated for each selected voxel were made using voxel-by-voxel ANCOVA with the appropriate memory measure as the covariate of interest using SPM2. This generated a group-specific adjusted mean value and associated adjusted error variance for each voxel, forming an SPM t map (SPM{t}). Based on our a priori hypotheses, we assessed positive regressions and results were thresholded to a series of different levels of the t statistic (p = .001, .01, or .05) with a correction at the cluster level for volume to 100 voxels (extent threshold of 800 mm2). An initial analysis was carried out with the highest threshold (p = .001). Where suprathreshold areas containing less than 100 voxels were found, subsequent analyses were performed at lower thresholds (p = .01 and p = .05) to explore the extent of areas with weaker correlation and confirm whether certain areas (e.g., hippocampus) were not involved.

ROI Analysis

Our ROI method has been described in detail elsewhere (Koutroumanidis et al., 2000; Barrington et al., 1998). Reconstructed images were aligned to a plane parallel to the long temporal axis using an interactive alignment program, resliced, and adjacent planes summed to give 15 planes. The resulting images were sampled using 7 mm diameter circular ROI that were placed with reference to the Talairach and Tournoux (1988) stereotaxic atlas and adjusted with visual inspection according to individual [18F]FDG-PET anatomy. One or more ROIs were located within the following 10 structures: dorsolateral frontal cortex, inferolateral aspect of the precentral and postcentral gyri, superior temporal gyrus, middle temporal gyrus, inferior temporal gyrus, infratemporal cortex, amygdala and hippocampus, inferior parietal lobule, thalamus, as well as caudate and putamen. Counts within these regions were averaged to provide anatomically paired regions. Activity measurements were expressed as a ratio to averaged global cerebral ROI values for each individual.

In order to establish as to whether our findings are specific to the memory domain, we have also carried out correlations between [18F]FDG uptake in the ROIs and other cognitive domains (FSIQ, verbal and nonverbal intelligence quotient [VIQ and PIQ]).

Statistical Analysis

Data analysis was carried out using a Statistical Package for Social Sciences 11.0 (SPSS, Chicago, IL). Nominal clinical characteristics were analyzed by Yates' corrected Chi-square test. Numerical data were compared using two-way or three-way ANOVA with the laterality of language dominance, the side of epileptogenic focus, and/or the side of the IAT as independent variables. Correlations between each memory score and ROI values in each region were analyzed using Pearson product–moment correlation test. Differences were considered statistically significant if p < .05. Statistical methods used for SPM were discussed earlier.

RESULTS

Patient Characteristics

Sixty-two patients (40 men, 22 women) were included in the study. The median age at the time of recording PET and IAT was 28.5 years (range = 16 to 52 years), the median age at seizure onset was 5 years (range = 0.5 to 31 years), and the median disease duration was 18.5 years (range = 3.5 to 50.0 years). All patients had video-EEG telemetry with scalp electrodes, which was conclusive in 23 patients. Among the remaining 39 patients, 28 patients had video telemetry recordings with bilateral foramen ovale electrodes, 8 patients had recordings with bilateral eight-contact subdural frontal and/or subtemporal, and 3 patients had bilateral temporal depth electrodes in the anterior and posterior hippocampus and in the amygdala. Habitual seizures arose from the left temporal lobe in 35 and from the right in 27 patients. Fifty-two were right-handed, nine were left-handed, and one was ambidextrous. MRI scanning showed unilateral medial temporal abnormalities suggestive of mesial temporal sclerosis (MTS) in 47 patients and was reported as normal in 15 patients. All patients showed decreased metabolism in the epileptogenic temporal lobe. The mean FSIQ score was 90.2 (SD = 13.7). The IAT showed that 50 patients were left hemisphere dominant for language and 12 were nonleft hemisphere dominant (11 were mixed and 1 was right-hemisphere dominant). Patient characteristics are summarized in Table 1, grouped according to the laterality of language dominance. Statistical comparison with t- and Chi-square tests showed no significant differences in demographic and clinical characteristics between patients with left hemisphere language dominance and those with nonleft hemisphere dominance.

Table 1.  

Clinical and Demographic Characteristics According to Laterality of Language Dominance


All
Language Dominance
Left
Nonleft
a
Number of patients 62 50 12 
Side of focus (left/right) 35/27 26/24 9/3 
Sex (male/female) 40/22 34/16 6/6 
Handedness (right/left/ambidextrous) 52/9/1 44/5/1 8/4/0 
MRI findings (MTS/non-MTS) 47/15 38/12 9/3 
Age at IAT/PET (years) [median (range)] 28.5 (16–52) 28.5 (16–52) 30.0 (16–44) 
Age at onset of epilepsy (years) [median (range)] 5.0 (0.5–31) 5.0 (0.5–31) 4.5 (0.5–17) 
Duration of epilepsy (years) [median (range)] 18.5 (3.5–50) 17.75 (3.5–50) 22.0 (7–40) 
Full-scale IQ [mean (SD)] 90.2 (13.7) 91.0 (14.5) 86.5 (9.3) 
Leftb amobarbital dose (mg) [mean (SD)] 95.8 (14.4) 94.8 (14.2) 100.0 (15.1) 
Rightb amobarbital dose (mg) [mean (SD)] 95.3 (12.9) 94.5 (13.1) 98.3 (12.1) 

All
Language Dominance
Left
Nonleft
a
Number of patients 62 50 12 
Side of focus (left/right) 35/27 26/24 9/3 
Sex (male/female) 40/22 34/16 6/6 
Handedness (right/left/ambidextrous) 52/9/1 44/5/1 8/4/0 
MRI findings (MTS/non-MTS) 47/15 38/12 9/3 
Age at IAT/PET (years) [median (range)] 28.5 (16–52) 28.5 (16–52) 30.0 (16–44) 
Age at onset of epilepsy (years) [median (range)] 5.0 (0.5–31) 5.0 (0.5–31) 4.5 (0.5–17) 
Duration of epilepsy (years) [median (range)] 18.5 (3.5–50) 17.75 (3.5–50) 22.0 (7–40) 
Full-scale IQ [mean (SD)] 90.2 (13.7) 91.0 (14.5) 86.5 (9.3) 
Leftb amobarbital dose (mg) [mean (SD)] 95.8 (14.4) 94.8 (14.2) 100.0 (15.1) 
Rightb amobarbital dose (mg) [mean (SD)] 95.3 (12.9) 94.5 (13.1) 98.3 (12.1) 

MRI = magnetic resonance image; MTS = medial temporal sclerosis; SD = standard deviation; IAT = intracarotid amobarbital test; PET = positron emission tomography; IQ = intelligence quotient.

a

Right-sided dominance in 3 patients and mixed dominance in 9 patients.

b

Refer to the side of performance (contralateral to the injection).

IAT Memory Scores

Among all 62 patients, the mean of the standardized left IAT total score was 0.582 (SD = 0.262, range = 0.00 to 1.75), of the standardized left subscore for words, 0.681 (SD = 0.531, range = −1.00 to 3.00), for pictures, 0.531 (SD = 0.221, range = 0.00 to 1.00), and for faces, 0.526 (SD = 0.628, range = −1.00 to 2.00). The mean of the standardized right IAT total score was 0.418 (SD = 0.262, range = −0.75 to 1.00), of the standardized right subscore for words, 0.319 (SD = 0.531, range = −2.00 to 2.00), for pictures, 0.469 (SD = 0.221, range = 0.00 to 1.00), and for faces, 0.474 (SD = 0.628, range = −1.00 to 2.00). Table 2 shows the means of standardized IAT memory scores according to the side of epileptogenesis and laterality of language dominance. Exploratory analysis showed normal distribution of unilateral IAT scores for both standardized total and subscores, justifying the use of parametric statistical methods. A three-way ANOVA model was employed to investigate the effects of side of IAT, laterality of language dominance, and side of seizure focus and the interactions. This showed a significant effect of IAT side on standardized total IAT score: Standardized left total IAT scores were higher than standardized right side scores. There was also a trend toward standardized left subscores for words being higher than standardized right subscores for words. In addition, a highly significant interaction between side of IAT and side of focus was observed for standardized total IAT score and subscore for pictures. The standardized left total score and subscore for pictures were lower in patients with left TLE than in those with right TLE, whereas the standardized right total score and subscore for pictures were lower in patients with right TLE. No other significant interactions were shown.

Table 2.> 

Standardized IAT Scores as a Function of Language Dominance and Side of Epileptogenic Focus

Language Dominance
Side of Focus
IAT Material
Total
Word
Picture
Face
Left IAT scores [n; mean (SD, range)] 
Left Left n = 25 n = 24 n = 26 n = 21 
0.483 (0.212, 0.00 to 0.89) 0.664 (0.575, 0.00 to 3.00) 0.432 (0.238, 0.00 to 1.00) 0.454 (.639, −1.00 to 2.00) 
Right n = 24 n = 23 n = 24 n = 20 
0.740 (0.291, 0.41 to 1.75) 0.799 (0.560, −1.00 to 2.00) 0.667 (0.185, 0.25 to 1.00) 0.505 (0.694, −1.00 to 2.00) 
Nonleft Left n = 9 n = 9 n = 9 n = 9 
0.430 (0.067, 0.35 to 0.50) 0.450 (0.342, 0.00 to 1.00) 0.459 (0.054, 0.33 to 0.50) 0.689 (0.524, 0.00 to 1.50) 
Right n = 3 n = 3 n = 3 n = 2 
0.596 (0.038, 0.56 to 0.63) 0.611 (0.192, 0.50 to 0.83) 0.530 (0.026, 0.50 to 0.55) 0.750 (0.354, 0.50 to 1.00) 
 
Right IAT scores [n; mean (SD, range)] 
Left Left n = 25 n = 24 n = 26 n = 21 
0.517 (0.212, 0.11 to 1.00) 0.336 (0.575, −2.00 to 1.00) 0.568 (0.238, −1.00 to 2.00) 0.546 (0.639, 0.00 to 1.00) 
Right n = 24 n = 23 n = 24 n = 20 
0.260 (0.291, −0.75 to 0.59) 0.201 (0.560, −1.00 to 2.00) 0.334 (0.185, −1.00 to 2.00) 0.495 (0.694, 0.00 to 0.75) 
Nonleft Left n = 9 n = 9 n = 9 n = 9 
0.570 (0.067, 0.50 to 0.65) 0.550 (0.342, 0.00 to 1.00) 0.541 (0.054, −0.50 to 1.00) 0.311 (0.524, 0.50 to 0.67) 
Right n = 3 n = 3 n = 3 n = 2 
0.404 (0.038, 0.37 to 0.44) 0.389 (0.192, 0.17 to 0.50) 0.470 (0.026, 0.00 to 0.50) 0.250 (0.353, 0.45 to 0.50) 
 
Three-way ANOVA 
Main effect [F value (p value)] 
 Side of IAT  4.390 (.038) 3.679 (.058) 0.761 (.385) 1.093 (.298) 
 Language dominance  0.000 (1.000) 0.000 (1.000) 0.000 (1.000) 0.000 (1.000) 
 Side of focus  0.000 (1.000) 0.000 (1.000) 0.000 (1.000) 0.000 (1.000) 
Interaction [F value (p value)] 
 Side of focus × Side of IAT  12.623 (.001) 1.172 (.281) 9.278 (.003) 0.087 (.768) 
 Language dominance × Side of IAT  2.764 (.099) 2.153 (.145) 1.177 (.280) 1.585 (.211) 
 Language dominance × Side of focus  0.000 (1.000) 0.000 (1.000) 0.000 (1.000) 0.000 (1.000) 
 Language dominance × Side of focus × Side of IAT  0.582 (.447) 0.009 (.926) 2.647 (.106) 0.001 (.980) 
Language Dominance
Side of Focus
IAT Material
Total
Word
Picture
Face
Left IAT scores [n; mean (SD, range)] 
Left Left n = 25 n = 24 n = 26 n = 21 
0.483 (0.212, 0.00 to 0.89) 0.664 (0.575, 0.00 to 3.00) 0.432 (0.238, 0.00 to 1.00) 0.454 (.639, −1.00 to 2.00) 
Right n = 24 n = 23 n = 24 n = 20 
0.740 (0.291, 0.41 to 1.75) 0.799 (0.560, −1.00 to 2.00) 0.667 (0.185, 0.25 to 1.00) 0.505 (0.694, −1.00 to 2.00) 
Nonleft Left n = 9 n = 9 n = 9 n = 9 
0.430 (0.067, 0.35 to 0.50) 0.450 (0.342, 0.00 to 1.00) 0.459 (0.054, 0.33 to 0.50) 0.689 (0.524, 0.00 to 1.50) 
Right n = 3 n = 3 n = 3 n = 2 
0.596 (0.038, 0.56 to 0.63) 0.611 (0.192, 0.50 to 0.83) 0.530 (0.026, 0.50 to 0.55) 0.750 (0.354, 0.50 to 1.00) 
 
Right IAT scores [n; mean (SD, range)] 
Left Left n = 25 n = 24 n = 26 n = 21 
0.517 (0.212, 0.11 to 1.00) 0.336 (0.575, −2.00 to 1.00) 0.568 (0.238, −1.00 to 2.00) 0.546 (0.639, 0.00 to 1.00) 
Right n = 24 n = 23 n = 24 n = 20 
0.260 (0.291, −0.75 to 0.59) 0.201 (0.560, −1.00 to 2.00) 0.334 (0.185, −1.00 to 2.00) 0.495 (0.694, 0.00 to 0.75) 
Nonleft Left n = 9 n = 9 n = 9 n = 9 
0.570 (0.067, 0.50 to 0.65) 0.550 (0.342, 0.00 to 1.00) 0.541 (0.054, −0.50 to 1.00) 0.311 (0.524, 0.50 to 0.67) 
Right n = 3 n = 3 n = 3 n = 2 
0.404 (0.038, 0.37 to 0.44) 0.389 (0.192, 0.17 to 0.50) 0.470 (0.026, 0.00 to 0.50) 0.250 (0.353, 0.45 to 0.50) 
 
Three-way ANOVA 
Main effect [F value (p value)] 
 Side of IAT  4.390 (.038) 3.679 (.058) 0.761 (.385) 1.093 (.298) 
 Language dominance  0.000 (1.000) 0.000 (1.000) 0.000 (1.000) 0.000 (1.000) 
 Side of focus  0.000 (1.000) 0.000 (1.000) 0.000 (1.000) 0.000 (1.000) 
Interaction [F value (p value)] 
 Side of focus × Side of IAT  12.623 (.001) 1.172 (.281) 9.278 (.003) 0.087 (.768) 
 Language dominance × Side of IAT  2.764 (.099) 2.153 (.145) 1.177 (.280) 1.585 (.211) 
 Language dominance × Side of focus  0.000 (1.000) 0.000 (1.000) 0.000 (1.000) 0.000 (1.000) 
 Language dominance × Side of focus × Side of IAT  0.582 (.447) 0.009 (.926) 2.647 (.106) 0.001 (.980) 

Scores and subscores refer to the tested side (contralateral to the injection side) and are calculated as (Score Left or Right)/(Score Left + Right).

Patients where the denominator of this formula is zero have been excluded from the analysis.

n = number of patients; SD = standard deviation; IAT = intracarotid amobarbital test; ANOVA = analysis of variance.

With regard to nonstandardized IAT scores, the mean of the nonstandardized left IAT total score was 8.74 (SD = 3.343, range = −1 to 15), of the nonstandardized left subscore for words was 2.82 (SD = 1.542, range −1 to 5), for pictures was 4.63 (SD = 1.822, range = 0 to 6), and for faces was 1.29 (SD = 2.068, range = −3 to 6). The mean of the nonstandardized right IAT total score was 6.98 (SD = 4.123, range = −3 to 15), of the nonstandardized right subscore for words was 1.79 (SD = 2.058, range = −2 to 6), for pictures was 3.95 (SD = 1.778, range 0 to 6), and for faces was 1.24 (SD = 1.753, range = −3 to 6).

SPM Correlations between [18F]FDG Uptake and IAT Memory Performance

Total Memory Scores

Correlations between [18F]FDG uptake and standardized total IAT scores are shown in Table 3 and are displayed in Figure 1. The standardized left total IAT scores were strongly correlated with [18F]FDG uptake in the left temporal lobe, both in the lateral and the medial temporal regions, and in the left ventrolateral frontal cortex (n = 61, left TLE = 34, right TLE = 27). A weak correlation was observed with the contralateral right cerebellar hemisphere. Correlations between [18F]FDG uptake and nonstandardized left total IAT scores essentially showed similar results except for the absence of correlations between nonstandardized left IAT scores and [18F]FDG uptake in the left medial temporal regions.

Figure 1. 

Statistical parametric maps showing regions with significant correlations between standardized total IAT scores and [18F]FDG uptake in the whole patient sample (n = 61). (A) Suprathreshold areas for the standardized left (in red) and right (in blue) total IAT scores at p < .01 rendered on the surface of the Montreal Neurological Institute (MNI)-averaged brain image. (B) Left: Suprathreshold areas at p < .01 for the standardized left total IAT score at (−24 −18 20) projected on the averaged MNI T1-weighted MRI with a color scale of t values. Right: Suprathreshold areas at p < .01 for the standardized right total IAT score at (28 18 −10) projected on the averaged MNI T1-weighted MRI with a color scale of t values.

Figure 1. 

Statistical parametric maps showing regions with significant correlations between standardized total IAT scores and [18F]FDG uptake in the whole patient sample (n = 61). (A) Suprathreshold areas for the standardized left (in red) and right (in blue) total IAT scores at p < .01 rendered on the surface of the Montreal Neurological Institute (MNI)-averaged brain image. (B) Left: Suprathreshold areas at p < .01 for the standardized left total IAT score at (−24 −18 20) projected on the averaged MNI T1-weighted MRI with a color scale of t values. Right: Suprathreshold areas at p < .01 for the standardized right total IAT score at (28 18 −10) projected on the averaged MNI T1-weighted MRI with a color scale of t values.

Table 3.> 

Regions with Suprathreshold t-statistics for Regression between Standardized Memory Scores and [18F]FDG Uptake in all Patients

Memory scores
Side
Regions
Size
x
y
z
t-Statistic
Left scores 
Total Left Medial temporal structures, STG, MTG, ITG 18,738 −68 −56 −14 5.44 
Left Inferior frontal gyrus 293 −20 68 −20 4.46 
Left CBL 524 −36 −94 −28 4.32 
Right CBL* – 40 −44 −50 2.88 
Right CBL# – 30 −96 −36 2.26 
Word Left CBL 115 −26 −90 −26 3.59 
Left MTG; STG*, ITG*, cuneus*, occipital gyri* 218 −66 −38 −2 3.53 
Left Inferior frontal gyrus# – −30 66 −8 2.47 
Picture Left Parahippocampal gyrus, MTG, ITG; hippocampus*, amygdala*, STG*; PCG# 10,814 −66 −40 −30 4.92 
Left Middle frontal gyrus* – −56 38 18 3.66 
Left ACG# – −4 38 14 2.18 
Left Inferior frontal gyrus# – −14 74 −10 2.04 
 
Right scores 
Total Right IPL; MTG*, ITG* 2992 58 −54 40 4.84 
Right Superior frontal gyrus 1028 38 62 16 4.30 
Right Precentral gyrus 401 52 48 3.87 
Bilat ACG 780 10 44 3.83 
Right PCG* – 10 −54 36 3.40 
Right Thalamus* – 26 −14 3.02 
Right MTG* – 42 16 −48 2.62 
Word Bilat/Right Bilat ACG*; right middle frontal gyrus# – −18 40 3.36 
Right STG*, MTG* – 46 28 −34 3.30 
Picture Bilat/Right Right IPL, bilat ACG 5741 18 −10 48 4.64 
Right STG*, MTG*; hippocampus#, parahippocampal gyrus# – 58 −12 3.24 
Right ITG* – 42 16 −50 3.22 
Right Insula*; thalamus#, caudate nucleus# – 30 −20 10 2.89 
Memory scores
Side
Regions
Size
x
y
z
t-Statistic
Left scores 
Total Left Medial temporal structures, STG, MTG, ITG 18,738 −68 −56 −14 5.44 
Left Inferior frontal gyrus 293 −20 68 −20 4.46 
Left CBL 524 −36 −94 −28 4.32 
Right CBL* – 40 −44 −50 2.88 
Right CBL# – 30 −96 −36 2.26 
Word Left CBL 115 −26 −90 −26 3.59 
Left MTG; STG*, ITG*, cuneus*, occipital gyri* 218 −66 −38 −2 3.53 
Left Inferior frontal gyrus# – −30 66 −8 2.47 
Picture Left Parahippocampal gyrus, MTG, ITG; hippocampus*, amygdala*, STG*; PCG# 10,814 −66 −40 −30 4.92 
Left Middle frontal gyrus* – −56 38 18 3.66 
Left ACG# – −4 38 14 2.18 
Left Inferior frontal gyrus# – −14 74 −10 2.04 
 
Right scores 
Total Right IPL; MTG*, ITG* 2992 58 −54 40 4.84 
Right Superior frontal gyrus 1028 38 62 16 4.30 
Right Precentral gyrus 401 52 48 3.87 
Bilat ACG 780 10 44 3.83 
Right PCG* – 10 −54 36 3.40 
Right Thalamus* – 26 −14 3.02 
Right MTG* – 42 16 −48 2.62 
Word Bilat/Right Bilat ACG*; right middle frontal gyrus# – −18 40 3.36 
Right STG*, MTG* – 46 28 −34 3.30 
Picture Bilat/Right Right IPL, bilat ACG 5741 18 −10 48 4.64 
Right STG*, MTG*; hippocampus#, parahippocampal gyrus# – 58 −12 3.24 
Right ITG* – 42 16 −50 3.22 
Right Insula*; thalamus#, caudate nucleus# – 30 −20 10 2.89 

MNI coordinates, t-statistics, and sizes (the number of voxels included at p < .001) of the highest t-statistics peak within each cluster with a threshold of p < .001 are shown.

Peak coordinates and t-statistics of regions depicted with a threshold of p < .01 (*) and p < .05 (#) are also shown.

Because the lower thresholds (p < .01 and p < .05) were employed for exploratory purposes, the size of each cluster detected at p < .01 or p < .05 was not shown.

Medial temporal structures include the hippocampus, amygdala, subiculum, and parahippocampal gyrus.

[18F]FDG = [18F]fluorodeoxyglucose; STG = superior temporal gyrus; MTG = middle temporal gyrus; ITG = inferior temporal gyrus; CBL = cerebellar hemisphere; IPL = inferior parietal lobule; Bilat = bilateral; ACG = anterior cingulate gyrus; PCG = posterior cingulate gyrus.

The standardized right total scores were significantly correlated with [18F]FDG uptake in the right inferior parietal lobule, the right dorsolateral frontal cortex, the right precentral gyrus, and the caudal portion of the right anterior cingulate cortex (n = 61, left TLE = 34, right TLE = 27). With a less stringent significance level of p < .01, correlations were found with the posterior portion of the right middle and inferior temporal gyri, the right posterior cingulate cortex including the retrosplenial cortex, and the right thalamus. Uptake by the right medial temporal regions did not correlate even with the least stringent threshold of p < .05. Correlations between [18F]FDG uptake and nonstandardized right total IAT scores essentially showed similar results except for the absence of correlations between nonstandardized right IAT scores and [18F]FDG uptake in the right thalamus.

Memory Subscores for Words

The standardized left IAT subscores for words were significantly correlated with [18F]FDG uptake in the posterior portion of the left middle temporal gyrus with a significance level of p < .001 (n = 59, left TLE = 33, right TLE = 26) (Table 3). Less significant correlations were found with the left lateral and medial occipital cortex (p < .01); and in the left ventrolateral frontal cortex and the posterior part of the left lateral temporal region (p < .05). No correlations were found with regional metabolic rates in the left medial temporal regions. There was no region showing significant correlations with nonstandardized left IAT subscores for words.

The standardized right subscores for words did not correlate with any regions with a threshold of p < .001 (n = 59, left TLE = 33, right TLE = 26) (Table 3). With more lenient significance levels, correlations were observed with the tip of the right superior and middle temporal gyri and the anterior cingulate cortex bilaterally (p < .01), and the right dorsolateral frontal cortex (p < .05). When using nonstandardized IAT subscores for words, a correlation was found with the right occipital gyri (p < .01) and right inferior parietal lobule (p < .05). Again, uptake by the right medial temporal region did not correlate with standardized or nonstandardized right subscores for words.

Memory Subscores for Pictures

The standardized left IAT subscore for pictures significantly correlated with [18F]FDG uptake in the left middle and inferior temporal gyri and part of the left medial temporal structures (the parahippocampal gyrus, but not with the hippocampus and amygdala) (p < .001) (n = 62, left TLE = 35, right TLE = 27) (Table 3). A weaker correlation was observed with all gyri of the left lateral temporal cortex, the hippocampus, the amygdala, the left dorsolateral frontal cortex (p < .01), the tip of the left ventrolateral frontal cortex, the anterior cingulate cortex, and the posterior cingulate cortex (including the retrosplenial cortex) (p < .05). Correlations between [18F]FDG uptake and nonstandardized left IAT subscores for pictures essentially showed similar results except for the presence of a weaker correlation in the left medial temporal structures (hippocampus and parahippocampal gyrus; p < .05) and the absence of correlations between left IAT subscores for pictures and [18F]FDG uptake in the left anterior and posterior cingulate cortices.

The standardized right subscores for pictures correlated with [18F]FDG uptake in the right inferior parietal lobule and the anterior cingulate cortex bilaterally (p < .001) (n = 62, left TLE = 35, right TLE = 27) (Table 3). A weaker association was found in the anterior part of the right lateral temporal cortex and the insula (p < .01), as well as the right thalamus and the head of the caudate nucleus (p < .05). Within the right medial temporal structures, the body of the hippocampus and the middle portion of the parahippocampal gyrus were weakly correlated with the standardized right subscore for pictures (p < .05). Correlations between [18F]FDG uptake and nonstandardized right IAT subscores for pictures essentially showed similar results except for the absence of correlations between right IAT subscores for pictures and [18F]FDG uptake in right medial and lateral temporal regions, the right thalamus, and the right caudate nucleus.

Memory Subscores for Faces

There was no region showing significant correlations with left or right standardized (n = 52, left TLE = 30, right TLE = 22) or nonstandardized subscores for faces.

ROI Correlations between [18F]FDG Uptake and IAT Memory Performance

To exclude potential artifacts arising from spatial normalization of the original images used by SPM, we carried out ROI analysis using standardized IAT scores. In particular, this was to examine the question of whether our initial SPM analysis had missed significant correlations in the right hippocampus/amygdala complex despite lowering the voxelwise threshold to the liberal level of p < .05. Among all 62 patients, ROI values from the left amygdala and hippocampus were significantly correlated with the standardized left total score (n = 61, r = .516, p < .001) and the standardized left subscore for pictures (n = 62, r = .365, p = .004), but not with the standardized left subscore for words (n = 59, r = .140, p = .291) or for faces (n = 52, r = .106, p = .453). Concerning the standardized right total and subscores, no significant correlations were observed with the ROI values of the right amygdala and hippocampus (n = 61, r = .097, p = .456 for the standardized total score; n = 59, r = .067, p = .612 for the standardized subscore for words; n = 62, r = .179, p = .164 for the standardized subscore for pictures; and n = 52, r = −.136, p = .336 for the standardized subscore for faces).

Effect of Medial Temporal Sclerosis and Normal Neuroimaging on the Correlation between IAT and [18F]FDG Uptake

Although we excluded patients with detectable focal lesions on MRI, our subjects were not pathologically homogenous. We therefore carried out a subsidiary analysis using standardized IAT scores to compare those with MRI evidence of MTS (n = 47, left TLE = 27, right TLE = 21) to those with normal MRI (n = 15, left TLE = 8, right TLE = 6). Among the patients who had MRI evidence of MTS, the topography of the correlated areas was not strikingly different from that of the whole group, albeit with some reduction in statistical significance. In the subgroup of patients with normal MRI, the topography of brain regional correlations was not different from that among patients with MRI evidence of MTS with the exception that the standardized left subscore for words was correlated with [18F]FDG uptake in the left hippocampus and in the amygdala.

Effect of Laterality of Language Dominance on the Relationship between [18F]FDG Uptake and IAT

To explore topographical differences in the correlation between standardized IAT scores and [18F]FDG uptake according to the language dominance, we studied patients with left (n = 50, left TLE = 26, right TLE = 24) and nonleft (n = 12, left TLE = 9, right TLE = 3) dominance for language separately. The results are shown in Figure 2 and Table 4. In patients with left dominance for language, the topography of the correlated areas was not different from that of the whole group, albeit there was some reduction in statistical significance (Table 4A). In contrast, the topography of the correlated areas in patients with nonleft dominance for language was different: Correlations with the standardized left total scores showed reduced involvement of left lateral and medial temporal regions (Figure 2, compare the left column, upper and lower panels). In addition, correlations between the standardized right total score (and the standardized right subscore for pictures) were found with right fronto-temporal regions, including the inferior, medial, and lateral temporal regions, and the right ventrolateral and ventromedial frontal cortices (Figure 2, compare the right column, upper and lower panels, and Table 4B).

Figure 2. 

Topographical distribution of significant correlations between [18F]FDG uptake and standardized total IAT scores in patients according to laterality of language dominance. The left column shows the areas with significant correlations with the standardized left total IAT scores, whereas the right column displays those showing correlations with the standardized right total IAT scores in patients with left (top) and nonleft (bottom) dominance for language. Suprathreshold areas at a threshold of p < .05 are projected on the averaged Montreal Neurological Institute T1-weighted MRI with a color scale of t values.

Figure 2. 

Topographical distribution of significant correlations between [18F]FDG uptake and standardized total IAT scores in patients according to laterality of language dominance. The left column shows the areas with significant correlations with the standardized left total IAT scores, whereas the right column displays those showing correlations with the standardized right total IAT scores in patients with left (top) and nonleft (bottom) dominance for language. Suprathreshold areas at a threshold of p < .05 are projected on the averaged Montreal Neurological Institute T1-weighted MRI with a color scale of t values.

Table 4. 

Regions with Suprathreshold t-statistics for Regression between Standardized Memory Scores and [18F]FDG Uptake in Patients with Left and Nonleft Dominance for Language.

Memory Material
Side
Regions
Size
x
y
z
t-Statistic
(A) Results in the patients with left dominance for language 
Left scores 
Total Left Inferior frontal gyrus 342 −18 66 −26 5.08 
Left Medial temporal structures, ITG, MTG; STG*; insula# 9612 −64 −56 −20 4.92 
Left CBL 126 −34 −94 −30 3.88 
Right CBL* – 42 −38 −42 3.12 
Word Left CBL* – −26 −90 −26 3.36 
Left MTG*, ITG*; STG#, insula# – −68 −40 3.18 
Left Precuneus* – −10 −64 −4 2.94 
Left Cuneus# – −16 −84 30 2.36 
Left Inferior frontal gyrus# – −30 66 −8 2.18 
Left PCG# – −20 32 2.03 
Picture Left STG, MTG, ITG; medial temporal structures*; PCG# 5122 −66 −40 −30 4.48 
Left Insula 398 −42 −6 4.26 
Left Middle frontal gyrus# – −58 36 18 3.28 
 
Right scores 
Total Right Middle frontal gyrus 624 38 62 16 4.54 
Right IPL; PCG*; occipital gyri# 1923 58 −54 38 4.41 
Right Precentral gyrus 124 56 46 3.64 
Right ACG 124 44 3.58 
Right MTG* – 76 −28 −14 3.08 
Right Thalamus# – 16 −12 10 2.57 
Word Right STG*, MTG* – 50 30 −38 4.46 
Left ACG# – −18 40 3.14 
Left Middle frontal gyrus# – −18 28 22 2.55 
Right Middle frontal gyrus# – 28 34 12 2.45 
Picture Right ACG, PCG, IPL; hippocampus#, thalamus#, putamen#, globus pallidum#, insula# 2933 78 −24 26 4.17 
Right STG*; MTG# – 56 10 −12 3.15 
 
(B) Results in patients with nonleft dominance for laanguage 
Left scores 
Total Left MTG, ITG; STG# 417 −64 −20 −26 6.48 
Left Parahippocampal gyrus*; hippocampus#, CBL# – −40 −32 −12 6.96 
Left Occipital gyri*; IPL# – −42 −94 −2 3.24 
Left ACG# – −12 38 12 3.27 
Left Middle frontal gyrus# – −54 40 10 2.81 
Right CBL# – 24 −44 −50 2.61 
Word Left PCG# – −14 −46 10 4.71 
Left IPL# – −68 −64 22 4.17 
Left Cuneus# – −16 −80 14 2.53 
Right CBL – 36 −40 −38 2.41 
Left CBL – −38 −40 −34 2.18 
Picture Left MTG; ITG# 156 −70 −18 −20 8.03 
Left Occipital gyri 139 −44 −82 −4 5.56 
Left IPL 276 −50 −52 22 5.16 
Left CBL 434 −44 −46 −30 5.13 
Left Middle frontal gyrus* – −18 46 22 5.53 
Left Parahippocampal gyrus* – −40 −34 −12 5.27 
Right CBL* – 30 −72 −42 4.03 
 
Right scores 
Total Right STG, MTG, ITG; parahippocampal gyrus*, insula*, orbital gyri*; hippocampus#, amygdala#; putamen#, globus pallidum# 628 56 −28 4.85 
Right IPL* – 66 −52 26 3.69 
Right Frontal pole*, superior frontal gyrus# – 14 64 32 3.59 
Right PCG# – 14 −36 30 2.88 
Word Right STG*, MTG*, orbital gyri* – 62 14 −16 3.73 
Left Superior parietal lobule# – −52 −38 54 3.47 
Left Orbital gyri#, inferior frontal gyrus# – −26 50 −4 3.34 
Right Middle frontal gyrus#, inferior frontal gyrus# – 42 −8 38 3.08 
Bilat ACG#, PCG# – −24 54 2.66 
Right ACG# – 14 52 2.39 
Picture Right Putamen, insula; caudate nucleus*; globus pallidum# 1138 30 −12 5.88 
Right IPL 289 62 −54 38 5.34 
Right MTG, ITG; STG*, orbital gyri*, inferior frontal gyrus*; medial temporal structures# 562 56 10 −30 5.10 
Right Middle frontal gyrus 111 40 42 −4 5.09 
Left Putamen*, caudate nucleus*; globus pallidum# – −24 3.94 
Right Cuneus* – 14 −90 20 3.79 
Left Orbital gyri* – −24 22 −14 3.38 
Bilat PCG# – 16 −38 28 3.72 
Left Cuneus# – −16 −88 −8 3.11 
Memory Material
Side
Regions
Size
x
y
z
t-Statistic
(A) Results in the patients with left dominance for language 
Left scores 
Total Left Inferior frontal gyrus 342 −18 66 −26 5.08 
Left Medial temporal structures, ITG, MTG; STG*; insula# 9612 −64 −56 −20 4.92 
Left CBL 126 −34 −94 −30 3.88 
Right CBL* – 42 −38 −42 3.12 
Word Left CBL* – −26 −90 −26 3.36 
Left MTG*, ITG*; STG#, insula# – −68 −40 3.18 
Left Precuneus* – −10 −64 −4 2.94 
Left Cuneus# – −16 −84 30 2.36 
Left Inferior frontal gyrus# – −30 66 −8 2.18 
Left PCG# – −20 32 2.03 
Picture Left STG, MTG, ITG; medial temporal structures*; PCG# 5122 −66 −40 −30 4.48 
Left Insula 398 −42 −6 4.26 
Left Middle frontal gyrus# – −58 36 18 3.28 
 
Right scores 
Total Right Middle frontal gyrus 624 38 62 16 4.54 
Right IPL; PCG*; occipital gyri# 1923 58 −54 38 4.41 
Right Precentral gyrus 124 56 46 3.64 
Right ACG 124 44 3.58 
Right MTG* – 76 −28 −14 3.08 
Right Thalamus# – 16 −12 10 2.57 
Word Right STG*, MTG* – 50 30 −38 4.46 
Left ACG# – −18 40 3.14 
Left Middle frontal gyrus# – −18 28 22 2.55 
Right Middle frontal gyrus# – 28 34 12 2.45 
Picture Right ACG, PCG, IPL; hippocampus#, thalamus#, putamen#, globus pallidum#, insula# 2933 78 −24 26 4.17 
Right STG*; MTG# – 56 10 −12 3.15 
 
(B) Results in patients with nonleft dominance for laanguage 
Left scores 
Total Left MTG, ITG; STG# 417 −64 −20 −26 6.48 
Left Parahippocampal gyrus*; hippocampus#, CBL# – −40 −32 −12 6.96 
Left Occipital gyri*; IPL# – −42 −94 −2 3.24 
Left ACG# – −12 38 12 3.27 
Left Middle frontal gyrus# – −54 40 10 2.81 
Right CBL# – 24 −44 −50 2.61 
Word Left PCG# – −14 −46 10 4.71 
Left IPL# – −68 −64 22 4.17 
Left Cuneus# – −16 −80 14 2.53 
Right CBL – 36 −40 −38 2.41 
Left CBL – −38 −40 −34 2.18 
Picture Left MTG; ITG# 156 −70 −18 −20 8.03 
Left Occipital gyri 139 −44 −82 −4 5.56 
Left IPL 276 −50 −52 22 5.16 
Left CBL 434 −44 −46 −30 5.13 
Left Middle frontal gyrus* – −18 46 22 5.53 
Left Parahippocampal gyrus* – −40 −34 −12 5.27 
Right CBL* – 30 −72 −42 4.03 
 
Right scores 
Total Right STG, MTG, ITG; parahippocampal gyrus*, insula*, orbital gyri*; hippocampus#, amygdala#; putamen#, globus pallidum# 628 56 −28 4.85 
Right IPL* – 66 −52 26 3.69 
Right Frontal pole*, superior frontal gyrus# – 14 64 32 3.59 
Right PCG# – 14 −36 30 2.88 
Word Right STG*, MTG*, orbital gyri* – 62 14 −16 3.73 
Left Superior parietal lobule# – −52 −38 54 3.47 
Left Orbital gyri#, inferior frontal gyrus# – −26 50 −4 3.34 
Right Middle frontal gyrus#, inferior frontal gyrus# – 42 −8 38 3.08 
Bilat ACG#, PCG# – −24 54 2.66 
Right ACG# – 14 52 2.39 
Picture Right Putamen, insula; caudate nucleus*; globus pallidum# 1138 30 −12 5.88 
Right IPL 289 62 −54 38 5.34 
Right MTG, ITG; STG*, orbital gyri*, inferior frontal gyrus*; medial temporal structures# 562 56 10 −30 5.10 
Right Middle frontal gyrus 111 40 42 −4 5.09 
Left Putamen*, caudate nucleus*; globus pallidum# – −24 3.94 
Right Cuneus* – 14 −90 20 3.79 
Left Orbital gyri* – −24 22 −14 3.38 
Bilat PCG# – 16 −38 28 3.72 
Left Cuneus# – −16 −88 −8 3.11 

MNI coordinates, t-statistics, and sizes (the number of voxels included at p < .001) of the highest t-statistics peak within each cluster with a threshold of p < .001 are shown. Peak coordinates and t-statistics (but not sizes) of regions depicted with a threshold of p < .01 (*) and p < .05 (#) are also shown. Because the lower thresholds (p < .01 and p < .05) were employed for exploratory purposes, the size of each cluster detected at p < .01 or p < .05 was not shown. Medial temporal structures = hippocampus, amygdala, subiculum, and parahippocampal gyrus.

[18F]FDG = [18F]fluorodeoxyglucose; STG = superior temporal gyrus; MTG = middle temporal gyrus; ITG = inferior temporal gyrus; CBL = cerebellar hemisphere; PCG = posterior cingulate gyrus; IPL = inferior parietal lobule; ACG = anterior cingulate gyrus; Bilat = bilateral.

Correlations with Other Cognitive Domains

There were no correlations between [18F]FDG uptake in the ROIs and FSIQ, VIQ, or PIQ, suggesting that our findings are specific to the memory domain.

DISCUSSION

The present study shows a robust topographic association between unilateral episodic memory scores and ipsilateral cerebral metabolism in a group of patients with unilateral, nonlesional TLE. There were no correlations between unilateral IAT scores and contralateral cerebral regions. Our findings suggest that episodic memory is subserved by more widespread cortical regions beyond the core mesiotemporal lobe memory structures, and by different networks in the right and left hemispheres, particularly among left-dominant subjects. On the left, medial and lateral temporal cortices appear to be primarily involved, whereas on the right, the fronto-parietal cortex and the anterior cingulate appear to be most involved, and, to a lesser degree, the posterior lateral temporal cortex. Interestingly, no involvement of the right hippocampus was found, possibly due to the small size of this structure compared to our extent threshold in SPM analysis. Our findings also indicate a degree of topographic specialization for material-specific memory and suggest that the functional anatomy of episodic memory may depend on laterality of the language dominance. To our knowledge, this is the first study that has mapped brain regions subserving episodic memory using an objective, voxel-based image analysis (SPM), corroborated by ROI approach. In general, correlations between [18F]FDG uptake and IAT scores showed essentially similar results, regardless of whether standardized or nonstandardized IAT measures were used, except that regions correlating with nonstandardized scores tended to be more localized and correlations with standardized scores occurred at higher significance levels. These effects might be due to the fact that standardized IAT scores are really a measure of memory lateralization rather than a measure of absolute unilateral memory performance.

Regions Involved in Memory on the Left Hemisphere

In the left hemisphere, standardized IAT memory scores not only correlated robustly with medial temporal structures but also with lateral temporal and extratemporal regions, including the retrosplenial cortex and orbitomedial frontal regions. The medial temporal cortex has unequivocal importance in subserving episodic memory, supported by evidence in normal and memory-disordered subjects (Squire, Stark, & Clark, 2004; Kopelman, 2002). In addition, our findings suggest involvement of the left inferolateral temporal cortex in supporting episodic memory, in line with evidence from studies that have compared verbal memory and IQ with [18F]FDG-PET in patients with TLE (Arnold et al., 1996; Rausch, Henry, Ary, Engel, & Mazziotta, 1994). This observation may explain why patients who undergo left standard anterior temporal lobectomy tend to show greater verbal memory decline than those who have more restricted temporal resections such as left amygdalo-hippocampectomy (Lacruz et al., 2004; Pauli et al., 1999; Goldstein & Polkey, 1993; Wieser & Yasargil, 1982). In contrast to findings from standardized IAT scores, the correlation between hippocampal [18F]FDG uptake and nonstandardized IAT scores was only significant for the left subscore for pictures.

In addition to temporal structures, we have identified the involvement of the ventrolateral frontal cortex. This region has been shown to be involved in paired associate learning (Shallice et al., 1994) and novel word list learning (Kopelman, Stevens, Foli, & Grasby, 1998). Functional neuroimaging studies in normal healthy individuals suggest that encoding verbal information activates the inferior frontal cortex (Kirchhoff, Wagner, Maril, & Stern, 2000; Wagner et al., 1998; Rugg, Fletcher, Frith, Frackowiak, & Dolan, 1996). In the current study, the anesthetization was at encoding, but conditions that show the associations involved words or material that is verbally encoded (objects). These findings are also consistent with neuroimaging results that suggest the posterior inferior frontal cortex supports access to and is involved in the maintenance of phonological codes (e.g., Poldrack et al., 1999; Fiez & Petersen, 1998). Finally, the left IAT measurement was also correlated with the retrosplenial cortex, a structure that appears to be involved in episodic memory networks in a number of studies from our own group (Reed et al., 1999, 2003, 2005) and others (e.g., Desgranges et al., 1998).

Regions Involved in Memory on the Right Hemisphere

In contrast to the substantial evidence of temporal lobe participation in the left hemisphere, IAT memory function in the right hemisphere correlated with metabolic uptake in ipsilateral frontal lobes (the dorsolateral frontal cortex, the right precentral gyrus, and the caudal portion of the right anterior cingulate cortex) and parietal lobes, as well as with subcortical regions such as the thalamus and basal ganglia. There was no evidence of medial temporal or limbic involvement. This finding was corroborated by using ROI and with analysis under less stringent statistical criteria which decrease the chance of Type II errors (false negatives). Only some involvement of the right hippocampus and parahippocampal gyrus occurred in association with the standardized subscore for pictures. fMRI studies using encoding-associated tasks have shown bilateral medial temporal activation in normal subjects and asymmetrical activations in patients with unilateral TLE, in favor of the side contralateral to the epileptogenic hemisphere (Deblaere et al., 2005; Richardson, Strange, Duncan, & Dolan, 2003; Golby et al., 2002; Detre et al., 1998). These studies often used laterality indices with peak values or volumes of voxels activated. fMRI laterality indices were often concordant with IAT laterality indices. It has also been shown that patients with TLE showed more activation outside temporal regions than normal subjects (Dupont et al., 2000, 2002), suggesting reorganization of functions in patients with medial temporal damage. Our results are consistent with a distinct fronto-parietal neuroanatomical network subserving episodic memory in the right hemisphere, different from that in the left hemisphere, particularly among left-dominant subjects. Interestingly, these are the regions involved in supporting working memory as suggested by functional imaging and neuropsychological studies in normal subjects and in patients (for review, see Wager & Smith, 2003; Fletcher & Henson, 2001; Duncan & Owen, 2000). Right frontal involvement for encoding pictorial material has been observed in functional neuroimaging studies, but this tends to be in the inferior frontal region (Kirchhoff et al., 2000; Kelley et al., 1998; Wagner et al., 1998) and has been linked to the notion that this region supports control processes that provide access to and maintenance of visuospatial material and visuo-object representations (Wagner, 1999; Awh & Jonides, 1998).

Regions Involved in Material-specific Tasks

Unilateral subscores for faces showed no significant associations with cerebral metabolism, probably due to the lower number of subjects included in the analysis of this task because of the high number of zero scores.

Unilateral left subscores for words or pictures were correlated with metabolism in the left lateral temporal, lateral frontal, and occipital lobes, in keeping with those regions being correlated with left total IAT scores. Interestingly, the metabolic rate of the left medial temporal region correlated with subscore for pictures (possibly due to verbal encoding of the pictures of objects), but not for words. Memory for words and pictures when testing the right hemisphere was correlated with the metabolic rate in the right fronto-parietal areas, again in keeping with those regions being correlated with right total IAT scores. There was also a trend toward a correlation between standardized right subscores for pictures and right medial temporal metabolism. The absence of correlation between left medial temporal metabolism and left subscores for words, and the poor correlations found with right medial temporal metabolism are rather puzzling. Absence of such correlations might be due to the lower memory scores achieved by the right hemisphere, or to technical issues, such as partial volume effects within our extent threshold or spatial normalization, or to the fact that the internal carotid artery does not supply the posterior third of the hippocampus, which might not have been anesthetized during the IAT. Partial volume effects increase with stronger atrophy, enhancing the possibility of detecting negative correlations between episodic memory performance with memory performance in the medial temporal lobe.

Effects of Laterality of Language Dominance

In patients with left dominance for language, a significant correlation was found between metabolism of the entire left temporal lobe and left hemisphere memory performance, both for standardized total scores and for standardized subscores for pictures. The effect of the laterality of the epileptic focus on these findings should be minimal, as foci were evenly distributed between the two sides. Among the patients with nonleft language dominance, standardized right and left total memory scores and subscores for pictures correlated with ipsilateral temporal metabolism rather symmetrically, and standardized right subscores for words correlated with right temporal metabolism. This is in contrast to the findings seen among left-dominant patients where right IAT scores were correlated with metabolism in right fronto-parietal areas, and showed no right temporal involvement. The effect of the epileptogenic side within this patient group may not be negligible (9 of the 12 patients had left TLE in nonleft language dominance). This is in line with recent findings showing increased nonleft language dominance in patients with left TLE, an effect most marked in those with epilepsy onset at an earlier age (Brazdil, Zakopcan, Kuba, Fanfrdlova, & Rektor, 2003). It is possible that in TLE, epileptogenesis may develop before language and memory lateralization, thus precluding normal left-hemispheric specialization.

Limitations

It might be surprising that a relation between unilateral memory performance with resting state metabolic activity provides insight into the networks necessary for memory performance. Pathology in one region of the brain may have distant metabolic effects on others. It is possible that areas with structural pathology are driving the metabolic changes seen in the regions reported and that these are not necessary for memory processing per se. Further, it is not clear why resting state metabolism would reveal networks utilized during the actual memory task, although there is work suggesting the role of “default” networks in memory function (e.g., Greicius, Krasnow, Reiss, & Menon, 2003; Mazoyer et al., 2001). As mentioned above, the relationship between unilateral memory performance and [18F]FDG-PET is ipsilateral, suggesting that the results are not random or artifacts. However, they may reflect more general issues with the function of that hemisphere as opposed to being specific for memory function.

The activity of hippocampal structures might have been underrepresented in our results due to their small volume compared to our extent threshold (Park et al., 2004; Hammers et al., 2002).

Conclusion

In the present study, we have shown that unilateral episodic memory as estimated by the IAT correlated with cortical and subcortical metabolism in ipsilateral brain regions, suggesting that the findings are not artifactual or due to randomness. Different structures were involved in the left and the right hemispheres. On the left, lateral temporal regions, and probably medial temporal regions, appear to be primarily involved, whereas on the right, the fronto-parietal cortex and the anterior cingulate appear to be most involved, and, to a lesser degree, the posterior lateral temporal cortex. The correlation patterns were different in patients with nonleft dominance for language, where right memory performance correlated with right temporal metabolism. Our findings are consistent with the growing evidence that extrahippocampal temporal and frontal cortices are involved in memory processing in addition to medial temporal structures. We found that these regions are involved to a different degree in each hemisphere. The lack of correlations between [18F]FDG uptake and FSIQ, VIQ, or PIQ suggests that our findings may be specific to the memory domain.

Acknowledgments

Dr. Akanuma was supported in part by Guy's and St Thomas' Charitable Foundation (Project Grant R011201).

Reprint requests should be sent to Nozomi Akanuma, South London & Maudsley NHS Foundation Trust, 308 Brixton Road, London SW9 6AA, UK, or via e-mail: Nozomi.Akanuma@slam.nhs.uk.

REFERENCES

Akanuma
,
N.
,
Alarcon
,
G.
,
Lum
,
F.
,
Kissani
,
N.
,
Koutroumanidis
,
M.
,
Adachi
N.
,
et al
(
2003a
).
Lateralising value of neuropsychological protocols for presurgical assessment of temporal lobe epilepsy.
Epilepsia
,
44
,
408
418
.
Akanuma
,
N.
,
Koutroumanidis
,
M.
,
Adachi
,
N.
,
Alarcon
,
G.
, &
Binnie
,
C. D.
(
2003b
).
Presurgical assessment of memory-related brain structures: The Wada test and functional neuroimaging.
Seizure
,
12
,
346
358
.
Alarcon
,
G.
,
Valentin
,
A.
,
Watt C., Selway
,
R. P.
,
Lacruz
,
M. E.
,
Elwes
,
R. D.
,
et al
(
2006
).
Is it worth pursuing surgery for epilepsy in patients with normal neuroimaging?
Journal of Neurology, Neurosurgery and Psychiatry
,
77
,
474
480
.
Arnold
,
S.
,
Schlaug
,
G.
,
Niemann
,
H.
,
Ebner
,
A.
,
Luders
,
H.
,
Witte
,
O. W.
,
et al
(
1996
).
Topography of interictal glucose hypometabolism in unilateral mesiotemporal epilepsy.
Neurology
,
46
,
1422
1430
.
Aupee
,
A. M.
,
Desgranges
,
B.
,
Eustache
,
F.
,
Lalevee
,
C.
,
de la Sayette
,
V.
,
Viader
,
F.
,
et al
(
2001
).
Voxel-based mapping of brain hypometabolism in permanent amnesia with PET.
Neuroimage
,
13
,
1164
1173
.
Awh
,
E.
, &
Jonides
,
J.
(
1998
).
Spatial working memory and spatial selective attention.
In R. Parsuraman (Eds.),
The attentive brain
(pp.
353
380
).
Cambridge
:
MIT Press
.
Barrington
,
S. F.
,
Koutroumanidis
,
M.
,
Agathonikou
,
A.
,
Marsden
,
P. K.
,
Binnie
,
C. D.
,
Polkey
,
C. E.
,
et al
(
1998
).
Clinical value of “ictal” FDG-positron emission tomography and the routine use of simultaneous scalp EEG studies in patients with intractable partial epilepsies.
Epilepsia
,
39
,
753
766
.
Baxendale
,
S. A.
,
Van Paesschen
,
W.
,
Thompson
,
P. J.
,
Duncan
,
J. S.
,
Shorvon
,
S. D.
, &
Connelly
,
A.
(
1997
).
The relation between quantitative MRI measures of hippocampal structure and the intracarotid amobarbital test.
Epilepsia
,
38
,
998
1007
.
Brazdil
,
M.
,
Zakopcan
,
J.
,
Kuba
,
R.
,
Fanfrdlova
,
Z.
, &
Rektor
,
I.
(
2003
).
Atypical hemispheric language dominance in left temporal lobe epilepsy as a result of the reorganization of language functions.
Epilepsy & Behavior
,
4
,
414
419
.
Buckner
,
R. L.
,
Kelley
,
W. M.
, &
Petersen
,
S. E.
(
1999
).
Frontal cortex contributes to human memory formation.
Nature Neuroscience
,
2
,
311
314
.
Davies
,
K. G.
,
Hermann
,
B. P.
, &
Foley
,
K. T.
(
1996
).
Relation between intracarotid amobarbital memory asymmetry scores and hippocampal sclerosis in patients undergoing anterior temporal lobe resections.
Epilepsia
,
37
,
522
525
.
Deblaere
,
K.
,
Backes
,
W. H.
,
Tieleman
,
A.
,
Vandemaele
,
P.
,
Defreyne
,
L.
,
Vonck
,
K.
,
et al
(
2005
).
Lateralized anterior mesiotemporal lobe activation: Semirandom functional MR imaging encoding paradigm in patients with temporal lobe epilepsy—Initial experience.
Radiology
,
236
,
996
1003
.
Desgranges
,
B.
,
Baron
,
J. C.
,
de la Sayette
,
V.
,
Petit-Taboue
,
M. C.
,
Benali
,
K.
,
Landeau
,
B.
,
et al
(
1998
).
The neural substrates of memory systems impairment in Alzheimer's disease. A PET study of resting brain glucose utilization.
Brain
,
121
,
611
631
.
Detre
,
J. A.
,
Maccotta
,
L.
,
King
,
D.
,
Alsop
,
D. C.
,
Glosser
,
G.
,
D'Esposito
,
M.
,
et al
(
1998
).
Functional MRI lateralization of memory in temporal lobe epilepsy.
Neurology
,
50
,
926
932
.
Duncan
,
J.
, &
Owen
,
A. M.
(
2000
).
Common regions of the human frontal lobe recruited by diverse cognitive demands.
Trends in Neurosciences
,
23
,
475
483
.
Dupont
,
S.
,
Samson
,
Y.
,
Van de Moortele
,
P. F.
,
Samson
,
S.
,
Poline
,
J. P.
,
Hasboun
,
D.
,
et al
(
2002
).
Bilateral hemispheric alteration of memory processes in right medial temporal lobe epilepsy.
Journal of Neurology, Neurosurgery and Psychiatry
,
73
,
478
485
.
Dupont
,
S.
,
Van de Moortele, PF., Samson
,
S.
,
Hasboun
,
D.
,
Poline
,
J.
,
Adam
,
C.
,
et al
(
2000
).
Episodic memory in left temporal lobe epilepsy: A functional MRI study.
Brain
,
123
,
1722
1732
.
Eichenbaum
,
H.
(
2000
).
A cortical–hippocampal system for declarative memory.
Nature Reviews Neuroscience
,
1
,
41
50
.
Fazio
,
F.
,
Perani
,
D.
,
Gilardi
,
M. C.
,
Colombo
,
F.
,
Cappa
,
S. F.
,
Vallar
,
G.
,
et al
(
1992
).
Metabolic impairment in human amnesia: A PET study of memory networks.
Journal of Cerebral Blood Flow and Metabolism
,
12
,
353
358
.
Ferrier
,
C. H.
,
Alarcon
,
G.
,
Glover
,
A.
,
Koutroumanidis
,
M.
,
Morris
,
R. G.
,
Simmons
,
A.
,
et al
(
2000
).
N-Acetylaspartate and creatine levels measured by (1)H MRS relate to recognition memory.
Neurology
,
55
,
1874
1883
.
Fiez
,
J. A.
, &
Petersen
,
S. E.
(
1998
).
Neuroimaging studies of word reading.
Proceedings of the National Academy of Sciences, U.S.A.
,
95
,
914
921
.
Fletcher
,
P. C.
, &
Henson
,
R. N.
(
2001
).
Frontal lobes and human memory: Insights from functional neuroimaging.
Brain
,
124
,
849
881
.
Frey
,
S.
, &
Petrides
,
M.
(
2000
).
Orbitofrontal cortex: A key prefrontal region for encoding information.
Proceedings of the National Academy of Sciences, U.S.A.
,
97
,
8723
8727
.
Gabrieli
,
J. D.
,
Brewer
,
J. B.
,
Desmond
,
J. E.
, &
Glover
,
G. H.
(
1997
).
Separate neural bases of two fundamental memory processes in the human medial temporal lobe.
Science
,
276
,
264
266
.
Golby
,
A.
,
Poldrack
,
R.
,
Illes
,
J.
,
Chen
,
D.
,
Desmond
,
J.
, &
Gabrieli
,
J.
(
2002
).
Memory lateralization in medial temporal lobe epilepsy assessed by functional MRI.
Epilepsia
,
43
,
855
863
.
Goldstein
,
L. H.
, &
Polkey
,
C. E.
(
1993
).
Short-term cognitive changes after unilateral temporal lobectomy or unilateral amygdalo-hippocampectomy for the relief of temporal lobe epilepsy.
Journal of Neurology, Neurosurgery and Psychiatry
,
56
,
135
140
.
Greicius
,
M. D.
,
Krasnow
,
B.
,
Reiss
,
A. L.
, &
Menon
,
V.
(
2003
).
Functional connectivity in the resting brain: A network analysis of the default mode hypothesis.
Proceedings of the National Academy of Sciences, U.S.A.
,
100
,
253
258
.
Habib
,
R.
,
Nyberg
,
L.
, &
Tulving
,
E.
(
2003
).
Hemispheric asymmetries of memory: The HERA model revisited.
Trends in Cognitive Sciences
,
7
,
241
245
.
Hammers
,
A.
,
Koepp
,
M. J.
,
Hurlemann
,
R.
,
Thom
,
M.
,
Richardson
,
M. P.
,
Brooks
,
D.
,
et al
(
2002
).
Abnormalities of grey and white matter [11C]flumazenil binding in temporal lobe epilepsy with normal MRI.
Brain
,
125
,
2257
2271
.
Hong
,
S. B.
,
Roh
,
S. Y.
,
Kim
,
S. E.
, &
Seo
,
D. W.
(
2000
).
Correlation of temporal lobe glucose metabolism with the Wada memory test.
Epilepsia
,
41
,
1554
1559
.
Incisa della Rocchetta
,
A.
, &
Milner
,
B.
(
1993
).
Strategic search and retrieval inhibition: The role of the frontal lobes.
Neuropsychologia
,
31
,
503
524
.
Janowsky
,
J. S.
,
Shimamura
,
A. P.
,
Kritchevsky
,
M.
, &
Squire
,
L. R.
(
1989
).
Cognitive impairment following frontal lobe damage and its relevance to human amnesia.
Behavioral Neuroscience
,
103
,
548
560
.
Kelley
,
W. M.
,
Miezin
,
F. M.
,
McDermott
,
K. B.
,
Buckner
,
R. L.
,
Raichle
,
M. E.
,
Cohen
,
N. J.
,
et al
(
1998
).
Hemispheric specialization in human dorsal frontal cortex and medial temporal lobe for verbal and nonverbal memory encoding.
Neuron
,
20
,
927
936
.
Kelley
,
W. M.
,
Ojemann
,
J. G.
,
Wetzel
,
R. D.
,
Derdeyn
,
C. P.
,
Moran
,
C. J.
,
Cross
,
D. T.
,
et al
(
2002
).
Wada testing reveals frontal lateralization for the memorization of words and faces.
Journal of Cognitive Neuroscience
,
14
,
116
125
.
Kennedy
,
A. M.
,
Newman
,
S. K.
,
Frackowiak
,
R. S.
,
Cunningham
,
V. J.
,
Roques
,
P.
,
Stevens
,
J.
,
et al
(
1995
).
Chromosome 14 linked familial Alzheimer's disease. A clinico-pathological study of a single pedigree.
Brain
,
118
,
185
205
.
Kirchhoff
,
B. A.
,
Wagner
,
A. D.
,
Maril
,
A.
, &
Stern
,
C. E.
(
2000
).
Prefrontal–temporal circuitry for episodic encoding and subsequent memory.
Journal of Neuroscience
,
20
,
6173
6180
.
Kopelman
,
M. D.
(
2002
).
Disorders of memory.
Brain
,
125
,
2152
2190
.
Kopelman
,
M. D.
,
Stevens
,
T. G.
,
Foli
,
S.
, &
Grasby
,
P.
(
1998
).
PET activation of the medial temporal lobe in learning.
Brain
,
121
,
875
887
.
Koutroumanidis
,
M.
,
Binnie
,
C. D.
,
Elwes
,
R. D.
,
Polkey
,
C. E.
,
Seed
,
P.
,
Alarcon
,
G.
,
et al
(
1998
).
Interictal regional slow activity in temporal lobe epilepsy correlates with lateral temporal hypometabolism as imaged with 18FDG PET: Neurophysiological and metabolic implications.
Journal of Neurology, Neurosurgery and Psychiatry
,
65
,
170
176
.
Koutroumanidis
,
M.
,
Hennessy
,
M. J.
,
Seed
,
P. T.
,
Elwes
,
R. D.
,
Jarosz
,
J.
,
Morris
,
R. G.
,
et al
(
2000
).
Significance of interictal bilateral temporal hypometabolism in temporal lobe epilepsy.
Neurology
,
54
,
1811
1821
.
Lacruz
,
M. E.
,
Alarcon
,
G.
,
Akanuma
,
N.
,
Lum
,
F. C.
,
Kissani
,
N.
,
Koutroumanidis
,
M.
,
et al
(
2004
).
Neuropsychological effects associated with temporal lobectomy and amygdalohippocampectomy depending on Wada test failure.
Journal of Neurology, Neurosurgery and Psychiatry
,
75
,
600
607
.
Loring
,
D. W.
,
Murro
,
A. M.
,
Meador
,
K. J.
,
Lee
,
G. P.
,
Gratton
,
C. A.
,
Nichols
,
M. E.
,
et al
(
1993
).
Wada memory testing and hippocampal volume measurements in the evaluation for temporal lobectomy.
Neurology
,
43
,
1789
1793
.
Mazoyer
,
B.
,
Zago
,
L.
,
Mellet
,
E.
,
Bricogne
,
S.
,
Etard
,
O.
,
Houdé
,
O.
,
et al
(
2001
).
Cortical networks for working memory and executive functions sustain the conscious resting state in man.
Brain Research Bulletin
,
54
,
287
298
.
McDermott
,
K. B.
,
Buckner
,
R. L.
,
Petersen
,
S. E.
,
Kelley
,
W. M.
, &
Sanders
,
A. L.
(
1999
).
Set- and code-specific activation in frontal cortex: An fMRI study of encoding and retrieval of faces and words.
Journal of Cognitive Neuroscience
,
11
,
631
640
.
Meunier
,
M.
,
Bachevalier
,
J.
, &
Mishkin
,
M.
(
1997
).
Effects of orbital frontal and anterior cingulate lesions on object and spatial memory in rhesus monkeys.
Neuropsychologia
,
35
,
999
1015
.
Milner
,
B.
,
Branch
,
C.
, &
Rasmussen
,
T.
(
1962
).
Study of short-term memory after intracarotid injection of sodium amytal.
Transactions of the American Neurological Association
,
87
,
224
226
.
Morris
,
R. G.
, &
Mayes
,
A. R.
(
2004
).
Long-term spatial memory: Introduction and guide to the special section.
Neuropsychology
,
18
,
403
404
.
Morris
,
R. G.
,
Nunn
,
J.
,
Abrahams
,
S.
,
Feigenbaum
,
J. D.
, &
Recce
,
M.
(
1999
).
The hippocampus and spatial memory in humans.
In N. Burgess, K. Jeffrey, & J. O'Keefe (Eds.),
Spatial functions of the hippocampal formation and parietal cortex
(pp.
259
289
).
Oxford
:
Oxford University Press
.
Morton
,
N.
,
Polkey
,
C. E.
,
Cox
,
T.
, &
Morris
,
R. G.
(
1996
).
Episodic memory dysfunction during sodium amytal testing of epileptic patients in relation to posterior cerebral artery perfusion.
Journal of Clinical and Experimental Neuropsychology
,
18
,
24
37
.
Owen
,
A. M.
(
2003
).
HERA today, gone tomorrow?
Trends in Cognitive Sciences
,
7
,
383
384
.
Park
,
H. J.
,
Levitt
,
J.
,
Shenton
,
M. E.
,
Salisbury
,
D. F.
,
Kubicki
,
M.
,
Kikinis
,
R.
,
et al
(
2004
).
An MRI study of spatial probability brain map differences between first-episode schizophrenia and normal controls.
Neuroimage
,
22
,
1231
1246
.
Pauli
,
E.
,
Pickel
,
S.
,
Schulemann
,
H.
,
Buchfelder
,
M.
, &
Stefan
,
H.
(
1999
).
Neuropsychologic findings depending on the type of the resection in temporal lobe epilepsy.
Advances in Neurology
,
81
,
371
377
.
Petrides
,
M.
,
Alivisatos
,
B.
,
Evans
,
A. C.
, &
Meyer
,
E.
(
1993
).
Dissociation of human mid-dorsolateral from posterior dorsolateral frontal cortex in memory processing.
Proceedings of the National Academy of Sciences, U.S.A.
,
90
,
873
877
.
Ploner
,
C. J.
,
Gaymard
,
B. M.
,
Ehrle
,
N.
,
Rivaud-Pechoux
,
S.
,
Baulac
,
M.
,
Brandt
,
S. A.
,
et al
(
1999
).
Spatial memory deficits in patients with lesions affecting the medial temporal neocortex.
Annals of Neurology
,
45
,
312
319
.
Poldrack
,
R. A.
,
Wagner
,
A. D.
,
Prull
,
M. W.
,
Desmond
,
J. E.
,
Glover
,
G. H.
, &
Gabrieli
,
J. D.
(
1999
).
Functional specialization for semantic and phonological processing in the left inferior prefrontal cortex.
Neuroimage
,
10
,
15
35
.
Rausch
,
R.
,
Babb
,
T. L.
,
Engel
,
J.
, Jr., &
Crandall
,
P. H.
(
1989
).
Memory following intracarotid amobarbital injection contralateral to hippocampal damage.
Archives of Neurology
,
46
,
783
788
.
Rausch
,
R.
,
Henry
,
T. R.
,
Ary
,
C. M.
,
Engel
,
J.
, Jr., &
Mazziotta
,
J.
(
1994
).
Asymmetric interictal glucose hypometabolism and cognitive performance in epileptic patients.
Archives of Neurology
,
51
,
139
144
.
Reed
,
L. J.
,
Lasserson
,
D.
,
Marsden
,
P.
,
Bright
,
P.
,
Stanhope
,
N.
, &
Kopelman
,
M. D.
(
2005
).
Correlations of regional cerebral metabolism with memory performance and executive function in patients with herpes encephalitis or frontal lobe lesions.
Neuropsychology
,
19
,
555
565
.
Reed
,
L. J.
,
Lasserson
,
D.
,
Marsden
,
P.
,
Stanhope
,
N.
,
Stevens
,
T.
,
Bello
,
F.
,
et al
(
2003
).
FDG-PET findings in the Wernicke–Korsakoff syndrome.
Cortex
,
39
,
1027
1045
.
Reed
,
L. J.
,
Marsden
,
P.
,
Lasserson
,
D.
,
Sheldon
,
N.
,
Lewis
,
P.
,
Stanhope
,
N.
,
et al
(
1999
).
Memory
,
7
,
599
612
.
Richardson
,
M.
,
Strange
,
B.
,
Duncan
,
J.
,
Dolan
,
R.
(
2003
).
Preserved verbal memory function in left medial temporal pathology involves reorganisation of function to right medial temporal lobe.
Neuroimage
,
20(Suppl. 1)
,
S112
S119
.
Rossi
,
S.
,
Cappa
,
S.
,
Babiloni
,
C.
,
Pasqualetti
,
P.
,
Miniussi
,
C.
,
Carducci
,
F.
,
et al
(
2001
).
Prefrontal cortex in long-term memory: An “interference” approach using magnetic stimulation.
Nature Neuroscience
,
4
,
948
952
.
Rugg, M. D, Fletcher
,
P. C.
,
Frith
,
C. D.
,
Frackowiak
,
R. S.
, &
Dolan
,
R. J.
(
1996
).
Differential activation of the prefrontal cortex in successful and unsuccessful memory retrieval.
Brain
,
119
,
2073
2083
.
Salanova
,
V.
,
Markand
,
O.
, &
Worth
,
R.
(
2001
).
Focal functional deficits in temporal lobe epilepsy on PET scans and the intracarotid amobarbital procedure: Comparison of patients with unitemporal epilepsy with those requiring intracranial recordings.
Epilepsia
,
42
,
198
203
.
Salanova
,
V.
,
Morris
,
H. H.
, III.,
Rehm
,
P.
,
Wyllie
,
E.
,
Dinner
,
D. S.
,
Luders
,
H.
,
et al
(
1992
).
Comparison of the intracarotid amobarbital procedure and interictal cerebral 18-fluorodeoxyglucose positron emission tomography scans in refractory temporal lobe epilepsy.
Epilepsia
,
33
,
635
638
.
Sass
,
K. J.
,
Spencer
,
D. D.
,
Kim
,
J. H.
,
Westerveld
,
M.
,
Novelly
,
R. A.
, &
Lencz
,
T.
(
1990
).
Verbal memory impairment correlates with hippocampal pyramidal cell density.
Neurology
,
40
,
1694
1697
.
Schacter
,
D. L.
(
1987
).
Implicit expressions of memory in organic amnesia: Learning of new facts and associations.
Human Neurobiology
,
6
,
107
118
.
Scoville
,
W. B.
, &
Milner
,
B.
(
1957
).
Loss of recent memory after bilateral hippocampal lesions.
Journal of Neurology, Neurosurgery and Psychiatry
,
20
,
11
21
.
Shallice
,
T.
,
Fletcher
,
P.
,
Frith
,
C. D.
,
Grasby
,
P.
,
Frackowiak
,
R. S.
, &
Dolan
,
R. J.
(
1994
).
Brain regions associated with acquisition and retrieval of verbal episodic memory.
Nature
,
368
,
633
635
.
Signorini
,
M.
,
Paulesu
,
E.
,
Friston
,
K.
,
Perani
,
D.
,
Colleluori
,
A.
,
Lucignani
,
G.
,
et al
(
1999
).
Rapid assessment of regional cerebral metabolic abnormalities in single subjects with quantitative and nonquantitative [18F]FDG PET: A clinical validation of statistical parametric mapping.
Neuroimage
,
9
,
63
80
.
Smith
,
M. L.
(
1989
).
Memory disorders associated with temporal lobe lesions.
In L. R. Squire & G. Gainotti (Eds.),
Handbook of neuropsychology
(Vol.
3
, pp.
91
106
).
Amsterdam
:
Elsevier
.
Squire
,
L. R.
(
1992
).
Memory and the hippocampus: A synthesis from findings with rats, monkeys, and humans.
Psychological Review
,
99
,
195
231
.
Squire
,
L. R.
,
Stark
,
C. E.
, &
Clark
,
R. E.
(
2004
).
The medial temporal lobe.
Annual Review of Neuroscience
,
27
,
279
306
.
Swartz
,
B. E.
,
Thomas
,
K.
,
Simpkins
,
F.
,
Kovalik
,
E.
, &
Mandelkern
,
M. M.
(
1999
).
Rapid quantitative analysis of individual 18FDG-PET scans.
Clinical Positron Imaging
,
2
,
47
56
.
Talairach
,
J.
, &
Tournoux
,
P.
(
1988
).
Co-planar streotaxic atlas of the human brain.
New York
:
Thieme
.
Tulving
,
E.
,
Kapur
,
S.
,
Craik
,
F. I.
,
Moscovitch
,
M.
, &
Houle
,
S.
(
1994
).
Hemispheric encoding/retrieval asymmetry in episodic memory: Positron emission tomography findings.
Proceedings of the National Academy of Sciences, U.S.A.
,
91
,
2016
2020
.
Turner
,
M. S.
,
Cipolotti
,
L.
,
Yousry
,
T.
, &
Shallice
,
T.
(
2007
).
Qualitatively different memory impairments across frontal lobe subgroups.
Neuropsychologia
,
45
,
1540
1552
.
Wada
,
J. A.
, &
Rasmussen
,
T.
(
1960
).
Intracarotid injection of sodium amytal for the lateralization of cerebral speech dominance: Experimental and clinical observations.
Journal of Neurosurgery
,
17
,
266
282
.
Wager
,
T. D.
, &
Smith
,
E. E.
(
2003
).
Neuroimaging studies of working memory: A meta-analysis.
Cognitive, Affective & Behavioral Neuroscience
,
3
,
255
274
.
Wagner
,
A. D.
(
1999
).
Working memory contributions to human learning and remembering.
Neuron
,
22
,
19
22
.
Wagner
,
A. D.
,
Poldrack
,
R. A.
,
Eldridge
,
L. L.
,
Desmond
,
J. E.
,
Glover
,
G. H.
, &
Gabrieli
,
J. D.
(
1998
).
Material-specific lateralization of prefrontal activation during episodic encoding and retrieval.
NeuroReport
,
9
,
3711
3717
.
Weintrob
,
D. L.
,
Saling
,
M. M.
,
Berkovic
,
S. F.
,
Berlangieri
,
S. U.
, &
Reutens
,
D. C.
(
2002
).
Verbal memory in left temporal lobe epilepsy: Evidence for task-related localization.
Annals of Neurology
,
51
,
442
447
.
Wheeler
,
M. A.
,
Stuss
,
D. T.
, &
Tulving
,
E.
(
1995
).
Frontal lobe damage produces episodic memory impairment.
Journal of the International Neuropsychological Society
,
1
,
525
536
.
Wieser
,
H. G.
,
Engel
,
J.
, Jr.,
Williamson
,
P. D.
,
Babb
,
T. L.
, &
Goldman
,
S.
(
1993
).
Surgically remediable temporal lobe syndromes.
In J. Engel, Jr. (Ed.),
Surgical treatment of the epilepsies
(pp.
49
63
).
New York
:
Raven
.
Wieser
,
H. G.
, &
Yasargil
,
M. G.
(
1982
).
Selective amygdalohippocampectomy as a surgical treatment of mesiobasal limbic epilepsy.
Surgical Neurology
,
17
,
445
457
.