Memory functioning in patients with unilateral temporal lobe epilepsy: Neuroimaging indicators of functional

integrity in the hippocampus and beyond.

by

Alexander Barnett

A thesis submitted in conformity with the requirements for the degree of Master of Arts

Department of Psychology University of Toronto

© Copyright by Alexander Barnett 2012

ii

Memory functioning in patients with unilateral temporal lobe

epilepsy: Neuroimaging indicators of functional integrity in the

hippocampus and beyond.

Alexander Barnett

Master of Arts

Department of Psychology

University of Toronto

2012

Abstract

Temporal lobe epilepsy (TLE) is a common form of intractable epilepsy that can be treated with

surgical resection of the epileptogenic medial temporal lobe tissue, specifically the hippocampus.

This resection can lead to a variable degree of memory deficit and considerable research has

been directed at identifying predictors of these deficits. This thesis explores the relationship

between structural predictors and functional predictors in TLE. I looked at fMRI activation

asymmetry produced by a scene encoding task as well as volume asymmetry ratios within the

hippocampus and the relationship of these predictors to memory performance in patients with

TLE. Mediation analysis was performed according to Baron and Kenny (1986) and showed that

fMRI activation asymmetry mediated the relationship between volume asymmetry and memory

asymmetry in patients with TLE. This suggests that activation asymmetry may be a preferred

variable for assessing functional adequacy in the medial temporal region.

iii

Acknowledgments

First and foremost I would like to acknowledge Dr. Mary Pat McAndrews for all of her

mentorship and guidance throughout this thesis. I would also like to acknowledge Dr. Andy Lee

for his helpful comments and training that helped me better understand MTL anatomy. Finally, I

would like to acknowledge my lab mates Melanie Cohn, Irene Giannoylis, Conny McCormick,

Areeba Adnan, Marie St-Laurent and Massih Moayedi for their assistance in scanning patients

and with assistance performing analysis.

iv

Table of Contents

Chapter 1: Introduction 1

1 Predictors of Decline 3

1.1 Neuropsychological Testing 3

1.2 Neuroimaging 4

1.3 Multimodal 6

Chapter 2: Methods 10

1 Participants 10

2 Neuropsychological Testing 12

3 fMRI Data Acquisition 12

4 fMRI Task 13

5 Functional Data Processing 13

6 Data Analysis 14

7 Structural Data Analysis 16

8 Group Comparison 16

9 Neuroimaging Measures and Behaviour 17

Chapter 3: Results 19

1 Demographics 19

2 Task Activation 19

3 Asymmetry Ratios 22

4 Neuroimaging Measures and Behaviour 22

Chapter 4: Discussion 26

v

List of Tables Table 1 11

Table 2 24

vi

List of Figures Figure 1 15

Figure 2 18

Figure 3 20

Figure 4 21

Figure 5 23

Figure 6 25

Figure 7 27

Chapter 1 Introduction

Epilepsy is a prevalent neurological disorder. In 2001 the World Health Organization estimated

that approximately 50 million people worldwide were affected by epilepsy (Levav & Rutz,

2002). Temporal lobe epilepsy (TLE) specifically is one of the most common presentations of

this disorder (Bell, Lin, Seidenberg, & Hermann, 2011). Epilepsy is characterized by recurrent

seizures, but also involves a profile of cognitive deficits, the most apparent being memory.

Medication is prescribed to reduce the frequency of seizures, but within this population can

sometimes be ineffective (Kwan & Brodie, 2000). Surgical interventions have been able to

successfully eliminate or greatly reduce the occurrence of seizures in patients with TLE (Engel,

1993; Engel et al., 2003). This surgery involves the resection of the varying amounts of anterior

portion of the hippocampus and the amygdala. Despite the overwhelming success of these

surgical procedures, post-surgical memory declines are very common (Gleissner, Helmstaedter,

Schramm, & Elger, 2004; Hermann, Seidenberg, Haltiner, & Wyler, 1995; Martin et al., 1998).

The canonical example of this effect is the case of Henry Molaison, who was unable to form any

new long term memory following bilateral resection of the anterior hippocampus (Scoville &

Milner, 1957). Since surgical treatment involves the resection of critical memory structures,

particularly the hippocampus, post-surgical memory decline is highly prevalent and variable

(Bell & Davies, 1998; Chelune, 1995). The case of Henry Molaison launched an extensive body

of research investigating the relationship between the hippocampus and memory as well the

investigation of predictors of post-surgical memory change.

As part of standard procedure in epilepsy surgery centres, all patients undergo an

extensive battery of neuropsychological testing pre- and post-surgically which allows clinicians

to understand the specific deficits of this patient population. This battery of testing also provides

2

insight to determining laterality of functional deficits as these tests have been shown to correlate

with areas of brain dysfunction and seizure focus (Hermann, Wyler, Somes, Berry, & Dohan,

1992; Jones-Gotman et al., 2010; McAndrews & Cohn, 2012; Saling et al., 1993). Patients with

left unilateral TLE (LTLE) tend to have poor verbal memory performance presurgically

compared to healthy controls whereas patients with right unilateral TLE (RTLE) tend to show

poor visual memory compared to healthy controls (Jones-Gotman et al., 2010; McAndrews &

Cohn, 2012). Similar trends were initially observed in post-surgical neuropsychological testing,

revealing that surgical resection of the left anterior hippocampus and amygdala lead to verbal

memory decline, while resection of the right is more related to visual memory decline, though

the evidence is less robust for RTLE (Bell & Davies, 1998; R C Martin et al., 1998; Milner,

1974; Morris, Abrahams, & Polkey, 1995). This led to the material specificity theory, which

states that the left hippocampus (in individuals with left-language dominance, which is the

majority of the population) is responsible for verbal memory while the right hippocampus is

responsible for visual memory (Milner, 1968). This theory is considered to be one of the

foundations of this field of research and has been looked at in terms of behavioural

neuropsychology measures and neuroimaging (Delaney, Rosen, Mattson, & Novelly, 1980; U

Gleissner, Helmstaedter, & Elger, 1998; Golby et al., 2001; Hermann, Wyler, Richey, & Rea,

1987; R C Martin et al., 2001).

Anatomical pathology has also been characterized within this population by examining

excised and post-mortem hippocampi taken from patients with TLE. Patients with TLE are very

likely to show signs of hippocampal sclerosis (HS), which is characterized by cell loss and

synaptic reorganization in the hippocampus and is the most common lesion in patients with TLE

(Blümcke, Thom, & Wiestler, 2002; Shamim et al., 2004). This lesion has been shown to be

related to memory deficits, with HS associated with poorer memory scores (Hermann et al.,

3

1992; Kneebone, Lee, Wade, & Loring, 2007; Marques et al., 2007; Saling et al., 1993; Trenerry

et al., 1993) and, with the advent of MRI, HS is typically diagnosed pre-operatively (Berkovic et

al., 1991). Of importance, structural neuroimaging studies using MRI for region of interest (ROI)

volumetry and voxel-based morphometry have shown atrophy not only in the hippocampus, but

also in a broader extent of medial temporal lobe including parahippocampal gyrus, fornix,

amygdala and entorhinal cortex (Bernasconi, 2003; Bernasconi et al., 1999, 2001; Kuzniecky et

al., 1999). The literature focuses heavily upon hippocampal changes in patients with TLE, but

there are clearly structural differences that occur in the MTL.

1 Predictors of Memory Decline

1.1 Neuropsychological Testing

Since the variability of memory decline is so high, researchers and clinicians have focused on

identifying factors that affect or predict the degree of memory decline due to surgery (Bell et al.,

2011). One of the first trends noted was that there is an inverse relationship between preoperative

memory functioning and postoperative memory change, with higher pre-operative scores

associated with greater post-operative declines (Chelune, Naugle, Lüders, & Awad, 1991;

Elshorst et al., 2009; Helmstaedter & Elger, 1996; Hermann et al., 1995). Chelune, (1995)

looked at a sample of 181 patients with TLE and observed that higher verbal memory

presurgically on the Wechsler Memory Scale Revised (Wechsler, 1987) resulted in greater

postsurgical decline of verbal memory in patients LTLE. They also found that that higher visual

memory presurgically resulted in greater postsurgical decline of visual memory in patients with

RTLE. Chelune (1995) explains this by reasoning that if a patient has high presurgical

functioning they may have a greater amount of functioning to lose after surgical resection.

4

1.2 Neuroimaging

Neuroimaging advances provide additional opportunities for predictive markers of post-surgical

memory decline. Structural MRI has allowed us to examine hippocampal atrophy. As discussed

earlier, hippocampal sclerosis has been shown to relate to cognitive deficits (Hermann et al.,

1992; Marques et al., 2007; Saling et al., 1993) and thus MRI measures of presurgical structural

integrity of the hippocampus may relate to the post-surgical outcome of memory. Several studies

have linked hippocampal volume in patients with TLE to post-surgical memory change

(Kneebone et al., 2007; Martin et al., 2001; Martin et al., 2002; Mechanic-Hamilton et al., 2009).

The results indicate that larger ipsilateral volumes, indicating a higher degree of intact tissue, are

related to greater declines in memory post-surgery. These results are often best seen using

asymmetry ratios (AR) [contralateral volume – ipsilateral volume] / [contralateral volume +

ipsilateral volume] since ARs tend to provide a greater spread separating patients with LTLE and

RTLE (Mechanic-Hamilton et al., 2009). The understanding here is that if the epileptogenic

hippocampus has little atrophy then it is more likely to support memory processes as it does in

the healthy brain. Therefore, removal of a more intact and useful hippocampal piece will result in

greater decline.

Functional MRI (fMRI) has allowed researchers to examine the functional integrity of the

to-be-resected tissue. Studies have attempted to use memory tasks during fMRI to allow for

quantification of hippocampal activity. Analysis of this activity, which is taken as a measure of

functional integrity, has been linked with memory changes following surgery (Bonelli et al.,

2010; Mechanic-Hamilton et al., 2009; Rabin et al., 2004; Richardson et al., 2004; Richardson,

Strange, Duncan, & Dolan, 2006). Rabin et al. (2004) examined 35 unilateral TLE patients (20

RTLE, 15 LTLE) and 30 healthy controls. This group used a scene encoding task (Stern et al.,

1996) to elicit medial temporal lobe activation in participants. All subjects were shown a series

5

of novel visual scenes in block design with a control scene that was degraded and retiled.

Contrasts between novel scenes and the control block were created for each individual and

activation was extracted from the hippocampus in the left and right hemispheres. Activation was

used to make an asymmetry ratio using the formula [(Lactivity- Ractivity)/(Lactivity+Ractivity)]. This

method was also used for a larger ROI encompassing the hippocampus, parahippocampus and

fusiform (HPF ROI). The ARs created from the larger HPF ROI were shown to significantly

correlate with post-surgical change in visual memory in patients with RTLE such that patients

having more right activation compared to left activation showed greater post-surgical decline.

The authors did not find significant correlations in their LTLE patients however and attribute this

to the scene encoding potentially being more sensitive to right hemisphere dysfunction.

Richardson et al. (2004) examined 10 LTLE patients presurgically on a verbal encoding

task in which patients were presented with a list of words and asked to make “living” or

“nonliving” judgments on the words, though not asked to memorize the words in the scanner.

Following the scan patients were given a surprise memory test on the words and asked to make

recognition judgments on the words. This judgment involved subjects giving a recollection

response (R), a familiarity response (K), or a new response (N) as described by Tulving (1985).

Recognition responses that are associated with recollection responses specifically activate the

hippocampus compared to recognition responses associated with familiarity (Eldridge,

Knowlton, Furmanski, Bookheimer, & Engel, 2000). They then contrasted the words for each

subject labeled correctly as R against the words labeled correctly as K. Activation from this

contrast was then used to create “encoding asymmetry” indices for each subject by subtracting

left activation from right activation using a small volume correction on the hippocampus. The

researchers then were able to find that this presurgical encoding asymmetry was able to

significantly predict post-surgical memory decline. They found that the greater the activation in

6

the left compared to the right hippocampus was linked with larger declines in memory following

surgical resection of the anterior portion of the left hippocampus in these patients.

Much of the data relating presurgical measures to predicting memory outcome support a

model called the functional adequacy theory. This theory posits that the amount of memory

decline seen in patients with TLE postsurgery is dependent upon the presurgical integrity of the

to-be-resected tissue. If the tissue is intact, and functionally involved in memory processes, then

removal of this tissue will presumably cause a larger reduction in memory functioning than if the

tissue was atrophied and not functionally involved in memory (Chelune, 1995).

1.3 Multimodal

Several studies have attempted to use multimodal methods for predicting post-surgical memory

change in patients with TLE. Elshorst et al.(2009) looked at presurgical neuropsychological

memory scores, MRI grading of hippocampal integrity and WADA memory performance as

predictors for post-surgical memory change in 59 patients with LTLE. They used the California

Verbal Learning Test (CVLT) and the Rey Auditory Verbal Learning Test (RAVLT) as

determinants of memory change and as predictors. They also used a ranking for MTS which was

based on a combination of anatomical scans to identify size (T1-weighted) and MR signal

increase (T2-weighted scans) which are indicators of neural atrophy. Finally, patients underwent

an intracarotid amytal procedure (IAP) in which one hemisphere of the brain is anesthetized

while the other hemisphere is given a memory task. This is used to identify the ability of one

hemisphere without the assistance of the anesthetized hemisphere. Each patient was shown a

series of 18 objects to remember on a subsequent recognition test that took place after clearance

of the drug that include the original 18 objects with 36 lures. Scores for each hemisphere were

determined in a pass/fail manner whereby the patient had to recognize at least 60% of the items

encoded at optimal anesthetization. Elshorst et al. (2009) found that presurgical

7

neuropsychological memory best predicted post-surgical memory outcome along with MRI

classification of hippocampal atrophy to explain around 45% of the variance in memory change.

In contrast, memory as assessed via the IAP did no add any predictive power to explaining post-

surgical memory outcome.

Mechanic-Hamilton et al. (2009) looked at the relationship between neuropsychological

functioning and WADA with fMRI as well as examining (in a separate analysis) the predictive

power of hippocampal volume and fMRI activation with memory outcome following surgery.

The study examined 49 patients with TLE (21 LTLE, 24 RTLE, 4 bilateral) and 25 controls.

Participants performed a visual scene encoding task while in the MRI scanner. Scenes were

presented in a block design fashion, alternating a series of novel scenes in the encoding blocks

with a single randomly retiled scene in the control blocks. Participants performed a self-paced

forced choice recognition test following the scan. After surgery, the majority of patients returned

to perform an alternative version of the visual scene encoding task outside the scanner and a pre-

to post- recognition change score was calculated. Activation was calculated by contrasting novel

blocks against the control block. ROI analysis yielded voxel counts for the hippocampus and a

mask that included the hippocampus, parahippocampus cortex and fusiform gyrus (HPF). The

resultant voxel counts were used to create ARs using the formula (voxelcountContra –

voxelcountIpsi) / (voxelcountContra + voxelcountIpsi). These ARs were used as predictors for

changes in clinical neuropsychological measures from pre- to post- surgery. Mechanic-Hamilton

et al. (2009) also used signal change within the ROIs as predictors of changes in clinical

neuropsychological measures.

Hippocampal volumes were calculated by manual tracing using published guidelines

(Watson et al., 1992). Volumes were also used to create ARs. In addition to examining changes

on the scene task, the study included pre- and post-operative measures from clinical memory

8

tests. For verbal memory, tasks included the CVLT and the Logical Memory subtest of the

WMS-III as measures of verbal memory. To measure visuospatial memory, they used the Faces

and visual reproduction subtests of the WMS-III. They calculated change scores for each subtest

by subtracting presurgical scores from postsurgical scores.

This study demonstrated a correlation between fMRI activation in the hippocampus in the

left hemisphere of LTLE patients and memory change determined by scene recognition.

Activation ARs did not however relate to changes in neuropsychological measures of memory

following surgery in either patients with LTLE or those with RTLE. Volume ARs were able to

correlate with several of the neuropsychological measures of memory following surgical

resection. Interestingly and in accord with Elshorst et al (2009), the strongest predictor of post-

surgical memory change was pre-surgical clinical memory scores.

Although these results provide important information regarding predictive values of

preoperative functional and structural measures, the authors did not directly explore a

relationship between functional ARs and volumetric ARs. If volumetric ARs are presumed to

provide us with some index of structural integrity we would expect that this would reflect the

functionality of the tissue, whether it be that ipsilateral atrophy results in hypo-activation of the

afflicted tissue or that ipsilateral atrophy results in a hyper-activation of the afflicted tissue for

compensation. Indeed, there are important reasons to expect that there may be a mediating

relationship between functional and structural integrity such that by understanding how much

tissue is available, and how efficiently the tissue is being used, we can better understand the link

between the brain and memory. In the literature on functional neuroimaging in

neurodegenerative conditions, Dickerson et al. (2004) examined the relationship between clinical

impairment and functional activation in individuals with Mild Cognitive Impairment (MCI).

They found that greater impairment (patients with ‘severe’ rather than ‘milder’ symptoms; all of

9

them non-demented by clinical criteria) was linked with increased MTL activation. Furthermore,

those individuals who activated a larger extent of the MTL showed accelerated cognitive decline.

In a later study the investigators reported that older adults with MCI showed greater MTL

activation compared to controls, whereas those with Alzheimer’s Disease showed significantly

less MTL activation (Dickerson et al., 2005). This evidence was taken to demonstrate that

increased activation may be a compensatory mechanism or might reflect a pathophysiological

mechanism leading towards Alzheimer ’s disease (AD). The authors suggested that there may be

a transitional continuum between normal aging and AD in which individuals with early MCI

show increased activation compared to controls, but as damage accrues in the MTL resulting in

more severe deficits, MTL activation also decreases. Such a relationship between volume,

activation and impairment may exist in TLE and is the driving force behind this study. We,

however, are not looking for a transitional continuum between healthy brain function and

epilepsy since our patient population has intractable epilepsy. Rather, we are looking at the

continuum between patients that have milder deficits to severe deficits in memory domains.

In summary, this study intends to build off of past research to better understand how

neuroimaging measures relate to memory in TLE. As discussed earlier, several studies have

shown that patients with TLE tend to have decreased volume in medial temporal lobe regions

beyond the hippocampus, including parahippocampal gyrus, fornix, amygdala and entorhinal

cortex (Bernasconi et al., 2003; Kuzniecky et al., 1999). Therefore, this thesis will attempt to

look at activation and volume, not solely in the hippocampus but also the parahippocampal

gyrus. Not only will this thesis look at an extended view of the MTL, but also it will examine

both volume and functional activation and how they relate to memory in this clinical population.

With this in mind we will employ the use of a scene encoding task which has been shown to

produce robust activation in the hippocampus and parahippocampal gyrus (Stern et al., 1996)

10

within the MRI scanner to allow for visualization of MTL functionality. We hypothesize that the

functional integrity, as measured by weighted voxel count in our ROIs, will correlate

significantly with presurgical neuropsychological measures of memory. We also hypothesize

that volume measures for the hippocampus and parahippocampal gyrus will correlate with

presurgical neuropsychological measures of memory. Finally, this thesis proposes that measures

of MTL activation mediate the relationship between volume and memory. In the current study

we are using presurgical memory performance so that we may examine the relationship between

volume and activation with memory when the MTL is intact. The implications of understanding

the relationship between functional activation and volume extend beyond TLE and may help

explain phenomenon that occur other disease populations with altered hippocampal functioning

such as MCI, AD, or depression.

Chapter 2 Methods

1 Participants Twenty three patients with pharmacologically intractable unilateral TLE were recruited from the

Neuropsychology and Epilepsy Surgery program at Toronto Western Hospital. Thirteen

presented with RTLE (5 men, 7 women; mean age = 35.5 years, range = 22-58 years) and 11

presented with LTLE (4 men, 7 women; mean age = 36 years, range = 19-53 years). Seizure

focus was determined using scalp EEG, and (if necessary) intracranial EEG. Twelve controls

were also recruited (8 men, 4 women; mean age = 29.4 years, range = 23-38). Exclusion criteria

included any history of neurological disorders or severe head trauma. Consent was obtained from

all participants. A list of demographics can be found in Table 1.

11

Table 1 Group Demographics, range in brackets

Controls RTLE LTLE

Age (years) 29.4 (23-38) 35.5 (22-58) 36 (19-53)

Years of education 17.9 (10-26) 14.4 (8-20) 14.6 (10-20)

Gender

Female

Male

4

8

7

5

7

4

Seizure Duration -- 16.2 (4-39) 18.4 (3-32)

12

2 Neuropsychological Testing Patients were administered a standard neuropsychological battery as part of the clinic’s standard

pre-operative investigation which included several tests assessing learning/memory. These

include verbal and visual long term memory tests. Two verbal tests were available and included

the Rey Auditory Verbal Learning Test (RAVLT) and Warrington Recognition Memory Test for

words. The visual long term memory tests included the Rey Visual Design Learning Test

(RVDLT) and Warrington Recognition Memory Test for faces. These measures were used to

create behavioural ARs according to the formula [(Verbal memory – Visual memory)/(Verbal

memory + Visual memory)]. One behavioural AR was calculated using scores from the RAVLT

and the RVDLT (REY-AR) and one was calculated using scores from the Warrington memory

test for words and the Warrington memory test for faces (WARR-AR). These tests were used to

examine the relationship of recall (via REY-AR) and recognition (via WARR-AR) on other

neuroimaging measurements. Both are clinically relevant, but there is no consensus on which

may be superior for characterizing hippocampal function (McAndrews & Cohn, 2012).

3 fMRI Data Acquisition Data was collected on a 3-T Signa MR system (GE Medical Systems, Milwaukee WI). A high-

resolution 3D anatomic scan was collected first for visualization, normalization of fMRI data to a

common anatomic template and volumetric analysis (T1-weighted sequence, FOV 220mm, 146

slices, flip angle=12degree, TE=3ms, TR=8ms, 256 x 256 matrix, resulting in voxel size of

.85939 x .85939 x 1.0). Echo-Planar Imaging sequences (TE=20 ms, TR=2000 msec, 32 5mm

oblique slices angled to be orthogonal to the long axis of the hippocampus to maximize signal

and minimize partial volume effects in the MTL) were run during the 6 ½ min functional scan.

13

4 fMRI Task Participants performed a commonly used scene encoding task during an fMRI scan. While in the

scanner they passively viewed a series of 60 complex visual scenes for 3500ms each. Two of

these visual scenes were exposed repeatedly immediately prior to the fMRI run and were

presented 15 times each during the critical run to serve as ‘baseline’ trials while the other 30

were novel. Presentation order of scenes was randomized for each subject. Previous literature

suggests that hippocampal activation will be seen for the novel>repeated contrast (Stern et al.,

1996).

5 Functional Data Preprocessing Preprocessing was done using SPM8 (Statistical Parameter Mapping 8; Wellcome Department of

Imaging, London) run through Matlab 7.9 (Mathworks, Inc). The first three images of the

functional scan were dropped to remove acquisition artifact. Preprocessing first included

realignment of anatomical and functional scans to the anterior commissure and then co-

registration of the functional scans to the anatomical scan. The functional scans were then

realigned and unwarped to correct to scanner motion artifact and subject movement. The

anatomical scan was segmented into gray matter, white matter and cerebral spinal fluid. The

resulting parameter files of this segmentation were then used to normalize functional and

anatomical scans into MNI space using affine registration. Finally, the functional images were

smoothed using a full-width at half maximum Gaussian kernel to 8 x 8 x 8 mm resolution. Data

went through a high-pass filter to account for low-frequency drift. Each stimulus event was then

modeled by SPM8’s canonical haemodynamic response function.

14

6 Data Analysis Novel encoding scenes were contrasted against repeated events using a general linear model

(GLM) in SPM8 for each subject. The resulting contrast images were used for group level

analyses to compare LTLE vs. controls, RTLE vs. controls and LTLE vs. RTLE. Healthy

individuals were used as a positive control to ensure implementation of the scene encoding

paradigm was successful. Group differences between healthy controls and our patient groups

were performed to see global changes between healthy and disease state.

ROI masks derived from MARINA (Walter et al., 2003) were applied to each patient’s

contrast to find voxel counts and t-values at a threshold of p<0.15, uncorrected. This liberal

threshold was chosen to allow for an inclusive representation of activated voxels, while

excluding those that are very likely to be due to chance. The application of low thresholding for

this purpose has been applied in the past (Branco et al., 2006; Mechanic-Hamilton et al., 2009).

Individual masks were made for left and right hippocampus and left and right parahippocampus,

as shown in Figure 1. Voxel counts were weighted by their t-scores using these masks. Weighted

voxel counts have been shown to provide a better measure than voxel count alone (Chlebus et al.,

2007). These weighted voxel counts (WVC) were used to create activation ARs using the

formula [(Left WVC-Right WVC) / (Left WVC + Right WVC)]. Larger activation ARs indicate

greater activity in the left ROIs compared to right ROIs. An AR of zero indicates equal activity

bilaterally in the ROIs and negative ARs indicate greater activity in the right ROIs. This resulted

in two activation ARs for each patient (one for the hippocampus and one for the

parahippocampus). If there was no activation above threshold in either the left or right ROI, then

that data point was discarded for that patient. At the p<0.15 threshold there was 5 discarded (1

RTLE, 4 LTLE) for the hippocampal analysis.

15

Figure 1. These coronal MRI images depict the masks used for small volume correction in each patient to produce

activation asymmetry ratios (ARs). Right hippocampus in blue, left hippocampus in red, right parahippocampal

gyrus in purple, left hippocampal gyrus in green.

16

7 Structural Data Analysis FreeSurfer [Martino Center for biomedical Imaging, Harvard-MIT, Boston USA;

http://surfer.nmr.mgh.harvard.edu] was used to find volumetric values for the hippocampi, and

cortical thickness for the parahippocampal gyrus. Cortical thickness rather than volume is used

for the parahippocampal gyrus because the methods used by FreeSurfer tend to underestimate

cortical volume due to its surface based measurement procedures. FreeSurfer has been

previously described in detail (Fischl et al., 2004) and has been assessed in terms of validity and

accuracy (Dickerson et al., 2008; Han et al., 2006; Lee et al., 2006; Morey et al., 2010; Pardoe,

Pell, Abbott, & Jackson, 2009). Preprocessing included intensity normalization, removal of non-

brain tissue, Talaraich transformation and segmentation of tissue into grey matter, white matter

and cerebral spinal fluid. Subcortical structures are then segmented. For cortical thickness, the

white and grey matter boundaries are identified and the distance between the two surfaces are

calculated. The cortex is then automatically parcellated. Thickness is then calculated for a

structure based on this cortical parcellation. Structural ARs were calculated for the hippocampus

according to the formula [(left volume – right volume) / (left volume + right volume)] while ARs

for the parahippocampal gyrus used the formula [(leftthickness-

rightthickness)/(leftthickness+rightthickness)].

8 Group Comparison Independent samples t-tests were used to look at group differences between left and right TLE

patients on neuropsychological memory test ARs, activation ARs, and structural ARs using

SYSTAT 13 (Systat Software Inc. Chicago IL).

17

9 Neuroimaging Measures and Behaviour Mediation analysis was performed following the methods outlined in Baron and Kenny (1986)

using linear regression in SYSTAT 13 (Systat Software Inc. Chicago IL). The hippocampal

activation AR (acting as the predicted mediator) was regressed onto the hippocampal volume AR

(acting as the independent variable). Each behavioural AR (the dependent variables) was then

regressed onto the hippocampal volume ARs, separately. Finally, each behavioural AR was

regressed onto both the hippocampal activation ARs and the hippocampal volume ARs. A

mediation relationship could exist if there is a significant decrease in structural volume ARs

ability to predict memory once activation ARs are added into the regression model. A model of

this is shown in Figure 2. In Figure 2, if c’ is significantly smaller than c, then this suggests a

mediation interaction is present. This significance can be assessed by testing the indirect effect as

outlined by Sobel (1982) which is as follows:

Where b is the slope of the mediator for predicting the dependent, in this case structural

AR predicting memory scores, sb is the standard error of the slope represented by b, a represents

the slope of the independent variable on the mediator, in this case structural AR on activation AR

and sa is the standard error of the slope represented by a. This test produces a z-score which can

be used to test for significance. Additionally, behavioural ARs scores were regressed onto the

parahippocampal activation AR and parahippocampal thickness ARs (separately) for each

subject to test the contribution of the parahippocampus to understanding memory presurgically.

18

A.

B.

Figure 2. A model of the interaction between structural volume, fMRI activation and memory. A. The influence of

volume on memory is shown as path c. B. Shows the model of influence provided a mediation relationship exists.

Volume now indirectly influences memory through activation and the direct path (c’) should be significantly smaller

than the previous direct path (c).

c

c’

a b

19

Chapter 3 Results

1 Demographics There was no difference in age between the RTLE, the LTLE group, and the control group,

F(32,2) = 1.63, p = 0.21. There was a significant group level difference in years of education

between patients with RTLE, patients with LTLE and healthy controls, F(32,2) = 3.5 , p = .042.

Post-hoc tests revealed no difference in years of education between patients with RTLE and

those with LTLE, t(21) = .16, p = .87, no difference in years of education between patients with

RTLE and healthy controls, t(22) = 2.2, p = .12 and no difference between patients with LTLE

and healthy controls, t(21) = 2.15, p = .13. There was no difference in seizure duration between

patients with RTLE and those with LTLE, t(21) = .384, p = .7. These demographic variables are

summarized in Table 1.

2 Task Activation As shown in Figure 3, all groups showed bilateral activation in occipital, fusiform gyrus,

parahippocampal gyrus and hippocampus (p<0.005, uncorrected). Group level analysis, shown in

Figure 4, controls activated bilateral parahippocampal to a greater extent than patients with

LTLE (p<0.005, uncorrected). There were no voxels in the MTL that activated greater in patients

with LTLE compared to healthy controls subjects. Controls had greater activation in the left

parahippocampal and compared to patients with RTLE (p<0.005, uncorrected). There were no

voxels in the MTL that activated more strongly in patients with RTLE compared to healthy

control subjects. Patients with RTLE had greater left parahippocampal and greater left

hippocampal activation compared to patients with LTLE patients (p<0.005, uncorrected). There

20

.

Figure3. Whole brain analysis of functional activation for the contrast novel > repeated scenes for LTLE, RTLE and

control groups at p<0.005, uncorrected, showing activation bilaterally in occipital, fusiform, parahippocampal and

hippocampal regions. Colour bars display t-values.

L

y=-28 z=-8 x=-28

LTLE novel>repeat

Control novel>repeat

L

y=-20 x=-28 z=-16

RTLE novel>repeat

L

y=28 x=-20 z=-10

21

Figure 4. Whole brain group analysis of functional activation for the contrast novel> repeated scenes contrasting

LTLE>RTLE, Controls>LTLE and Controls>RTLE at p<0.005, uncorrected. Colour bars display t-values.

RTLE>LTLE for novel>repeat

L

y=-20 x=-20 z=-20

Controls>LTLE for novel>repeat

y=-48 x=-32 z=-14

L

Controls>RTLE for novel>repeat

y=-32 x=-54 z=-8

L

22

were no voxels in the MTL that demonstrated greater activation in patients with LTLE compared

to those with RTLE at this threshold.

3 Asymmetry Ratios There was no significant difference for behavioural ARs between the LTLE and RTLE groups

Rey Test AR t(21) = 1.85, p = .078, Warrington test AR t(21) = 1.62, p = .12). These null results

are likely a function of small sample size as the Cohen’s d value for REY-AR was 0.79, which is

considered a large effect size, while the Cohen’s d value for the WARR-AR was 0.70 which is

considered to be medium-large. Turning to functional activation, patients with RTLE had

significantly more left lateralized activity compared to LTLE in the hippocampal activation ARs,

t(16) = 3.11, p = .007, but this was not observed for the parahippocampal activation ARs, t(16) =

1.56, p = 0.14. Patients with RTLE also had a significantly different hippocampal volume ARs

compared to patients with LTLE, t(21) = 4.97, p <0.001 but there was no significant difference

in parahippocampal thickness ARs, t(21) = 0.42, p = .68 (See Figure 5 and Table 2).

4 Neuroimaging Measures and Behaviour Hippocampal volume ARs were able to explain a significant proportion of variance in

hippocampal activation ARs, R2= .31, F(1,16) = 7.31, p = .016, establishing a link between the

independent variable and the mediator. This regression is shown in Figure 6A. Hippocampal

volume ARs were not able to significantly explain the variance in WARR-AR, R2 = .061,

F(1,21) = 1.37, p = .26, nor was it able to significantly explain the variance in REY-AR, R2=

.004, F(1,21) = .094, p = .762. This failed to produce a link between the independent variable

and the dependent variable. Hippocampal activation ARs were able to explain a significant

proportion of variance in the WARR-ARs (Figure 6B), R2 = .28, F(1,16) = 6.26, p = .024, but not

REY-ARs, R2 = .003, F(1,16) = .046, p = .833. This established the link between the mediator

23

RTLE LTLE-0.8

-0.4

0.0

0.4

0.8

Rey T

est

AR

RTLE LTLE-0.250

-0.125

0.000

0.125

0.250

Warr

ingto

n T

est

AR

RTLE LTLE-1.0

-0.5

0.0

0.5

1.0

Hip

po A

ctivation A

R

RTLE LTLE-1.0

-0.5

0.0

0.5

1.0

Para

Activation A

R

RTLE LTLE-0.50

-0.25

0.00

0.25

0.50

Hip

po V

olu

me A

R

RTLE LTLE-0.2

-0.1

0.0

0.1

0.2

Para

Thic

kness A

R

Figure 5. Behavioural ARs, activation ARs, hippocampal volume ARs and parahippocampal cortical thickness ARs

using “box-and-whisker” plots between patients with RTLE and LTLE. The centre horizontal line represents the

group median, while the edges of the box represent the 25th

and 75th

percentiles. The ends of the whiskers represent

the 1.5 interquartile range (IQR) of the distribution. Asterisks represent values that fall beyond the 1.5 (IQR).

denotes significant group differences (p < .05) in ARs after correcting for multiple comparisons.

24

Table 2. A summary of the means and standard deviations of the asymmetry ratios (ARs)

produced for the LTLE and RTLE groups.

RTLE LTLE

M SD M SD

REY-AR .22 .22 .054 .2

WARR-AR .073 .051 .037 .054

Hippo activation AR -.050 .24 -.458 .32

Hippo volume AR .11 .11 -.13 .12

Para activation AR .29 .21 .112 .282

Para thickness AR -.011 .031 -.019 .053

25

A

B

Figure 6. Regression models for the mediation analysis. A. hippocampal volume ARs shown as a predictor for

hippocampal activation ARs. B. Hippocampal activation ARs shown as a predictor for WARR-ARs. Patients who

did not show activation in both hemispheres were excluded.

26

and one of the behavioural ARs. When both hippocampal volume ARs and hippocampal

activation ARs were included as predictors of WARR-ARs the overall model failed to explain a

significant amount of variance in WARR-ARs, R2 = .29, F(1,16) = 3.04, p = .078. Activations

ARs were still significant predictors, β = .586, t(16) = 2.23, p = .042, and Volume ARs were not,

β = -.099, t(16) = -.378, p = .711.

Despite the failure to establish a link between volume ARs and a behavioural AR, we

proceeded to test the indirect effect of a mediation relationship using Sobel’s Test (1982), since

the link between volume ARs and behaviour has been reported previously in the literature

(Mechanic-Hamilton et al., 2009). According the Sobel’s Test (1982) there is a significantly

large indirect effect of this mediation model, z = 1.74, p = 0.041, which indicates that the effect

of volume on memory flows through activation. The model is summarized in Figure 7.

Parahippocampal activation ARs and Parahippocampal thickness ARs did not correlate

significantly with any behavioural asymmetry ratios.

Chapter 4 Discussion

This thesis examined the relationship between volume, functional activation during

encoding, and behavioural measures of memory. During the novel scene encoding task all

patients and healthy controls activated bilateral MTL regions including the hippocampus and the

parahippocampal gyrus. This was in support of previous work (Rabin et al., 2004; Stern et al.,

1996). Healthy controls showed greater parahippocampal activation bilaterally compared to

patients with LTLE. This could be due to the widespread structural changes in the MTL that are

seen in TLE (Bell et al., 2011; Bernasconi et al., 2003). The opposite contrast did not reveal

27

Hippo Volume AR

Hippo Activation AR

WARR-AR

r=.56*

controlling for Volume AR

r=.499*

Path c: r=.247

controlling for Activation AR Path c’ r=.054

Figure 7. A depiction of the proposed mediation model accompanied by path correlations showing a

significant correlation between volume ARs and activation ARs for path a, a significant correlations

between activation ARs and WARR-ARs when controlling for volume ARs in path b and a significant

reduction in path c to path c’.

28

any significant regions in the MTL that were activated in LTLE compared to controls.

Interestingly, healthy controls also had greater left parahippocampal activation compared to

patients with RTLE. Patients with RTLE did not activate any MTL regions to a greater degree

than healthy controls. Patients with RTLE activated the left hippocampus and parahippocampus

greater than patients with LTLE during novel scene encoding. This could be due to the fact that

patients with LTLE have damaged left hippocampi leading to weaker activation. There were no

regions in the MTL that were significantly more active in patients with LTLE compared to those

in RTLE, even at very liberal thresholds.

This study was unable to find a significant difference between neuropsychological

asymmetry ratios between patients with LTLE and RTLE though there was a trend shown in

which RTLE patients tended to perform better on verbal memory compared to visual memory

and LTLE patients tended to perform better on visual memory tests than verbal memory tests.

We would expect that an increase in sample size would produce a significant difference that is

shown in the literature in line with material specificity (Marques et al., 2007; McAndrews &

Cohn, 2012; Milner, 1968). The effect sizes between groups for both the Rey memory tests and

the Warrington memory tests were substantial suggesting that a larger sample would yield results

more consistent with the literature.

Functional activation asymmetry within the hippocampus during a novel scene encoding

task differed significantly between our RTLE and LTLE group. As predicted, patients with

LTLE displayed a greater proportion of right hippocampal activation compared to left

hippocampal activation, while patients with RTLE displayed a greater amount of left

hippocampal activation compared to right hippocampal activation. This is in support of previous

findings (Detre et al., 1998; Mechanic-Hamilton et al., 2009). However, there was not a

significant difference in parahippocampal activation asymmetry between patients with LTLE and

29

those with RTLE. Rather, there was a non-significant trend that showed patients with LTLE

preferentially activating the right parahippocampus compared to the left and patients with RTLE

preferentially activating the left parahippocampus compared to the right. Several studies have

included the parahippocampus in ROI analysis, but as a part of a large ROI including the

hippocampus and the fusiform (Mechanic-Hamilton et al., 2009; Rabin et al., 2004) but none

have included a separate mask for the parahippocampal gyri specifically. We expect that a larger

sample would reveal that a significant difference between these two patient groups does in fact

exist.

This study also saw a significant difference in hippocampal volume asymmetry between

patients with LTLE and those with RTLE. This is in line with past studies (Bernasconi et al.,

2003; Mechanic-Hamilton et al., 2009) and showed that patients with LTLE had smaller left

hippocampi compared to their right hippocampus and those with RTLE had smaller right

hippocampi compared to their left hippocampus. There was not a significant asymmetry in

parahippocampal thickness between patients with RTLE and those with LTLE. This could be due

to the extensive boundaries considered for parahippocampal cortex in thickness analysis. It has

been previously shown that parahippocampal atrophy may be more restricted to anterior portion,

specifically the entorhinal and to a lesser degree the perirhinal cortices whereas the posterior

portion of the parahippocampal cortex does not show much atrophy when comparing patients

with TLE to healthy controls (Bernasconi et al., 1999, 2003). Our boundaries for measuring

cortical thickness included entorhinal, perirhinal and posterior parahippocampal cortices together

as one measurement. This may have diluted any differences that may exist in this patient group.

Future studies would therefore be encouraged to use a finer method of examining

parahippocampal cortical measurements that differentiate entorhinal, perirhinal and posterior

parahippocampal cortices.

30

The results of this study also demonstrated that the link between hippocampal volume

and memory is mediated by hippocampal activation. The link between volume and memory was

not found to be significant in this study, but has been previously shown (Mechanic-Hamilton et

al., 2009; Reminger et al., 2004). Despite not finding a significant link between volume and

memory, the influence of hippocampal volume on memory was found to decrease significantly

following the introduction of hippocampal activation asymmetry. Hippocampal volume

asymmetry was correlated with hippocampal activation asymmetry, which in turn was correlated

with memory asymmetry in Warrington words and faces. This relationship leaves hippocampal

activation as the middle link between volume and memory. This finding represents a link that has

not previously been explored in the literature and suggests that the relationship between

hippocampal volume and memory is partially explained through the activation (extent and

magnitude in this case) that tissue is able to generate during encoding. This may suggest that

hippocampal activation asymmetry may be a better marker of hippocampal adequacy than

hippocampal volume is. Two individuals with TLE may both have significant ipsilateral

hippocampal damage, but if one of those individuals has activation biased to their ipsilateral side

and shows relatively unimpaired presurgical memory, that individual may be at greater risk of

post-surgical memory loss than their counterpart who may have activation biased more towards

the contralateral side. The first individual has a decrease in hippocampal volume, yet is still

using their ipsilateral hippocampus to its best ability, whereas the second individual has a

decrease in hippocampal volume, and is relying on their contralateral hippocampus, suggesting

that ipsilateral removal would not cause as much of a disruption.

There was very little overlap between volume asymmetry scores for patients with LTLE

and those with RTLE. This lack of variance however may restrict the ability of volume

asymmetry to correlate with memory. It may have better clinical utility therefore in serving to aid

31

in the assessment of seizure laterality. Conversely, activation asymmetry has a larger variance

that correlates strongly with memory asymmetry. This would allow activation asymmetry to

better assess presurgical functional adequacy of the to-be-removed hippocampus. According to

the functional adequacy theory (Chelune, 1995), individuals who demonstrate a higher degree of

hippocampal adequacy presurgically are at greater risk of post-surgical memory decline.

The relationship between MRI measures of structural and functional integrity is one that

has not been investigated extensively but has more recently become explored (Poppenk &

Moscovitch, 2011; Voets et al., 2012). Poppenk & Moscovitch (2011) looked at healthy controls

across multiple studies to find a relationship between posterior-anterior hippocampal volume

ratios and recollective memory. They went on to find that this relationship between hippocampal

volume and recollective memory was mediated by post-encoding resting-state connectivity of the

posterior hippocampus to precuneus, inferior parietal and lateral frontal cortex. Voets et al. 2012

used resting-state connectivity, voxel-based morphometry and diffusion tensor imaging (DTI) to

look at patients with TLE to identify voxel-wise associations between resting state connectivity

and white matter coherence (determined by fractional anisotropy from DTI). They found the

functional connectivity between parahippocampal cortex, frontal and temporal neocortex was

associated with white matter coherence in the pathways connecting these regions. Their overall

findings suggest that cortical atrophy and white matter coherence are associated with functional

connectivity. Now that neuroimaging has developed tools to look at structural integrity measures

via DTI, volumetry, voxel based morphometry and at functional integrity via activation

contrasts, functional connectivity and multivoxel pattern analysis, we are better prepared to

identify and exploit connections between structure integrity, activation and connectivity. Within

the hippocampus, we have shown a link between volume asymmetry and activation asymmetry

32

and this may generalize to other disorders that may have MTL disruption such as AD,

schizophrenia and depression.

This study attempted to find a link between activation asymmetry and thickness

asymmetry in the parahippocampal gyrus, but was unsuccessful. This may be explained through

having a somewhat small sample size, and having overly inclusive boundaries that would have

been better restricted to certain regions of parahippocampal gyrus such as entorhinal cortex.

Despite the failure of finding significant correlations or differences in parahippocampal regions,

I maintain that exploring parahippocampal parameters as potential predictors post-surgical

memory change may still be of interest, since parahippocampal cortex is also removed during

temporal lobectomy procedures. There have been changes in volume seen in regions of the

parahippocampal cortex (Bernasconi et al., 1999, 2001, 2003) some of which are independent of

hippocampal volume changes, specifically the entorhinal cortex. The entorhinal cortex is thought

to be a gate to the hippocampus and an independent change in the entorhinal cortex could

therefore affect memory performance. This study therefore suggests a more precise look at

entorhinal cortex structure and function measures as potential predictors of post-surgical memory

in future studies following this thesis.

Several limitations to this study exist. Firstly, all patients were using anticonvulsive

medication during all stages of the study. It is unclear as to how this medication may have

affected the memory performance or functional activation during fMRI scanning. For example,

we do not know how these medications may affect vasodilation which would ultimately alter the

BOLD signal. This limitation exists in most studies of this nature. The use of ARs may temper

part of this limitation because it measures a ratio of activation between the hemispheres. Also,

the control group used was not age matched with the patient sample. The control group however

was used to ensure proper task implementation, while the true results of this thesis focusses

33

specifically on differences between LTLE and RTLE groups and on correlations between the

neuroimaging measures with patient memory performance.

Future work should follow up on the reported patients following surgery, to correlate

post-surgical memory change with presurgical activation and volume asymmetries in both the

hippocampus and parahippocampal cortices.

The main goal of this thesis was to explore the relationship between volume, activation

and memory in patients with TLE as well as exploring any influence the parahippocampus may

have on memory. We discovered that activation asymmetry mediates the relationship between

volume asymmetry and memory asymmetry in patients with TLE. This is in line with recent

research that has shown that underlying brain processes mediate the relationship between volume

and memory (Poppenk & Moscovitch, 2011) and suggests the need for more structure-activation

based analyses, such as DTI and resting-state connectivity. Based on our findings we conclude

that hippocampal activation asymmetry is indeed a good predictor of hippocampal adequacy,

perhaps even more so than volume asymmetry. Despite our concerns about the independent

value of activation measures (given the patterns seen in MCI and AD), current results make us

more confident in interpreting regional activation as a viable ‘biomarker’.

34

References

Baron, R. M., & Kenny, D. a. (1986). The moderator-mediator variable distinction in social

psychological research: conceptual, strategic, and statistical considerations. Journal of

personality and social psychology, 51(6), 1173-82. Retrieved from

http://www.ncbi.nlm.nih.gov/pubmed/3806354

Bell, B., & Davies, K. G. (1998). Anterior temporal lobectomy, hippocampal sclerosis, and

memory: recent neuropsychological findings. Neuropsychology review, 8(1), 25-41.

Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9585921

Bell, B., Lin, J. J., Seidenberg, M., & Hermann, B. (2011). The neurobiology of cognitive

disorders in temporal lobe epilepsy. Nature reviews. Neurology, 7(3), 154-64. Nature

Publishing Group. doi:10.1038/nrneurol.2011.3

Berkovic, S. F., Andermann, F., Olivier, A., Ethier, R., Melanson, D., Robitaille, Y., Kuzniecky,

R., et al. (1991). Hippocampal sclerosis in temporal lobe epilepsy demonstrated by

magnetic resonance imaging. Annals of neurology, 29(2), 175-82.

doi:10.1002/ana.410290210

Bernasconi, N., Bernasconi, A., Andermann, F., Dubeau, F., Feindel, W., & Reutens, D. C.

(1999). Entorhinal cortex in temporal lobe epilepsy: a quantitative MRI study. Neurology,

52(9), 1870-6. AAN Enterprises. Retrieved from

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation

&list_uids=10371536

Bernasconi, N., Bernasconi, A., Caramanos, Z., Antel, S. B., Andermann, F., & Arnold, D. L.

(2003). Mesial temporal damage in temporal lobe epilepsy: a volumetric MRI study of the

hippocampus, amygdala and parahippocampal region. Brain : a journal of neurology,

126(Pt 2), 462-9. doi:10.1093/brain/awg034

Bernasconi, N., Bernasconi, A., Caramanos, Z., Dubeau, F., Richardson, J., Andermann, F., &

Arnold, D. L. (2001). Entorhinal cortex atrophy in epilepsy patients exhibiting normal

hippocampal volumes. Neurology, 56(10), 1335-9. Retrieved from

http://www.neurology.org/content/56/10/1335.1.short

Blümcke, I., Thom, M., & Wiestler, O. D. (2002). Ammon’s horn sclerosis: a maldevelopmental

disorder associated with temporal lobe epilepsy. Brain pathology (Zurich, Switzerland),

12(2), 199-211. doi:10.1111/j.1750-3639.2002.tb00436.x

Bonelli, S. B., Powell, R. H. W., Yogarajah, M., Samson, R. S., Symms, M. R., Thompson, P. J.,

Koepp, M. J., et al. (2010). Imaging memory in temporal lobe epilepsy: predicting the

effects of temporal lobe resection. Brain : a journal of neurology, 133(Pt 4), 1186-99.

doi:10.1093/brain/awq006

Branco, D. M., Suarez, R. O., Whalen, S., O’Shea, J. P., Nelson, A. P., da Costa, J. C., & Golby,

A. J. (2006). Functional MRI of memory in the hippocampus: Laterality indices may be

35

more meaningful if calculated from whole voxel distributions. NeuroImage, 32(2), 592-602.

doi:10.1016/j.neuroimage.2006.04.201

Chelune, G. J. (1995). Hippocampal adequacy versus functional reserve: predicting memory

functions following temporal lobectomy. Archives of clinical neuropsychology : the official

journal of the National Academy of Neuropsychologists, 10(5), 413-32. Retrieved from

http://www.ncbi.nlm.nih.gov/pubmed/14588901

Chelune, G. J., Naugle, R. I., Lüders, H., & Awad, I. A. (1991). Prediction of cognitive change

as a function of preoperative ability status among temporal lobectomy patients seen at 6-

month follow-up. Neurology, 41(3), 399-404. Retrieved from

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation

&list_uids=2006008

Chlebus, P., Mikl, M., Brázdil, M., Pazourková, M., Krupa, P., & Rektor, I. (2007). fMRI

evaluation of hemispheric language dominance using various methods of laterality index

calculation. Experimental brain research. Experimentelle Hirnforschung. Expérimentation

cérébrale, 179(3), 365-74. doi:10.1007/s00221-006-0794-y

Delaney, R. C., Rosen, A. J., Mattson, R. H., & Novelly, R. A. (1980). Memory function in focal

epilepsy: a comparison of non-surgical, unilateral temporal lobe and frontal lobe samples.

Cortex, 16(1), 103-117. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/6769639

Detre, J. A., Maccotta, L., King, D., Alsop, D. C., Glosser, G., D’Esposito, M., Zarahn, E., et al.

(1998). Functional MRI lateralization of memory in temporal lobe epilepsy. Neurology,

50(4), 926-32. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9566374

Dickerson, B. C., Fenstermacher, E., Salat, D. H., Wolk, D. a, Maguire, R. P., Desikan, R.,

Pacheco, J., et al. (2008). Detection of cortical thickness correlates of cognitive

performance: Reliability across MRI scan sessions, scanners, and field strengths.

NeuroImage, 39(1), 10-8. doi:10.1016/j.neuroimage.2007.08.042

Dickerson, B. C., Salat, D. H., Bates, J. F., Atiya, M., Killiany, R. J., Greve, D. N., Dale, A. M.,

et al. (2004). Medial temporal lobe function and structure in mild cognitive impairment.

Annals of neurology, 56(1), 27-35. doi:10.1002/ana.20163

Dickerson, B. C., Salat, D. H., Greve, D. N., Chua, E. F., Rand-Giovannetti, E., Rentz, D. M.,

Bertram, L., et al. (2005). Increased hippocampal activation in mild cognitive impairment

compared to normal aging and AD. Neurology, 65(3), 404-11.

doi:10.1212/01.wnl.0000171450.97464.49

Eldridge, L. L., Knowlton, B. J., Furmanski, C. S., Bookheimer, S. Y., & Engel, S. a. (2000).

Remembering episodes: a selective role for the hippocampus during retrieval. Nature

neuroscience, 3(11), 1149-52. doi:10.1038/80671

Elshorst, N., Pohlmann-Eden, B., Horstmann, S., Schulz, R., Woermann, F., & McAndrews, M.

P. (2009). Postoperative memory prediction in left temporal lobe epilepsy: the Wada test is

36

of no added value to preoperative neuropsychological assessment and MRI. Epilepsy &

behavior : E&B, 16(2), 335-40. Elsevier Inc. doi:10.1016/j.yebeh.2009.08.003

Engel, J. (1993). Update on surgical treatment of the epilepsies. Summary of the Second

International Palm Desert Conference on the Surgical Treatment of the Epilepsies (1992).

Neurology, 43(8), 1612-7. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8102482

Engel, J., Wiebe, S., French, J., Sperling, M., Williamson, P., Spencer, D., Gumnit, R., et al.

(2003). Practice parameter: temporal lobe and localized neocortical resections for epilepsy:

report of the Quality Standards Subcommittee of the American Academy of Neurology, in

association with the American Epilepsy Society and the American Association of Neuro.

Neurology, 60(4), 538-47. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12601090

Fischl, B., van der Kouwe, A., Destrieux, C., Halgren, E., Ségonne, F., Salat, D. H., Busa, E., et

al. (2004). Automatically parcellating the human cerebral cortex. Cerebral cortex (New

York, N.Y. : 1991), 14(1), 11-22. doi:10.1093/cercor/bhg087

Gleissner, U, Helmstaedter, C., & Elger, C. E. (1998). Right hippocampal contribution to visual

memory: a presurgical and postsurgical study in patients with temporal lobe epilepsy.

Journal of neurology, neurosurgery, and psychiatry, 65(5), 665-9. Retrieved from

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2170342&tool=pmcentrez&ren

dertype=abstract

Gleissner, Ulrike, Helmstaedter, C., Schramm, J., & Elger, C. E. (2004). Memory outcome after

selective amygdalohippocampectomy in patients with temporal lobe epilepsy: one-year

follow-up. Epilepsia, 45(8), 960-2. doi:10.1111/j.0013-9580.2004.42203.x

Golby, a J., Poldrack, R. a, Brewer, J. B., Spencer, D., Desmond, J. E., Aron, a P., & Gabrieli, J.

D. (2001). Material-specific lateralization in the medial temporal lobe and prefrontal cortex

during memory encoding. Brain : a journal of neurology, 124(Pt 9), 1841-54. Retrieved

from http://www.ncbi.nlm.nih.gov/pubmed/11522586

Han, X., Jovicich, J., Salat, D., van der Kouwe, A., Quinn, B., Czanner, S., Busa, E., et al.

(2006). Reliability of MRI-derived measurements of human cerebral cortical thickness: the

effects of field strength, scanner upgrade and manufacturer. NeuroImage, 32(1), 180-94.

doi:10.1016/j.neuroimage.2006.02.051

Helmstaedter, C., & Elger, C. E. (1996). Cognitive consequences of two-thirds anterior temporal

lobectomy on verbal memory in 144 patients: a three-month follow-up study. Epilepsia,

37(2), 171-80. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8635428

Hermann, B. P., Seidenberg, M., Haltiner, a, & Wyler, a R. (1995). Relationship of age at onset,

chronologic age, and adequacy of preoperative performance to verbal memory change after

anterior temporal lobectomy. Epilepsia, 36(2), 137-45. Retrieved from

http://www.ncbi.nlm.nih.gov/pubmed/7821270

37

Hermann, B. P., Wyler, a R., Richey, E. T., & Rea, J. M. (1987). Memory function and verbal

learning ability in patients with complex partial seizures of temporal lobe origin. Epilepsia,

28(5), 547-54. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/3653060

Hermann, B. P., Wyler, A. R., Somes, G., Berry, A. D., & Dohan, F. C. (1992). Pathological

status of the mesial temporal lobe predicts memory outcome from left anterior temporal

lobectomy. Neurosurgery, 31(4), 652-656; discussion 656-657. Retrieved from

http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med3

&AN=1407450

Jones-Gotman, M., Smith, M. L., Risse, G. L., Westerveld, M., Swanson, S. J., Giovagnoli, A.

R., Lee, T., et al. (2010). The contribution of neuropsychology to diagnostic assessment in

epilepsy. Epilepsy & behavior : E&B, 18(1-2), 3-12. Elsevier Inc.

doi:10.1016/j.yebeh.2010.02.019

Kneebone, A. C., Lee, G. P., Wade, L. T., & Loring, D. W. (2007). Rey Complex Figure: figural

and spatial memory before and after temporal lobectomy for intractable epilepsy. Journal of

the International Neuropsychological Society : JINS, 13(4), 664-71.

doi:10.1017/S1355617707070828

Kuzniecky, R., Bilir, E., Gilliam, F., Faught, E., Martin, R., & Hugg, J. (1999). Quantitative MRI

in temporal lobe epilepsy: evidence for fornix atrophy. Neurology, 53(3), 496-501.

Retrieved from http://www.neurology.org/cgi/content/abstract/53/3/496

Kwan, P., & Brodie, M. J. (2000). Early identification of refractory epilepsy. The New England

journal of medicine, 342(5), 314-9. doi:10.1056/NEJM200002033420503

Lee, J. K., Lee, J.-M., Kim, J. S., Kim, I. Y., Evans, A. C., & Kim, S. I. (2006). A novel

quantitative cross-validation of different cortical surface reconstruction algorithms using

MRI phantom. NeuroImage, 31(2), 572-84. doi:10.1016/j.neuroimage.2005.12.044

Levav, I., & Rutz, W. (2002). The WHO World Health Report 2001 new understanding--new

hope. (World Health Organization, Ed.)The Israel journal of psychiatry and related

sciences, 39(1), 50-6. World Health Organization. Retrieved from

http://www.ncbi.nlm.nih.gov/pubmed/12013710

Marques, C. M., Caboclo, L. O. S. F., da Silva, T. I., Noffs, M. H. D. S., Carrete, H., Lin, K.,

Lin, J., et al. (2007). Cognitive decline in temporal lobe epilepsy due to unilateral

hippocampal sclerosis. Epilepsy & behavior : E&B, 10(3), 477-85.

doi:10.1016/j.yebeh.2007.02.002

Martin, R C, Sawrie, S. M., Knowlton, R. C., Bilir, E., Gilliam, F. G., Faught, E., Morawetz, R.

B., et al. (2001). Bilateral hippocampal atrophy: consequences to verbal memory following

temporal lobectomy. Neurology, 57(4), 597-604. Retrieved from

http://www.ncbi.nlm.nih.gov/pubmed/11524466

Martin, R C, Sawrie, S. M., Roth, D. L., Gilliam, F. G., Faught, E., Morawetz, R. B., &

Kuzniecky, R. (1998). Individual memory change after anterior temporal lobectomy: a base

38

rate analysis using regression-based outcome methodology. Epilepsia, 39(10), 1075-82.

Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9776328

Martin, Roy C, Kretzmer, T., Palmer, C., Sawrie, S., Knowlton, R., Faught, E., Morawetz, R., et

al. (2002). Risk to verbal memory following anterior temporal lobectomy in patients with

severe left-sided hippocampal sclerosis. Archives of neurology, 59(12), 1895-901. Retrieved

from http://www.ncbi.nlm.nih.gov/pubmed/12470177

McAndrews, M. P., & Cohn, M. (2012). Neuropsychology in Temporal Lobe Epilepsy:

Influences from Cognitive Neuroscience and Functional Neuroimaging. Epilepsy Research

and Treatment, 2012, 1-13. doi:10.1155/2012/925238

Mechanic-Hamilton, D., Korczykowski, M., Yushkevich, P. A., Lawler, K., Pluta, J., Glynn, S.,

Tracy, J. I., et al. (2009). Hippocampal volumetry and functional MRI of memory in

temporal lobe epilepsy. Epilepsy & behavior : E&B, 16(1), 128-38.

doi:10.1016/j.yebeh.2009.07.012

Milner, B. (1968). Disorder of memory after brain lesions in man. Preface: material-specific and

generalized memory loss. Neuropsychologia, 6, 175-179.

Milner, B. (1974). Hemispheric specialization: scope and limits. In F. Schmitt & W. FG (Eds.),

The neurosciences: third study program (pp. 75–89). Massachusetts: MIT Press.

Morey, R. A., Selgrade, E. S., Wagner, H. R., Huettel, S. A., Wang, L., & McCarthy, G. (2010).

Scan-rescan reliability of subcortical brain volumes derived from automated segmentation.

Human brain mapping, 31(11), 1751-62. doi:10.1002/hbm.20973

Morris, R. G., Abrahams, S., & Polkey, C. E. (1995). Recognition memory for words and faces

following unilateral temporal lobectomy. The British journal of clinical psychology the

British Psychological Society, 34 ( Pt 4)(Pt 4), 571-576. Retrieved from

http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med3

&AN=8563663

Pardoe, H. R., Pell, G. S., Abbott, D. F., & Jackson, G. D. (2009). Hippocampal volume

assessment in temporal lobe epilepsy: How good is automated segmentation? Epilepsia,

50(12), 2586-92. doi:10.1111/j.1528-1167.2009.02243.x

Poppenk, J., & Moscovitch, M. (2011). A hippocampal marker of recollection memory ability

among healthy young adults: contributions of posterior and anterior segments. Neuron,

72(6), 931-7. Elsevier Inc. doi:10.1016/j.neuron.2011.10.014

Rabin, M. L., Narayan, V. M., Kimberg, D. Y., Casasanto, D. J., Glosser, G., Tracy, J. I., French,

J. a, et al. (2004). Functional MRI predicts post-surgical memory following temporal

lobectomy. Brain : a journal of neurology, 127(Pt 10), 2286-98. doi:10.1093/brain/awh281

Reminger, S. L., Kaszniak, A. W., Labiner, D. M., Littrell, L. D., David, B. T., Ryan, L.,

Herring, A. M., et al. (2004). Bilateral hippocampal volume predicts verbal memory

39

function in temporal lobe epilepsy. Epilepsy & behavior : E&B, 5(5), 687-95.

doi:10.1016/j.yebeh.2004.06.006

Richardson, M. P., Strange, B. A., Duncan, J. S., & Dolan, R. J. (2006). Memory fMRI in left

hippocampal sclerosis: optimizing the approach to predicting postsurgical memory.

Neurology, 66(5), 699-705. doi:10.1212/01.wnl.0000201186.07716.98

Richardson, M. P., Strange, B. a, Thompson, P. J., Baxendale, S. a, Duncan, J. S., & Dolan, R. J.

(2004). Pre-operative verbal memory fMRI predicts post-operative memory decline after

left temporal lobe resection. Brain : a journal of neurology, 127(Pt 11), 2419-26.

doi:10.1093/brain/awh293

Saling, M. M., Berkovic, S. F., O’Shea, M. F., Kalnins, R. M., Darby, D. G., & Bladin, P. F.

(1993). Lateralization of verbal memory and unilateral hippocampal sclerosis: evidence of

task-specific effects. Journal of clinical and experimental neuropsychology, 15(4), 608-18.

doi:10.1080/01688639308402582

Scoville, W. B., & Milner, B. (1957). Loss of recent memories after bilateral hippocampal

lesions. Journal of Neurology Neurosurgery and Psychiatry, 20, 11-21.

Shamim, S., Hasler, G., Liew, C., Sato, S., Theodore, W. H., & Wieser, H.-G. (2004). ILAE

Commission Report. Mesial temporal lobe epilepsy with hippocampal sclerosis. Epilepsia,

45(6), 695-714. Blackwell Publishing Inc. doi:10.1111/j.0013-9580.2004.09004.x

Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equation

models. (S. Leinhardt, Ed.)Sociological Methodology, 13(1982), 290-312. American

Sociological Association. doi:10.2307/270723

Stern, C. E., Corkin, S., González, R. G., Guimaraes, a R., Baker, J. R., Jennings, P. J., Carr, C.

a, et al. (1996). The hippocampal formation participates in novel picture encoding: evidence

from functional magnetic resonance imaging. Proceedings of the National Academy of

Sciences of the United States of America, 93(16), 8660-5. Retrieved from

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=38729&tool=pmcentrez&render

type=abstract

Trenerry, M. R., Jack, C. R., Ivnik, R. J., Sharbrough, F. W., Cascino, G. D., Hirschorn, K. A.,

Marsh, W. R., et al. (1993). MRI hippocampal volumes and memory function before and

after temporal lobectomy. Neurology, 43(9), 1800-1805. AAN Enterprises. Retrieved from

http://www.ncbi.nlm.nih.gov/pubmed/8414035

Tulving, E. (1985). Memory and consciousness. (E. Tulving & F. I. M. Craik, Eds.)Canadian

Psychology, 26(1), 1-12. Canadian Psychological Association. doi:10.1037/h0080017

Voets, N. L., Beckmann, C. F., Cole, D. M., Hong, S., Bernasconi, A., & Bernasconi, N. (2012).

Structural substrates for resting network disruption in temporal lobe epilepsy. Brain : a

journal of neurology, 2350-2357. doi:10.1093/brain/aws137

40

Walter, B., Blecker, C., Kirsch, P., Sammer, G., Schienle, A., Stark, R., & Vaitl, D. (2003).

MARINA: An easy to use tool for the creation of MAsks for Region of INterest Analyses.

NeuroImage, 19(2), CD-Rom. Retrieved from

http://scholar.google.com/scholar?q=MARINA:+An+easy+to+use+tool+for+the+creation+

of+MAsks+for+Region+of+INterest+Analyses&um=1&ie=UTF-

8&sa=N&hl=en&tab=ws#0

Watson, C., Andermann, F., Gloor, P., Jones-Gotman, M., Peters, T., Evans, a, Olivier, a, et al.

(1992). Anatomic basis of amygdaloid and hippocampal volume measurement by magnetic

resonance imaging. Neurology, 42(9), 1743-50. Retrieved from

http://www.ncbi.nlm.nih.gov/pubmed/1513464