Chapter 20 – Tracking Epilepsy Disease Progression with Neuroimaging

Chapter 20 Tracking Epilepsy Disease Progression with Neuroimaging

Boris C. Bernhardt , Ana Carolina Coan , Lorenzo Caciagli , Andrea Bernasconi and Neda Bernasconi

20.1 Introduction

Epilepsy is one of the most prevalent neurological disorders, affecting ~1% of the general population. Since the early hypothesis that “seizures beget seizures” by William Gowers,1 there has been a considerable body of animal and human research suggesting that several forms of epilepsy may be progressive.2, 3 Epilepsy progression may be defined as a cumulative impact of the disease over time. Domains affected by progression may range from clinical manifestations, such as progressively worse seizure control and seizure severity, to cumulatively altered EEG patterns and more extensive brain abnormalities, cognitive decline, together with increased challenges for adequate socioaffective functioning and quality of life.

Neuroimaging has become a key component of the workup in individual patients with epilepsy, as it allows for the in vivo localization of epileptogenic lesions.49 Neuroimaging is especially important in patients with pharmacoresistant focal epilepsies for whom surgery is, overall, the most effective treatment to arrest seizures.10, 11 Particularly magnetic resonance imaging (MRI) has been central to unveiling structural and functional manifestations of the disease. Quantitative imaging studies have furthermore been pivotal in characterizing whole-brain phenotypes,1219 providing meaningful complements to genetic studies and experimental approaches.

Throughout the last decades, individual- and group-level neuroimaging assessments have addressed progression in different epilepsy syndromes. In the current chapter, we aim to overview neuroimaging markers and study designs that have been used in the evaluation of disease progression. We evaluate the evidence put forward by previous imaging studies assessing disease progression, particularly those in temporal lobe epilepsy (TLE) and idiopathic generalized epilepsy (IGE). In TLE, multiple studies have suggested cumulative brain atrophy in patients with longer disease duration, particularly in the mesiotemporal lobe, but also in thalamic and neocortical regions.3, 20, 21 These findings have been complemented by diffusion MRI work suggesting cumulative derangement in white matter architecture and microstructure, together with functional and metabolic studies emphasizing cumulative reorganization in interregional networks.22, 23 Although progressive brain alterations have been less extensively studied in IGE than in TLE, several reports have also suggested progressive changes at the level of the thalamus and neocortical regions.24, 25 Collectively, findings largely support the concept that the disease course in several epileptic populations may be progressive. A critical evaluation of the literature, however, also reveals that the overwhelming majority of previous assessments were either cross-sectional designs with heterogeneous age-control procedures or single-cohort longitudinal studies restricted to small patient groups. We believe that evidence is still partial to determine which specific patient subgroups present with a progressive disease course. Moreover, our understanding on which mechanisms contribute to progression needs to be further refined.2, 26

20.2 Tracking Progression with Neuroimaging Studies

Neuroimaging modalities have been highly useful in the study of epilepsy, particularly MRI, positron emission tomography (PET), single-photon emission computed tomography (SPECT), and magnetoencephalography (MEG).27, 28 As these techniques probe complementary aspects of the nervous system, with variable trade-offs in terms of costs, spatial resolution, and temporal resolution, the promise rests on their combined application to further consolidate our understanding of pathological substrates, pathophysiological mechanisms, and disease progression in epileptic disorders.

Undoubtedly, MRI has played a pivotal role in the investigation and management of epilepsy, particularly the drug-resistant forms.27, 29 Indeed, since its introduction to research and the clinical practice in the late 1980s, ongoing advances in MRI acquisition and analysis have revolutionized diagnostic and therapeutic approaches.28 Given its unmatched spatial resolution, noninvasiveness, and whole-brain coverage, MRI has been the most important modality to reveal lesions co-occurring with the site of seizure origin.29 The existence of different MRI sequences provides flexibility to model multiple aspects of brain organization and pathology, ranging from alteration in morphology (through the analysis of volumetry or cortical thickness based on T1-weighted images) to gliosis (through analysis on FLAIR/T2 intensity), myelination and iron deposition (based on quantitative T1 mapping and T2*/SWI analysis), fiber architecture and tissue microstructure (inferred from diffusion MRI parameters), local function (derived from fMRI derivatives), and interregional connectivity (based on functional MRI connectivity, structural covariance, or diffusion tractography). In addition to being of high utility in defining the surgical target,6, 3032 recent work has also proposed several MRI markers to serve as prognostic biomarkers in the prediction of long-term surgical outcome.14, 3336

Given its noninvasiveness, MRI can provide markers that track disease progression in individual patients over time. So far, however, the overwhelming majority of studies on progressive brain changes have employed cross-sectional designs, which related MRI metrics to disease duration, estimates of seizure frequency, a history of generalized seizures, or age. Statistical methods are largely based on generalized linear models, including straightforward regressions (e.g., between duration of epilepsy and hippocampal volume) or group comparisons (e.g., comparisons of hippocampal volume between new-onset and long-duration patients). Cross-sectional designs are cost-effective and logistically less complex than longitudinal ones. As they do not carry the burden of repeated assessments over time, they may also allow the inclusion of a broader range of measures to explore variable interrelationships that may generate novel hypotheses about disease progression. Nevertheless, cross-sectional studies suffer from the crucial confound of mixing between- and within-subject effects.37, 38 In other words, these designs compare patient groups with short duration to those with long duration (or those with few seizures to those with frequent seizures); yet, results have been frequently interpreted to signify progressive disease trajectory of individual patients. In the latter case, differences across age groups in lifestyle, fitness, but also available treatment, together with other cohort effects, might have influenced findings.

Despite their high costs and logistic challenges, longitudinal assessments are undeniably the more appropriate design to test hypotheses regarding within-subject trajectories, and to infer causality.39 Within the broad category, single-cohort variants follow an individual cohort over time. If carried out over long intervals, such studies may become costly, patients’ follow-up may become increasingly difficult, and test-retest reliability may be compromised by updates in neuroimaging hardware and analysis techniques. Multicohort longitudinal designs overcome some of these limitations, and combine longitudinal and cross-sectional elements. They follow several patient and control cohorts. Further stratification is possible, for example, with respect to levels of drug-control (e.g., controlled vs. pharmacoresistant), duration (e.g., early onset vs. chronic long standing, or according to an interval scale), age, gender, and variables that may interact with disease trajectories, such as initial precipitating events or genetic markup. In effect, stratified multicohort designs theoretically provide the most accurate estimation of disease trajectories, control for age and cohort effects, and adequately model interindividual variability. Longitudinal designs can be analyzed using statistical methods that model both within- and between-subject effects on disease trajectories. A prominent design choice is multilevel modeling, with mixed-effects models being a prominent subcategory. These models estimate fixed effects of a given variable of interest, such as duration or time from baseline, on a dependent variable, such as hippocampal volume or cortical thickness, while also taking into account the within-participant dependence of observations. These techniques can flexibly model uneven sampling intervals, missing data, and an imbalanced number of samples across individuals. By including all data possible, they may increase statistical sensitivity compared to the more widely used repeated-measures ANOVAs in the case of imbalanced or incomplete data. Multilevel models can be used to infer population-level trajectories, to evaluate between-group differences of trajectories, and to examine individual variability in progressive disease course. Notably, and despite their increased costs and challenges to follow a given individual, longitudinal studies have the benefit of requiring drastically smaller samples than cross-sectional studies to capture subtle effects.40 In a power analysis by Steen and colleagues, for example, detecting a 5% difference in brain volume in a 2-sample cross-sectional study required a sample size of 73 per group, while a similar-sample longitudinal study demanded just 5 individuals per group of patients.40

Yet, and as we describe in the following section, longitudinal studies have been surprisingly scarce in the epilepsy neuroimaging literature and stratified multicohort studies are so far nonexistent. One particular challenge is the group of patients with drug-resistant focal epilepsy, who are recommended to undergo surgery once a lesion is detected. This implies that well-characterized patients are operated more readily, while surgery may be delayed for many years in more challenging candidates (with a possibly more severe disease course). Moreover, given that drug-resistant patients are treated with variable drugs and doses throughout the course of their disease, it has been almost impossible to properly evaluate the role of antiepileptic drugs in disease progression. Future prospective studies are, therefore, recommended to devise meaningful strategies that account for such selective attrition and therapy-related confounding effects.

20.2.1 Studies in Temporal Lobe Epilepsy

TLE is the most common drug-resistant epilepsy in adults and is commonly associated with hippocampal sclerosis, the marked cell loss, and/or gliosis seen on histological specimens.41 Quantitative MRI studies have been pivotal in revealing hippocampal pathology in vivo; indeed, hippocampal volume loss measured on T1-weighted images has been shown to correlate with the degree of neuronal loss42 while increased T2 signal is thought to index reactive astrogliosis.43

In the mid-1990s, several cross-sectional studies in TLE brought forward the first evidence that the extent of ipsilateral hippocampal damage as measured on MRI may increase as a function of disease duration,44 seizure frequency estimates,4547 and a history of generalized seizures.47 In a study by Kalviainen and colleagues, hippocampal volume was reported to correlate negatively with both the total number of partial or generalized seizures in patients with left TLE.47 These data were complemented by the finding of prolonged T2 relaxation times in patients with more frequent seizures and a longer disease duration.47

Subsequent cross-sectional assessments by multiple groups have reproduced these findings, mainly in drug-resistant cohorts.4852 For example, Theodore et al. demonstrated in 35 unilateral TLE patients with a history of secondary generalization that duration, but not age at seizure onset, correlated with the degree of ipsilateral hippocampal atrophy.53 Several studies measured hippocampal volumes using automated techniques, and were able to demonstrate results with similar effect sizes.48, 54, 55 Notably, while most assessments reported evidence for progression in the ipsilateral hippocampus, some studies have also suggested cumulative atrophy in the contralateral hippocampus.49, 56

It is notable that there has been no standardized methodology in these studies to control for confounds of normal aging. Indeed, approaches range from the complete omission of age control procedures to the reporting of no age effect in controls, corrections for age or age at seizure onset, or statistical interaction analysis of aging between patients and controls.

Longitudinal imaging comparisons between hippocampal volume changes in patients relative to controls may provide a rather direct control for aging, but these analyses are so far virtually inexistent. In addition to several interesting case studies,5759 only few single-cohort designs have suggested within-patient disease progression. Analyzing serial MRI data in 24 patients recruited from a first seizure clinic, Briellmann and colleagues observed ipsilateral hippocampal volume decrease of 9% over a period of 3.5 years.60 Atrophy rate was modulated by the number of generalized seizures between both scans, suggesting a harmful effect of even a few generalized seizures on the brain. In a longitudinal study evaluating a small sample of seizure-free patients and those with continuing seizures over a similar follow-up period, Fuerst and colleagues observed progressive ipsilateral atrophy in the latter but not former subgroup, also suggesting a link between seizures and disease progression.61

Cross-sectional volumetric studies have suggested that progressive changes are not limited to the hippocampus, but may also involve adjacent mesiotemporal regions, such as the entorhinal cortex and amygdala (Figure 20.1A).62 Cumulatively increased structural compromise in mesiotemporal networks may extend to both hemispheres, as evidenced by a recent surface-shape analysis.63 In this study, a high similarity was observed between cross-sectional findings in 134 patients and longitudinal analysis in a subgroup of 31 (Figure 20.1B).63 Notably, surface-based mapping of local atrophy has further refined the pattern of progressive changes in the hippocampus itself, by consistently pointing to high effect sizes particularly in the CA1 subregion.63, 64 Given that progressive atrophy preferentially colocalized with areas displaying marked neuronal loss on histology,41, 65 these results emphasize the ability of advanced structural MRI processing to unveil lesional tissue and progressive structural damage otherwise not detected on visual evaluation or global volumetry.6

Figure 20.1. Imaging biomarkers (left column) and findings in temporal lobe epilepsy (TLE, right column). Examples are shown for mesiotemporal volumetry (A), hippocampal surface-shape mapping (B), voxel-based morphometry (C), and cortical thickness analysis (D). The selected findings in TLE collectively suggest more marked atrophy in patients with a longer duration of epilepsy.

Adapted from Bernhardt et al.,12 Coan et al.,20 Bernasconi et al.,62 and Bernhardt et al.63 with permission.

Considering the thalamus, data derived from volumetry, MR spectroscopy, voxel-based morphology, as well as surface-shape mapping have collectively yielded findings that suggest progressive damage and neuronal dysfunction in TLE,6669 in line with the concept that the thalamus plays an important role in the pathophysiological network of this condition.70, 71

Cross-sectional assessments of the neocortex based on voxel-based morphometry,67, 72 whole-brain volumetric techniques,48 and MRI-based cortical thickness measurements15, 21, 7375 have shown widespread and multilobar atrophy increasing with longer disease duration. Again, the few longitudinal studies performed to date have been rather sensitive despite the modest samples studied, and could consistently reveal increased atrophy over time across all lobes (Figure 20.1C–D).12, 20, 21 These studies could furthermore suggest a modulation of progressive trajectories by several clinical variables, including seizure frequency,20, 21 focus lateralization,20 and duration at baseline.21

As in the case of cross-sectional designs, different approaches have been used to account for potential aging effects in these longitudinal studies. In a series of studies,21, 76 the absence of a longitudinal control group was compensated by cross-sectionally comparing age effects on structural markers between patients and controls. This analysis indicated more marked age-related cortical thinning in the former group, suggesting that progressive atrophy in patients is likely due to not typical aging but rather additional disease-specific effects. In the study of Coan et al.,20 the finding of marked gray matter loss between both scans in the patient sample visually contrasted with that of no significant gray matter changes in controls.

A recent systematic review and meta-analysis of MRI volumetric studies of the hippocampus indicated overall moderate effect sizes supportive of more severe ipsilateral atrophy in patients with longer epilepsy duration and more frequent seizures.77 Additional synthesis of whole-brain morphometric studies emphasized that changes often extend to extratemporal and subcortical regions, collectively supporting that TLE is likely progressive. Notably, quantitative synthesis of study design variability also indicated that previous work was mainly based on cross-sectional inference and effects of chronological aging were rather inconsistently addressed, emphasizing the need for future studies with longitudinal study designs and more rigorous age control procedures.77

Data from other neuroimaging modalities, although performed less frequently, also support the concept of TLE being a progressive disorder. In a cross-sectional diffusion tractography assessment, Keller and colleagues reported a correlation between epilepsy duration and decreased white matter fiber anisotropy (a marker of ordered fiber arrangement, axonal membranes, and myelination)78 in the ipsilateral temporal lobe, bilateral thalamus, and posterior corpus callosum.23 The hypothesis that progressive white matter pathology may be preferentially located in temporolimbic networks affected by seizure spread was also supported by the finding of cumulatively decreased anisotropy in the uncinate fasciculus and cingulum.75 The parallel analysis of cortical thickness in these patients furthermore revealed progressive cortical thinning in the ipsilateral parietal and contralateral frontal lobe.75

Findings in the structural domain have been complemented by reports of progressively altered intrinsic functional connectivity based on task-free (“resting-state”) functional MRI. Zhang et al.,79 for example, observed that the degree of functional connectivity decreases between mesiotemporal regions and parietal midline cortices correlated with epilepsy duration. Duration effects on interhemispheric hippocampal functional connectivity have been reported as well.80 More recent graph-theoretical analyses could show more marked functional connectome changes in patients with a longer disease duration,81 a finding in accordance to a longitudinal analysis of networks derived from cortico-cortical structural covariance patterns.13

Assessing [18F]-fluorodeoxyglucose (18FDG)-PET data in children with new-onset partial seizures (most of them having a temporal seizure focus), Gaillard, Theodore, and colleagues suggested decreased glucose metabolism may be less common in children than in studies of adults with chronic epilepsy.82 In a subsequent report based on 91 patients with drug-resistant TLE, the same group demonstrated a correlation between ipsilateral hippocampal hypometabolism and disease duration;83 moreover, longitudinal PET in children revealed a relation between seizure frequency and the degree of hypometabolism.84

20.2.2 Studies in Idiopathic Generalized Epilepsy

IGE refers to a group of epilepsy syndromes characterized by generalized spike and slow-wave discharges on EEG, occurring in runs of 2.5–4 Hz, with normal background activity.8587 Although their exact pathophysiological substrate remains unknown, abundant studies in animal models and human patients have suggested a major role of cortico-thalamic networks in the generation and maintenance of generalized epileptic activity.8891

As in TLE, the majority of previous neuroimaging studies in IGE has addressed progressive changes using cross-sectional designs. Using MR spectroscopy, Bernasconi and colleagues demonstrated a correlation between decreased thalamic NAA/Cr ratio, a marker for neuronal loss and dysfunction, and duration of epilepsy (Figure 20.2A).24 Thalamic duration effects have also been reported using voxel-based morphometric techniques92, 93 and volumetric assessments.25 In a sample of 50 IGE, a combined voxel-based morphometric and surface-shape analysis localized duration effects on atrophy primarily in anterior-medial and posterior-dorsal aspects of bilateral thalami.94

Figure 20.2. Imaging biomarkers (left column) and findings in idiopathic generalized epilepsy, IGE (right column). Examples are shown for thalamic spectroscopy (A) and a combined thalamic volumetric and cortical thickness analysis (B). The selected findings in IGE suggest more marked compromise of the thalamocortical system in patients with a longer duration of epilepsy.

Adapted from Bernasconi et al.24 and Bernhardt et al.25 with permission.

Several studies have also suggested progressive alterations in neocortical regions. In frontal cortices, there are cross-sectional data supporting progressive gray matter loss based on voxel-based morphometric and cortical thickness assessments in cohorts with juvenile myoclonic epilepsy95, 96 and with generalized tonic-clonic seizures only.25, 93 In the latter IGE subgroup, a previous study showed duration-related atrophy in both thalamic as well as neocortical regions, suggesting progressive compromise of the thalamocortical system (Figure 20.2B).25 Furthermore, the study could observe faster cortical thinning in pharmacoresistant patients relative to those with well-controlled seizures.25

Cross-sectional diffusion MRI analyses have complemented other structural neuroimaging assessments and provided broad support to the notion of abnormal subcortico-cortical connectivity in IGE.22, 23, 97, 98 Moreover, in samples with juvenile myoclonic epilepsy, patients with more frequent generalized seizures show reduced fiber anisotropy in thalamocortical tracts alone22 or in concert with other subcortical bundles.99 Progressive changes in thalamocortical connectivity have independently been suggested using resting-state functional connectivity analysis in a heterogeneous IGE population.100 Last, recent graph theoretical analysis has shown effects of duration on several parameters relating to network topology.101, 102

Longitudinal studies in adult IGE populations are so far nonexistent. In a series of cross-sectional and longitudinal imaging studies on new-onset pediatric cases with IGE, differences in brain structure were observed at baseline between patients and controls,103, 104 indicative of preexisting brain anomalies, together with abnormal trajectories early in the course of the disease, possibly suggesting combinations of alterations in developmental trajectories and progressive changes.104106

20.3 Conclusions

There is abundant neuroimaging data concluding that TLE, being among the most common and widely studied epilepsies, is likely progressive. While relatively consistent findings suggestive of disease progression have been reported at the level of the ipsilateral hippocampus, cumulative atrophy in TLE has been shown to also involve mesiotemporal, thalamic, and neocortical regions—not necessarily restricted to the hemisphere ipsilateral to the focus. Overall, these data together with recent meta-analytical synthesis77 support the concept that TLE is a progressive disorder with a system-level impact on brain networks.107, 108 While progression has been less frequently studied in IGE, there is repeated support for disease progression in several patient subgroups (e.g., juvenile myoclonic epilepsy and patients with generalized tonic-clonic seizures), likely involving thalamic and/or prefrontal and frontocentral networks.

It is noteworthy that most findings have been derived from cross-sectional studies with a relatively heterogeneous control for aging, and that some cross-sectional studies did not support evidence for progressive changes on variable neuroimaging measures in both TLE109 and IGE.110112 Carefully evaluating the few longitudinal studies published to date, a crucial shortcoming is the lack of statistical comparisons in population slopes between patients and controls. It, thus, still remains to be tested whether longitudinally measured progressive changes in TLE and IGE are indeed different from aging, and to which extent.

Future longitudinal studies should help to clarify mechanisms underlying disease progression. In animal models, single seizures have been shown to cause apoptotic changes or intermediate forms of neuronal death.113, 114 Similar findings have been observed after experimental115 and human status epilepticus.116 In animals, seizures can increase the susceptibility of network synchronization, lowering the threshold to generate new seizures.117, 118 Seizures have been suggested to damage the brain, possibly through the upregulation of neuronal-axonal excitability markers,119 particularly glutamate.120 Furthermore, disruptions in cortical GABAergic circuits have been described, potentially contributing to genesis or maintenance of seizure activity.121

Effects of antiepileptic drugs on brain networks are largely unknown. Phenytoin122 has been suggested to induce cerebellar atrophy, and valproic acid therapy has been associated with pseudoatrophy123 and cortical thinning.124 On the other hand, the same drugs have been attributed neuroprotective effects promoting neurogenesis.125, 126 Given that drug-resistant patients often have a history of multiple and combined drug trials, this cohort may pose practical challenges on the implementation of prospective studies with well-controlled and standardized medical treatment.

Progression may be not only heterogeneous across different epilepsy syndromes but also variable within a specific syndrome. Considering TLE, clinical observations and research findings suggest that some patients may have a more severe disease course and accelerated decline compared to others. Genetic studies may provide important data that could explain interindividual differences in susceptibility for seizure-related atrophy. Moreover, it remains to be evaluated how additive effects of challenges in psychosocial functioning, lifestyle choices, and comorbid depression can be distinguished from the effects of epilepsy and seizures.

In sum, further longitudinal studies that follow patient and control cohorts using advanced neuroimaging are needed to provide high-level evidence for disease progression and identify its underlying factors. Future studies should statistically compare trajectories and identify patient subgroups based on patterns of progression. As the necessity for treatment in drug-resistant cohorts precludes tracking of within-subject change over time, structured and accelerated designs that systematically enroll patients at different time points in their disease course, from new onset to chronic long standing, are recommended.127 In the ideal case, such an investigation should include both drug-resistant and well-controlled patients and closely monitor medication dose, seizure counts, and psychosocial functioning throughout the testing interval. Such a rich design clearly demands multicentric and multidisciplinary efforts.


1.Gowers WR. Epilepsy and Other Chronic Convulsive Disorders: Their Causes, Symptoms and Treatment. London: J&A Churchill; 1881.Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar

2.Sutula TP, Hagen J, Pitkanen A. Do epileptic seizures damage the brain? Curr Opin Neurol. 2003;16:189–95. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar

3.Cascino GD. Temporal lobe epilepsy is a progressive neurologic disorder: time means neurons! Neurology. 2009;72:1718–9. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

4.Bernasconi A, Antel SB, Collins DL, et al. Texture analysis and morphological processing of magnetic resonance imaging assist detection of focal cortical dysplasia in extra-temporal partial epilepsy. Ann Neurol. 2001;49:770–5. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

5.Colliot O, Bernasconi N, Khalili N, et al. Individual voxel-based analysis of gray matter in focal cortical dysplasia. NeuroImage. 2006;29:162–71. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

6.Bernasconi A, Bernasconi N, Bernhardt BC, et al. Advances in MRI for “cryptogenic” epilepsies. Nat Rev Neurol. 2011;7:99108. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

7.Hong S, Kim H, Bernasconi N, et al. Automated detection of cortical dysplasia type II in MRI-negative epilepsy. Neurology. 2014;83:4855. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

8.Huppertz HJ, Grimm C, Fauser S, et al. Enhanced visualization of blurred gray-white matter junctions in focal cortical dysplasia by voxel-based 3D MRI analysis. Epilepsy Res. 2005;67:3550. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

9.Wagner J, Weber B, Urbach H, et al. Morphometric MRI analysis improves detection of focal cortical dysplasia type II. Brain. 2011;134:2844–54. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

10.Wiebe S, Blume WT, Girvin JP, et al. A randomized, controlled trial of surgery for temporal-lobe epilepsy. N Engl J Med. 2001;345:311–8. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

11.Engel J Jr, McDermott MP, Wiebe S, et al. Early surgical therapy for drug-resistant temporal lobe epilepsy: a randomized trial. JAMA. 2012;307:922–30. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

12.Bernhardt BC, Bernasconi N, Concha L, et al. Cortical thickness analysis in temporal lobe epilepsy: reproducibility and relation to outcome. Neurology. 2010;74:1776–84. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar

13.Bernhardt BC, Chen Z, He Y, et al. Graph-theoretical analysis reveals disrupted small-world organization of cortical thickness correlation networks in temporal lobe epilepsy. Cereb Cortex. 2011;21:2147–57. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

14.Hong S, Bernhardt BC, Schrader DV, et al. Whole-brain MRI phenotying of dysplasia-related frontal lobe epilepsy. Neurology. 2016;86:643–50. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar

15.McDonald CR, Hagler DJ Jr, Ahmadi ME, et al. Regional neocortical thinning in mesial temporal lobe epilepsy. Epilepsia. 2008;49:794803. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

16.Keller SS, Roberts N. Voxel-based morphometry of temporal lobe epilepsy: An introduction and review of the literature. Epilepsia. 2008;49:741–57. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

17.Koepp MJ, Woermann FG. Imaging structure and function in refractory focal epilepsy. Lancet Neurol. 2005;4:4253. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

18.Woermann FG, Free SL, Koepp MJ, et al. Voxel-by-voxel comparison of automatically segmented cerebral gray matter—a rater-independent comparison of structural MRI in patients with epilepsy. NeuroImage. 1999;10:373–84. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

19.Woermann FG, Free SL, Koepp MJ, et al. Abnormal cerebral structure in juvenile myoclonic epilepsy demonstrated with voxel-based analysis of MRI. Brain. 1999;122:2101–8. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

20.Coan AC, Appenzeller S, Bonilha L, et al. Seizure frequency and lateralization affect progression of atrophy in temporal lobe epilepsy. Neurology. 2009;73:834–42. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

21.Bernhardt BC, Worsley KJ, Kim H, et al. Longitudinal and cross-sectional analysis of atrophy in pharmacoresistant temporal lobe epilepsy. Neurology. 2009;72:1747–54. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

22.Deppe M, Kellinghaus C, Duning T, et al. Nerve fiber impairment of anterior thalamocortical circuitry in juvenile myoclonic epilepsy. Neurology. 2008;71:1981–5. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

23.Keller SS, Schoene-Bake JC, Gerdes JS, et al. Concomitant fractional anisotropy and volumetric abnormalities in temporal lobe epilepsy: cross-sectional evidence for progressive neurologic injury. PLOS ONE. 2012;7:e46791. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

24.Bernasconi A, Bernasconi N, Natsume J, et al. Magnetic resonance spectroscopy and imaging of the thalamus in idiopathic generalized epilepsy. Brain. 2003;126:2447–54. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

25.Bernhardt BC, Rozen DA, Worsley KJ, et al. Thalamo-cortical network pathology in idiopathic generalized epilepsy: insights from MRI-based morphometric correlation analysis. NeuroImage. 2009;46:373–81. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

26.Coan AC, Cendes F. Epilepsy as progressive disorders: what is the evidence that can guide our clinical decisions and how can neuroimaging help? Epilepsy Behav. 2013;26:313–21. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

27.Duncan J. The current status of neuroimaging for epilepsy. Curr Opin Neurol. 2009;22:179–84.Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

28.Kuzniecky RI, Knowlton RC. Neuroimaging of epilepsy. Semin Neurol. 2002;22:279–88. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

Only gold members can continue reading. Log In or Register to continue

Jan 3, 2021 | Posted by in NEUROLOGY | Comments Off on Chapter 20 – Tracking Epilepsy Disease Progression with Neuroimaging
Premium Wordpress Themes by UFO Themes