V m P R a b a A R R A A K T H N D 1 p h s ( c f p a b F 1 h 0 Epilepsy Research 132 (2017) 100–108 Contents lists available at www.sciencedirect.com Epilepsy Research journa l h om epa ge: www.elsev ier .com/ locate /ep i lepsyres oxel-based analysis of diffusion tensor imaging in patients with esial temporal lobe epilepsy atrícia Sanchesa,b, Elaine Keiko Fujisaoa, Aline M.S. Bragaa, Nathalia Raquel Cristaldoa, oberto dos Reisa,b, Seizo Yamashitab, Luiz Eduardo Bettinga,∗ Departamento de Neurologia, Psicologia e Psiquiatria, Faculdade de Medicina de Botucatu – UNESP – Univ Estadual Paulista, Brazil Departamento de Doenç as Tropicais e Diagnóstico por Imagem, Faculdade de Medicina de Botucatu – UNESP – Univ Estadual Paulista, Brazil r t i c l e i n f o rticle history: eceived 14 May 2016 eceived in revised form 19 February 2017 ccepted 24 March 2017 vailable online 25 March 2017 eywords: emporal lobe epilepsy ippocampal sclerosis euroimaging iffusion tensor imaging a b s t r a c t Purpose: Quantitative techniques of diffusion analysis allow for an in-vivo investigation of the phys- iopathology of epilepsies. The objective of this study was to evaluate the variation of the main diffusion parameters and explore differences between two methodologies of voxel-wise analysis comparing a group of patients with mesial temporal lobe epilepsy (MTLE) with controls. Methods: 24 patients with a diagnosis of MTLE were selected. All patients and a control group of 36 individuals were submitted to 3 T magnetic resonance imaging. Diffusion parameters were obtained from the raw images. Based on the tensors, a customized template was created, and images were registered into standard space. Voxel-based comparisons between patients and controls was performed by whole brain voxel-wise analysis and tract-based spatial statistics (TBSS). Tract-specific analysis (TSA) was performed in the mostly damaged fasciculi. Results: 10 patients presented with right hippocampal sclerosis (HS), 11 with left HS and 3 with bilateral HS with left predominance. Whole brain voxel-wise analysis showed abnormalities mainly local- ized in the temporal lobes (total volume of 3859 mm3). TBSS showed more widespread abnormalities (21931 mm3). TSA pointed to abnormalities situated essentially in the temporal stem topography. Frac- tional anisotropy (FA) and radial diffusivity (RD) were the parameters that showed more abnormalities. Conclusion: Whole brain voxel-wise analysis was more restricted than TBSS. The methods were comple- mentary stressing the significance of the findings. The abnormalities were more frequently observed in FA and RD indicating the need for using several diffusion parameters for the investigation of patients with MTLE. . Introduction Investigations of the white matter in patients with mesial tem- oral lobe epilepsy (MTLE) using diffusion tensor imaging (DTI) ave been able to detect abnormalities predominantly, but unre- tricted to, the temporal lobe allocated in a centrifugal pattern Concha et al., 2012). This pattern is under investigation and hanges according to the heterogeneity of the patients and the dif- erent methods of neuroimaging analysis, including the acquisition arameters and post-processing algorithms. Voxel-based analyses are quantitative techniques which have s major advantages the ability to statistically compare groups of rains (Yasuda et al., 2010). In diffusion images, voxel-based anal- ∗ Corresponding author at: Departamento de Neurologia, Psicologia e Psiquiatria, aculdade de Medicina de Botucatu – UNESP – Univ Estadual Paulista, Zip Code 8618-687, Botucatu, SP, Brazil. E-mail address: betting@fmb.unesp.br (L.E. Betting). ttp://dx.doi.org/10.1016/j.eplepsyres.2017.03.004 920-1211/© 2017 Elsevier B.V. All rights reserved. © 2017 Elsevier B.V. All rights reserved. yses may be performed directly in the diffusion maps using an analogous methodology of the voxel-based morphometry (VBM) (Ashburner and Friston, 2000). The main drawback with this approach comes from incorrect registration of the images which, is considered a critical problem in DTI (Bookstein, 2001; Abe et al., 2010). Another widely used technique is tract-based spatial statistics (TBSS) (Smith et al., 2006). This methodology consists in projecting the diffusion data onto a simplified map of the white-matter. The main objective of TBSS creation was to overcome the registration problem. However, limitations, especially concerning this issue and the direction of the tracts, still exist (Bach et al., 2014; Schwarz et al., 2014). Improved voxel-based analyses of diffusion images may be achieved using tensor-based registration (Keihaninejad et al., 2013). Because whole brain voxel-wise analysis and TBSS are per- formed with different approaches, the main hypothesis is that they may show different aspects of the abnormalities in patients with dx.doi.org/10.1016/j.eplepsyres.2017.03.004 http://www.sciencedirect.com/science/journal/09201211 http://www.elsevier.com/locate/epilepsyres http://crossmark.crossref.org/dialog/?doi=10.1016/j.eplepsyres.2017.03.004&domain=pdf mailto:betting@fmb.unesp.br dx.doi.org/10.1016/j.eplepsyres.2017.03.004 Resea M o t i 2 w r D e d I p o e w n s 2 K 1 t p v 2 V o o s 1 e a p n T d s i 2 a e p e 1 4 8 c a m 2 5 b P. Sanches et al. / Epilepsy TLE. In addition, differences may also be observed depending n the diffusion parameter used. The objective of this investiga- ion was a multimodal investigation of tensor-based registered DTI mages of patients with MTLE. . Methods This study was approved by the local ethics committee. Patients ith refractory MTLE (n = 24, 16 women, mean age 42 ± 12 years, ange 20–68) and with hippocampal sclerosis (HS) were selected. iagnosis of MTLE was performed according to clinical and lectroencephalographic features following previous recommen- ations (Commission on Classification and Terminology of the nternational League Against Epilepsy, 1989; Berg et al., 2010). All atients had focal seizures which were detailed by clinical history btained from patients or relatives. Patients with dual pathology or xtra-temporal epilepsy were excluded. Control subjects (n = 36, 18 omen, mean age 33 ± 11 years, range 21–57) without histories of eurologic disease were recruited from the local community. All ubjects gave informed consent. .1. Electroencephalogram (EEG) Interictal EEG was performed with a 32-channel recorder (Nihon ohden, Tokyo, Japan). Electrodes were positioned according to the 0–20 international system of electrode placement and with addi- ional Silverman’s anterior temporal electrodes. All records were erformed with 20 min long, with photic stimulation and hyper- entilation. .2. Magnetic resonance imaging (MRI) MRI acquisitions were performed using a 3 T scanner (Siemens, erio, Erlangen, Germany) with a 12-channel head array coil. HS was diagnosed by conventional visual examination based n two main sequences: coronal perpendicular to the long axis f the hippocampus defined at the sagittal image Short T1 inver- ion recovery (STIR; slice thickness, 2.2 mm; field-of-view (FOV), 80 mm; matrix size, 230 × 256; repetition time (TR), 2100 ms; cho time (TE), 9.5 ms; inversion time (TI), 499 ms; flip angle, 150◦; cquisition time, 3.36 min) and T2 periodically rotated overlapping arallels lines with enhanced reconstruction (BLADE; slice thick- ess, 2.2 mm; FOV, 180 mm; matrix size, 230 × 256; TR, 4000 ms; E, 135 ms; flip angle, 120◦; acquisition time, 3.22 min). For hippocampal and whole brain volumetry, three- imensional Magnetization Prepared Rapid Gradient Echo T1 equences (MPRAGE; 192 sagittal slices; slice thickness, 1 mm; n plane resolution, 0.5 × 0.5 mm; FOV, 256 mm; matrix size, 56 × 256; TR, 2300 ms; TE, 2.47 ms; TI, 1100 ms; flip angle, 9◦; cquisition time, 5.21 min per volume) were used. Diffusion weighted images were acquired using single-shot cho planar imaging (65 axial slices; slice thickness, 2 mm; in lane resolution, 1.8 × 1.8 mm; 12 non collinear diffusion gradi- nts direction, b-value 1000 s/mm2; FOV, 230 mm; matrix size 28 × 128 mm; TR, 9100 ms; TE, 96 ms; TI, 200 ms; flip angle, 150◦; averages to improve signal-to-noise ratio; total acquisition time, .21 min). Images were acquired in Digital Imaging and Communi- ations in Medicine (DICOM) format and transformed into ANALYZE nd the Neuroimaging Informatics Technology Initiative (NIfTI) for- at using MRIcron software (Rorden and Brett, 2000). .3. Hippocampal volumetry The FreeSurfer (http://surfer.nmr.mgh.harvard.edu/, version .3, developed at the Martinos Center for Biomedical Imaging y the Laboratory for Computational Neuroimaging, Charlestown, rch 132 (2017) 100–108 101 MA, U.S.A.) image analysis suite was used to obtain hippocam- pal volumes and total intracranial volumes (TIV). Volumetry was performed to confirm HS and to group the patients according to the most atrophic hippocampus for bilateral HS. The standard pro- cessing pipeline implemented in the software was used. ANALYZE images were imported into the software, and they automatically underwent several processing steps including motion correction, non-brain tissue removal, automatic registration into Talairach space, segmentation of subcortical structures including the hip- pocampus, and tessellation of gray matter white matter boundaries (Fischl et al., 2002; Fischl et al., 2004). Hippocampal volumes were normalized according to the TIV. An asymmetry index was calculated by: 2 x (right hippocampus – left hippocampus) ÷ (right hippocampus + left hippocampus). The volumes and the asymmetry index obtained for each individual were standardized according to the value of normal controls using a z-score transformation. 2.4. Diffusion tensor imaging processing FSL software (FMRIB Software Library, www.fmrib.ox.ac.uk/fsl, version 5.0, created by the Analysis Group, FMRIB, Oxford, U.K.) was used to obtain the main diffusion maps (Jenkinson et al., 2012). NifTI images were imported and submitted to brain extraction fol- lowed by eddy currents correction using the Brain Extraction Tool (BET) and FMRIB’s Diffusion Toolbox (FDT) (Smith, 2002; Behrens et al., 2003). Finally, a diffusion tensor model was fitted at each voxel using DTIFIT (Behrens et al., 2003). After these steps, the DTI parameters obtained were imported into DTI-TK software (www. nitrc.org/projects/dtitk, developed at Penn Image Computing and Science Laboratory, University of Pennsylvania, Philadelphia, PA, U.S.A.) for further processing. This software provides an algorithm to register DTI images aligning white matter tracts by combining the fiber orientation in each voxel (Keihaninejad et al., 2013; Wang et al., 2011). Using DTI-TK, a customized template was created. All images were submitted to an initial rigid alignment with a template based on 550 normal individuals with ages between 20 and 80 years (Zhang et al., 2010). To obtain the customized template, images were subsequently submitted to affine linear alignment and to a diffeomorphic registration algorithm with 6 iterations. After the creation of this specific template, it was individually submitted to rigid, affine and diffeomorphic alignment with a specific enhanced DTI template in ICBM-152 space (International Consortium for Brain Mapping, Montreal Neurological Institute, MNI) (Zhang et al., 2011). Finally, deformation fields obtained for the creation and ICBM-152 transformation of the template were applied for each individual in the study. The final results were aligned images in the standard space with 1 mm isotropic voxels. For each subject, diffu- sion parameter maps of fractional anisotropy (FA), axial diffusivity (AD), radial diffusivity (RD) and mean diffusivity (MD) were finally extracted and used for statistical analysis. 2.5. Statistical analysis Two types of voxel-based comparisons were conducted. A whole brain voxel-based analysis was performed using SPM12 (www.fil. ion.ucl.ac.uk/spm, Wellcome Trust Centre for Neuroimaging, Uni- versity College London, London, U.K.) running under the Matlab R2012b platform (MathWorks, Natick, MA, U.S.A.). In this method, diffusion images were smoothed with an 8 mm Gaussian filter (Full Width at Half Maximum). For each diffusion parameter, compar- isons between patients and controls were performed using a full factorial design with two contrasts (searching for areas of increased and decreased abnormalities). http://surfer.nmr.mgh.harvard.edu/ http://surfer.nmr.mgh.harvard.edu/ http://surfer.nmr.mgh.harvard.edu/ http://surfer.nmr.mgh.harvard.edu/ http://surfer.nmr.mgh.harvard.edu/ http://surfer.nmr.mgh.harvard.edu/ http://surfer.nmr.mgh.harvard.edu/ http://www.fmrib.ox.ac.uk/fsl http://www.fmrib.ox.ac.uk/fsl http://www.fmrib.ox.ac.uk/fsl http://www.fmrib.ox.ac.uk/fsl http://www.fmrib.ox.ac.uk/fsl http://www.fmrib.ox.ac.uk/fsl http://www.nitrc.org/projects/dtitk http://www.nitrc.org/projects/dtitk http://www.nitrc.org/projects/dtitk http://www.nitrc.org/projects/dtitk http://www.nitrc.org/projects/dtitk http://www.fil.ion.ucl.ac.uk/spm http://www.fil.ion.ucl.ac.uk/spm http://www.fil.ion.ucl.ac.uk/spm http://www.fil.ion.ucl.ac.uk/spm http://www.fil.ion.ucl.ac.uk/spm http://www.fil.ion.ucl.ac.uk/spm http://www.fil.ion.ucl.ac.uk/spm 102 P. Sanches et al. / Epilepsy Research 132 (2017) 100–108 Table 1 Clinical, electroencephalographic and neuroimaging features of 24 patients with mesial temporal lobe epilepsy. N HS First Freq Medications EEG Age Gender RH vol RH z LH vol LH z AI z 1 L 8 0 CBZ 1200, LTG 100, CLB 60 Normal 52 Women 3766 0.8 2422 −4.3 9.1 2 L 25 3 LTG 300, CLB 20 Left 58 Men 3894 −0.3 2866 −3.8 6.2 3 R 1 2 TPM 300, PHT 300 Right, Slow 43 Women 2911 −3.2 4213 1.6 −8.6 4 R 11 2 OXC 2100, VPA 1000, CLB 20 Bilateral 33 Men 2809 −4.1 3700 −1.1 −6.5 5 L 12 2 LTG 450, CLB 40 Left 28 Women 3919 −0.4 2532 −5.2 9.0 6 L 10 4 CLB 20, PB 50, CBZ 1000 Left 37 Women 4469 1.4 3907 −0.1 2.5 7 L 7 3 CBZ 1000 Slow 22 Women 4122 1.2 2805 −3.5 7.9 8 L 3 3 PHT 300, CLB 10 Left 58 Men 4153 3.0 2749 −2.4 8.5 9 B 2 0 CBZ 1200, CLB 10 Right 49 Women 2489 −4.4 2411 −5.0 0.2 10 R 17 8 CBZ 600, CLB 20 Right 26 Men 3846 −2.6 4248 −1.4 −2.7 11 R 10 4 CBZ 1200 Right 39 Women 3814 −2.0 4633 0.8 −4.8 12 R 13 6 CBZ 600, LTG 400 Right 61 Women 3629 −1.3 4600 2.5 −5.7 13 R 28 20 CBZ 1200, CLB 20 Right 42 Women 4133 −0.8 4405 0.3 −2.0 14 L 10 4 CBZ 1200, CLB 40 Left 39 Women 4392 0.7 2884 −4.2 8.6 15 L 10 2 PB 100, CBZ 400 Left 41 Men 4304 0.9 3349 −2.2 5.0 16 R 12 20 CBZ 400, CLB 40, TPM 400 Right 37 Women 3110 −3.9 3967 −1.2 −5.8 17 L 5 12 CBZ 1200, CLB 20 Left 37 Women 3353 −1.0 2372 −4.9 7.1 18 R 21 10 CBZ 200, CLB 60 Bilateral 40 Women 3017 −3.5 4494 1.7 −9.2 19 B 7 1 TPM 125, CLB 10 Bilateral 60 Women 2918 −3.9 2205 −6.7 5.6 20 B 5 4 CBZ 1200 Left 43 Women 2442 −4.9 2231 −6.1 1.5 21 L 2 2 CBZ 600, PHT 200, PB 100 – 43 Men 3899 −0.6 2503 −5.4 9.1 22 R 42 8 CBZ 1000 Bilateral 58 Women 3420 −1.2 3852 0.6 −3.1 23 L 7 – CBZ 1000 Left 68 Men 3818 −1.5 2835 −4.7 6.0 24 R 4 3 CBZ 1200, CLB 20 Normal 20 Men 2577 −5.2 3888 −1.1 −9.4 N, Number of patients; HS, hippocampal sclerosis (R, right; L, left; B, bilateral); First, age at first seizure in years; Freq, estimated number of seizures per month; medications, d ic res t ital); E s ampu a in mm o g t p p m f m i a i r 2 p D a p r s a y u d a r w 3 3 ( aily doses in mg of the antiepileptic medications in use at the time of magnet opiramate; PHT, phenytoin; OXC, oxcarbazepine; VPA, valproate; PB, phenobarb low, slow background; Age reported in years; RH vol, volume of the right hippoc fter normalization; LH vol, volume of the left hippocampus (raw, not normalized) The second type of voxel-based analysis used was TBSS. Based n the FA images, a mean skeleton of the white matter of the whole roup was created. For each parameter, the values of white matter racts of all individuals were projected in this FA skeleton. Com- arisons between each parameter of patients and controls were erformed using Randomise, an FSL tool for nonparametric per- utation inference (Winkler et al., 2014). Two contrasts searching or areas of increased and decreased abnormalities and 5000 per- utations were used. After the statistical analysis, a Talairach atlas was used to dentify the nearest gray matter anatomical structure (Talairach nd Tournoux, 1988). For the identification of the main fasciculi nvolved, the Johns Hopkins University (JHU) white-matter tractog- aphy atlas was used (Mori, 2005). This probabilistic atlas identified 0 main structures in 28 individuals (Mori, 2005). Finally, to confirm the results a tract-specific analysis (TSA) was erformed for the main tracts depicted in the previous steps. Using TI-TK, samples of the white matter and tract-specific attributes cross the whole population were obtained. The mean value was rojected onto a medial two dimensional model representing the eal tridimensional tracts. For each individual the values corre- ponding to the spokes from each vertex of the mesh were obtained nd used for the analysis (Yushkevich et al., 2008). Statistical anal- sis was performed comparing the tracts of patients and controls sing a general linear model with two contrasts (increased and ecreased differences). For the three types of analyses, age, sex and TIV were introduced s covariates. The level of significance selected was p < 0.05 cor- ected for multiple comparisons (Family-Wise Error and for TBSS ith additional Threshold-Free Cluster Enhancement). . Results .1. Clinical and EEG findings The mean age of recurrent seizures onset was 10 ± 9 years range, 1–42). At the time of MRI acquisition, the mean fre- onance acquisition (CBZ, carbamazepine, LTG, lamotrigine; CLB, clobazam; TPM, EG, result of the record with indication of epileptiform discharge lateralization; s (raw, not normalized) in mm3; RH z/LH z, z-score of the right/left hippocampus 3; AI z, z-score of the asymmetry index after normalization. quency of focal seizures estimated according to clinical history was 5 ± 5 per month (0–20) and the patients were using a mean of 2 ± 1 antiepileptic drugs (1–3). Interictal EEG showed epileptiform discharges in the left anterior temporal region in 9 patients. Mean- while, 7 patients showed discharges in the right anterior temporal region, and 4 had bilateral discharges. Two patients presented with normal EEGs, 1 patient had abnormal background activity and 1 patient was not subjected to the exam. The clinical and EEG findings are reported in Table 1. 3.2. MRI findings Hippocampal volumetry confirmed that 10 patients presented with left HS, 11 with right HS, and 3 with bilateral HS with left predominance. The mean volume of the left hippocampus was 4315 ± 363 (3428–4961) for the controls and 3336 ± 848 (2205–4633) mm3 for the patients. The mean volume of the right hippocampus was 4412 ± 362 (3682–5268) for the controls and 3550 ± 626 (2442–4469) mm3 for the patients. After normalization, the z-scores obtained for the patients were −2.1 ± 2.7 (−6.7–2.8) for the left and −1.4 ± 2.2 (−5.2–3) for the right hippocampus. The mean asymmetry index was 1.1 ± 5.6 (−9.4–9.1). These findings are detailed in Table 1 and Fig. 1. 3.3. Whole brain voxel-wise analysis For the whole group of patients, we obtained: one cluster of increased AD, volume of 290 mm3, main localization in the right parahippocampal gyrus (x = 31, y = − 18, z = −24; T value = 6.11; Z = 5.32; p = 0.001); two clusters of increased RD, total volume of 1102 mm3, main localization in the right parahippocampal gyrus (x = 28, y = − 16, z = −25; T value = 6.65; Z = 5.67; p < 0.0001); two clusters of increased MD, total volume of 782 mm3, main localization in the right parahippocampal gyrus (x = 29, y = −17, z = −24; T value = 6.52; Z = 5.59; p < 0.0001); five clusters of reduced FA, total volume of 1685 mm3, main localization in the left temporal lobe involving the uncinate and inferior fronto-occipital fasci- P. Sanches et al. / Epilepsy Research 132 (2017) 100–108 103 Fig. 1. Magnetic resonance imaging findings of 24 patients with mesial temporal lobe epilepsy. Numbers indicate the patients (see Table 1 for more clinical and EEG details). For each individual, coronal slices in T1 (superior row) and T2 (inferior row) sequences are demonstrated. Images are in radiological orientation with the slice including the anterior portion of the hippocampus. 104 P. Sanches et al. / Epilepsy Research 132 (2017) 100–108 Table 2 Results of voxel-based analysis comparing axial diffusivity (AD), radial diffusivity (RD), mean diffusivity (MD), and fractional anisotropy (FA) maps of 24 patients with medial temporal lobe epilepsy to those of 36 controls according to two methodologies: whole brain voxel-wise analysis and tract-based spatial statistics. Method Parameter Volume (mm3) P-value Coordinates Localization Talairach JHU Whole brain voxel-wise analysis AD 290 0.001 31, −18, −24 Parahippocampal Right cingulum (hippocampus) RD 758 <0.0001 28, −16, −25 Parahippocampal Right cingulum (hippocampus) 344 <0.0001 −34, −10, −15 Temporal lobe Left inferior fronto-occipital, uncinate, inferior longitudinal MD 595 <0.0001 29, −17, −24 Parahippocampal Right cingulum (hippocampus) 187 0.003 −32, −10, −15 Parahippocampal Left inferior longitudinal, anterior thalamic radiation FA 719 < 0.0001 −33, −11, −13 Sub-lobar Left uncinate, inferior fronto-occipital 624 <0.0001 34, −3, −15 Temporal lobe Right inferior fronto-occipital, anterior thalamic radiation, inferior longitudinal 234 0.001 −40, −33, −3 Temporal lobe Left inferior fronto-occipital, inferior longitudinal, superior longitudinal 57 0.011 −22, −6, 32 Sub-lobar Left inferior longitudinal, anterior thalamic radiation 51 0.012 −26, −28, −2 Frontal lobe Left superior longitudinal Tract-based spatial statistics AD 423 0.028 −39, 7, −30 Superior temporal Left inferior longitudinal, uncinate, inferior fronto-occipital RD 1132 0.001 −38, −10, −19 Temporal lobe Left anterior thalamic radiation, inferior longitudinal, inferior fronto-occipital 412 0.001 28, 4, −35 Uncus Right inferior longitudinal MD 254 0.008 28, 4, −35 Uncus Right inferior longitudinal 69 0.009 −42, −7, −22 Temporal lobe Left inferior longitudinal, uncinate, superior longitudinal, inferior fronto-occipital FA 17581 <0.001 −39, −5, −36 Temporal lobe Left inferior longitudinal 1332 <0.001 42, −3, −35 Middle temporal Right inferior longitudinal 523 0.001 39, −40, 18 Superior temporal Right superior longitudinal 205 0.001 32, −49, 18 Temporal lobe Right inferior fronto-occipital, inferior longitudinal, superior longitudinal, forceps major Results correspond to areas of increased AD, RD, MD and reduced FA. Coordinates are in Montreal Neurological Institute space (MNI; x, y, z). Localization was identified according to the Talairach atlas and the Johns Hopkins University white-matter tractography atlas (Talairach and Tournoux, 1988; Mori, 2005). Fig. 2. Results of the voxel-based analysis comparing the maps of axial diffusivity (AD), radial diffusivity (RD), mean diffusivity (MD) and fractional anisotropy (FA) of 24 p ere us ( the bo b l abno c T M o r z R atients with medial temporal lobe epilepsy to 36 controls. Two methodologies w colored areas) are codified according to the p-value (depicted in the color scale at rain (Slices). A tridimensional inflated brain model also shows the projection of al uli (x = −33, y = −11, z = −13; T value = 6.44; Z = 5.54; p < 0.0001). able 2 and Fig. 2 show these findings. Areas of reduced AD, RD, D and increased FA were not disclosed. For patients with right MTLE, we obtained: two clusters f increased AD, volume of 203 mm3, main localization in the ight hippocampus and parahippocampal gyrus (x = 30, y = −17, = −24; T value = 5.9; Z = 5.16; p = 0.003); two clusters of increased D, total volume of 360 mm3, main localization in the right hip- ed (whole brain voxel-wise analysis and tract-based spatial statistics). The results ttom of the figure) and overlaid in coronal slices of an anatomical template of the rmal areas. Orientation is radiological. pocampus and parahippocampal gyrus (x = 26, y = −17, z = −25; T value = 6.36; Z = 5.47; p = 0.001); three clusters of increased MD, total volume of 338 mm3, main localization in the right hip- pocampus and parahippocampal gyrus (x = 27, y = −17, z = −25; T value = 6.30; Z = 5.43; p = 0.001); two clusters of reduced FA, total volume of 371 mm3, main localization in the right temporal lobe involving the uncinate, inferior fronto-occipital and inferior longi- tudinal fasciculi (x = −33, y = −1, z = −13; T value = 6.81; Z = 5.76; p P. Sanches et al. / Epilepsy Research 132 (2017) 100–108 105 Fig. 3. Results of the voxel-based analysis comparing the maps of axial diffusivity (AD), radial diffusivity (RD), mean diffusivity (MD) and fractional anisotropy (FA) of 10 patients with right (superior rows) and 14 with left (inferior rows) medial temporal lobe epilepsy to 36 controls. Two methodologies were used (whole brain voxel-wise a acco o siona i < a i p v v g z M p c f i l p a 3 i t i t i r c M i p m nalysis and tract-based spatial statistics). The results (colored areas) are codified verlaid in coronal slices of an anatomical template of the brain (Slices). A tridimen s radiological. 0.0001). Fig. 3 show these findings. Areas of reduced AD, RD, MD nd increased FA were not disclosed. For patients with left MTLE, we obtained: one cluster of ncreased AD, volume of 81 mm3, main localization in the left hip- ocampus and parahippocampal gyrus (x = −29, y = −17, z = −23; T alue = 5.66; Z = 5; p = 0.009); two clusters of increased RD, total olume of 735 mm3, main localization in the left parahippocampal yrus involving the anterior thalamic radiation (x = −32, y = −12, = −18; T value = 6.50; Z = 5.56; p < 0.0001); one cluster of increased D, total volume of 456 mm3, main localization in the left arahippocampal gyrus involving the inferior longitudinal fasci- ulus (x = −30, y = −12, z = −18; T value = 6.17; Z = 5.34; p = 0.005); our clusters of reduced FA, total volume of 1505 mm3, main local- zation in the left parahippocampal gyrus involving the inferior ongitudinal fasciculi (x = −27, y = −28, z = 0; T value = 5.82; Z = 5.11; = 0.001). Fig. 3 show these findings. Areas of reduced AD, RD, MD nd increased FA were not disclosed. .4. TBSS analysis For the whole group of patients, it was obtained: one cluster of ncreased AD, volume of 423 mm3, localization in the left superior emporal gyrus involving the inferior longitudinal, uncinate and nferior fronto-occipital fasciculi (x = −39, y = 7, z = −30; p = 0.028); wo clusters of increased RD, total volume of 1544 mm3, main local- zation in the left temporal lobe involving the anterior thalamic adiation, inferior longitudinal and inferior fronto-occipital fasci- uli (x = −38, y = −10, z = −19; p = 0.001); two clusters of increased D, total volume of 323 mm3, main localization in the right uncus nvolving the inferior longitudinal fasciculus (x = 28, y = 4, z = −35; = 0.008); four clusters of reduced FA, total volume of 19641 mm3, ain localization in the left temporal lobe involving the inferior rding to the p-value (depicted in the color scale at the bottom of the figure) and l inflated brain model also shows the projection of all abnormal areas. Orientation longitudinal fasciculus (x = −39, y = −5, z = −36; p < 0.001). Table 2 and Fig. 2 show these findings. Areas of reduced AD, RD, MD and increased FA were not disclosed. For patients with right MTLE, it was obtained: two clusters of increased RD, volume of 2663 mm3, main localization in the right temporal lobe involving the inferior longitudinal fasciculus (x = 42, y = −3, z = −27; p = 0.008); one cluster of increased MD, total vol- ume of 675 mm3, main localization in the right temporal lobe involving the inferior longitudinal fasciculus (x = 31, y = 8, z = −34; p = 0.023); one cluster of reduced FA, total volume of 514 mm3, main localization in the right middle temporal gyrus involving the inferior longitudinal fasciculus (x = 42, y = −3, z = −35; p = 0.005). Fig. 3 shows these findings. Areas of reduced AD, RD, MD and increased AD and FA were not disclosed. For patients with left MTLE, it was obtained: four clus- ters of increased RD, volume of 8478 mm3, main localization in the left superior temporal lobe involving the inferior longitu- dinal fasciculus and anterior thalamic radiation (x = −34, y = 4, z = −31; p = 0.001); three clusters of increased MD, total volume of 1643 mm3, main localization in the left temporal lobe involving mainly the inferior longitudinal fasciculus (x = −42, y = −7, z = −22; p = 0.011); one cluster of reduced FA, total volume of 17928 mm3, main localization in the left temporal lobe involving the inferior longitudinal fasciculus (x = −39, y = −5, z = −36; p < 0.001). Fig. 3 shows these findings. Areas of reduced AD, RD, MD and increased AD and FA were not disclosed. 3.5. TSA analysis TSA showed increased AD, RD and MD as well as reduced FA in patients with MTLE. For the whole group analysis, RD (5442 mm2) and MD (5496 mm2) were the parameters with the larger affected 106 P. Sanches et al. / Epilepsy Research 132 (2017) 100–108 Table 3 Results of tract-specific analysis comparing axial diffusivity (AD), radial diffusivity (RD), mean diffusivity (MD) and fractional anisotropy (FA) maps of 24 patients with medial temporal lobe epilepsy to those of 36 controls. Parameter Area (mm2) P-value T value (mean) Hemisphere Localization AD 645 0.002 2.49 left Inferior fronto-occipital 316 0.002 2.88 left Uncinate 319 0.0012 2.58 right Uncinate RD 217 0.044 1.87 left Inferior longitudinal 813 0.003 2.17 right Inferior longitudinal 425 0.02 1.98 right Inferior longitudinal 1108 0.001 2.87 left Inferior fronto-occipital 637 0.008 1.87 left Inferior fronto-occipital 797 0.007 2.98 right Inferior fronto-occipital 419 0.007 2.08 left Superior longitudinal 300 0.024 1.81 left Superior longitudinal 301 0.005 2.45 left Uncinate 425 <0.0001 3.03 right Uncinate MD 949 0.005 2.18 right Inferior longitudinal 369 0.043 2.03 right Inferior longitudinal 1006 0.002 2.96 left Inferior fronto-occipital 781 0.004 2.1 left Inferior fronto-occipital 1132 0.007 2.89 right Inferior fronto-occipital 496 0.003 1.89 left Superior longitudinal 322 0.003 2.67 left Uncinate 441 <0.0001 2.98 right Uncinate FA 291 0.032 1.83 right Inferior longitudinal 864 0.0006 2.41 left Inferior fronto-occipital 244 0.04 2.54 right Inferior fronto-occipital 259 0.022 2.13 left Superior longitudinal 180 0.017 2.77 right Uncinate Fig. 4. Results of the tract-specific analysis comparing the maps of axial diffusivity (AD), radial diffusivity (RD), mean diffusivity (MD) and fractional anisotropy (FA) of 24 patients with medial temporal lobe epilepsy to 36 controls. Results (colored areas) are codified according to the p-value (depicted in the color scale at the bottom of the figure) and overlaid in the flattened tridimensional model of the fasciculi. An overlay of all fasciculi with a tridimensional inflated brain model is also depicted. Resea a b ( b ( o ( f s a a f d n w ( a m g i i o o 4 w o i a q b s s e t d e a R o F a t m s t i c l fi w c m t c c r c P. Sanches et al. / Epilepsy reas, followed by FA (1838 mm2) and AD (1280 mm2). The distri- ution of the abnormalities was mainly in inferior fronto-occipital 7214 mm2) and inferior longitudinal (3064 mm2) regions followed y the uncinate (2304 mm2) and superior longitudinal fasciculi 1474 mm2). There were no areas of reduced AD, RD or MD, nor f increased FA. Table 3 and Fig. 4 show these findings in detail. For patients with right MTLE, RD (3167 mm2) and MD 2812 mm2) were the parameters with the larger affected areas, ollowed by FA (1878 mm2) and AD (272 mm2). RD, MD and FA howed bilateral areas of abnormalities. The distribution of the bnormalities was mainly in inferior fronto-occipital (3045 mm2) nd inferior longitudinal (3010 mm2) followed by the uncinate asciculi (2347 mm2). Inferior fronto-occipital and inferior longitu- inal fasciculi showed bilateral areas of abnormalities. There were o areas of reduced AD, RD or MD, nor of increased FA. For patients with left MTLE, RD (4527 mm2) and MD (3578 mm2) ere the parameters with the larger affected areas, followed by AD 1356 mm2) and FA (1298 mm2). AD, RD and MD showed bilateral reas of abnormalities. The distribution of the abnormalities was ainly in inferior fronto-occipital (4982 mm2) and superior lon- itudinal (2032 mm2) followed by the uncinate (1883 mm2) and nferior longitudinal fasciculi (1862 mm2). Inferior fronto-occipital, nferior longitudinal and uncinate fasciculi showed bilateral areas f abnormalities. There were no areas of reduced AD, RD or MD, nor f increased FA. . Discussion In this investigation different patterns of diffusion abnormalities ere observed in patients with MTLE depending on the methodol- gy and the diffusion parameter used. For the voxel-based analysis nvestigation, FA and RD were the parameters that showed more bnormalities. Therefore, they should be routinely used in the uantitative investigation of diffusion abnormalities in MTLE. Clinical interpretation of DTI parameters is complex and should e performed with care (Alexander et al., 2007). To maximize the pecificity of the DTI investigation, the use of several parameters, uch as FA, AD, and RD, is recommended. FA is reduced in sev- ral conditions and is considered a marker of fiber integrity but his is an over-simplification. FA basically indicates that orientation ependent aspects of the brain microstructure are different (Jones t al., 2013). Decreasing axonal density, increasing axonal caliber nd reducing the degree of myelination should all lead to increased D and reduced FA (Jones et al., 2013). AD indicates the main fiber rientation. Most investigations of patients with MTLE have used A and MD. In this study, AD was the parameter with the fewest bnormalities, followed by MD, RD and FA respectively. Considering he limitations, these findings may be related with an active inflam- atory process and are in line with pathological investigations of urgical patients (Crespel et al., 2002). The present whole brain voxel-wise and TBSS results indicated he presence of abnormalities in 80% of cases in the superior and nferior longitudinal, inferior fronto-occipital and uncinate fasci- uli. These white matter fasciculi have a diffuse intra hemispheric ocalization extending in the axial orientation of the brain. Our ndings demonstrate a massive involvement of the temporal lobe hite matter with abnormalities in the intricate network of fasci- uli that run through this region. The TSA results further revealed ajor abnormalities in temporal stem topography encompassing he inferior fronto-occipital, inferior longitudinal and uncinate fas- iculi in 90% of cases. Prior research investigating white matter onnectivity in MTLE patients has suggested that differences in esections of this circuitry may contribute to different surgical out- omes (Bonilha et al., 2013). rch 132 (2017) 100–108 107 There are investigations suggesting that patients with HS have more widespread abnormalities than patients with a normal MRI (Scanlon et al., 2013). Indeed, patients with left HS are also predis- posed to having more diffuse abnormalities (Ahmadi et al., 2009). However, these findings are not uniform (Rodriguez-Cruces and Concha, 2015). The patients investigated here were heterogeneous. Patients had mainly unilateral HS, but bilateral patients were also included. In our study, comparisons according to HS side reveled broad abnormalities in left MTLE. In patients with MTLE, white matter abnormalities have been localized to the temporal and extra-temporal regions with a cen- trifugal pattern (Gross, 2011; Otte et al., 2012). Astrogliosis and microstructure rearrangement near the focus and postictal vaso- genic edema distally may be related to this pattern (Concha et al., 2012). White matter abnormalities may be observed in up to 63% of patients submitted to surgical treatment (Kasper et al., 1999; Rodriguez-Cruces and Concha, 2015). The main histologi- cal abnormalities described are axonal and myelinic abnormalities, heterotopic neurons in the white matter, blurring of the gray-white matter boundary and white matter gliosis. Reduced axonal den- sity is consistently reported, and the increased extra-cellular space explains the increment of MD and reduction of FA observed in pre- vious investigations (Rodriguez-Cruces and Concha, 2015). It remains unclear which methodology of voxel-based analysis is the best to evaluate patients with MTLE. Two investigations com- pared voxel-based analysis using SPM versus TBSS. In one study in which FA and MD of 33 patients with MTLE were investigated, Focke et al. (2008) found widespread abnormalities in the temporal lobe mainly ipsilateral to the HS. Similar to the present study, they found that TBSS was more sensitive to changes for white matter abnor- malities. Subsequently, in a study examining FA and ellipsoidal area ratio in 19 patients with MTLE, Afzali et al. (2011) found that TBSS abnormalities were most prominent in the temporal lobe and parahippocampal gyrus. Meanwhile, SPM analysis disclosed abnor- malities mainly in the temporal lobes, corpus callosum and fornix. The authors concluded that TBSS was relatively more localized and that ellipsoidal area ratio was more sensitive to white matter abnor- malities than FA. Both investigations used conventional registration of the images. Registration is a critical issue in voxel-based analysis. Without careful performance and checking of this procedure, false positive results may occur (Bookstein, 2001; Schwarz et al., 2014). In this study, registration was performed with the use of the tensor map as is recommended for DTI (Wang et al., 2011). This approach uses full tensor information to drive a more precise alignment (Keihaninejad et al., 2013). The restriction of our whole brain voxel-wise analysis results to the temporal lobe may be related to the use of Family-Wise Error correction in the statistical analysis. It has been suggested that this threshold may be too rigorous and omit true positive results (Henley et al., 2010). Although a similar correction was applied in our TBSS, the different statistical approach used for the TBSS likely minimized this effect (Winkler et al., 2014), yielding the identifica- tion of more regionally extensive abnormalities, especially for FA. These findings are in agreement with the notion that MTLE patients suffer from extra-temporal abnormalities. One drawback of this investigation was the number of diffu- sion directions acquired. A large number of sampling orientations is recommended to avoid biases (Jones, 2004). However, for this investigation, this impact was reduced because the main objective was not to develop detailed representation of structural abnor- malities. This study was designed to access variation across the main diffusion parameters and to compare two techniques of voxel- based analysis when applied to DTI images acquired using the same protocol. 1 Resea 5 c M i C A S R A A A A A B B B B B C C C F 08 P. Sanches et al. / Epilepsy . Conclusion The findings here lend support that different approaches may ontribute to the current knowledge of the mechanisms behind TLE. The use of all diffusion parameters and distinct techniques s advised because they may show additional findings. onflicts of interest None. cknowledgement Supported by grants number 2011/02961-2 and 2016/17914-3, ão Paulo Research Foundation (FAPESP). eferences be, O., Takao, H., Gonoi, W., et al., 2010. Voxel-based analysis of the diffusion tensor. Neuroradiology 52, 699–710. fzali, M., Soltanian-Zadeh, H., Elisevich, K.V., 2011. Tract based spatial statistical analysis and voxel based morphometry of diffusion indices in temporal lobe epilepsy. Comput. Biol. Med. 41, 1082–1091. hmadi, M.E., Hagler Jr., D.J., McDonald, C.R., et al., 2009. Side matters: diffusion tensor imaging tractography in left and right temporal lobe epilepsy. AJNR Am. J. Neuroradiol. 30, 1740–1747. lexander, A.L., Lee, J.E., Lazar, M., Field, A.S., 2007. Diffusion tensor imaging of the brain. Neurotherapeutics 4, 316–329. shburner, J., Friston, K.J., 2000. Voxel-based morphometry–the methods. Neuroimage 11, 805–821. ach, M., Laun, F.B., Leemans, A., et al., 2014. Methodological considerations on tract-based spatial statistics (TBSS). Neuroimage 100, 358–369. ehrens, T.E., Woolrich, M.W., Jenkinson, M., et al., 2003. Characterization and propagation of uncertainty in diffusion-weighted MR imaging. Magn. Reson. Med. 50, 1077–1088. erg, A.T., Berkovic, S.F., Brodie, M.J., et al., 2010. Revised terminology and concepts for organization of seizures and epilepsies: report of the ILAE Commission on Classification and Terminology, 2005–2009. Epilepsia 51, 676–685. onilha, L., Helpern, J.A., Sainju, R., et al., 2013. Presurgical connectome and postsurgical seizure control in temporal lobe epilepsy. Neurology 81, 1704–1710. ookstein, F.L., 2001. Voxel-based morphometry should not be used with imperfectly registered images. Neuroimage 14, 1454–1462. ommission on Classification and Terminology of the International League Against Epilepsy, 1989. Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia 30, 389–399. oncha, L., Kim, H., Bernasconi, A., Bernhardt, B.C., Bernasconi, N., 2012. Spatial patterns of water diffusion along white matter tracts in temporal lobe epilepsy. Neurology 79, 455–462. respel, A., Coubes, P., Rousset, M.C., et al., 2002. Inflammatory reactions in human medial temporal lobe epilepsy with hippocampal sclerosis. Brain Res. 952, 159–169. ischl, B., Salat, D.H., Busa, E., et al., 2002. Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron 33, 341–355. rch 132 (2017) 100–108 Fischl, B., van der Kouwe, A., Destrieux, C., et al., 2004. Automatically parcellating the human cerebral cortex. Cereb. Cortex 14, 11–22. Focke, N.K., Yogarajah, M., Bonelli, S.B., et al., 2008. Voxel-based diffusion tensor imaging in patients with mesial temporal lobe epilepsy and hippocampal sclerosis. Neuroimage 40, 728–737. Gross, D.W., 2011. Diffusion tensor imaging in temporal lobe epilepsy. Epilepsia 52 (Suppl. 4), 32–34. Henley, S.M., Ridgway, G.R., Scahill, R.I., et al., 2010. Pitfalls in the use of voxel-based morphometry as a biomarker: examples from huntington disease. AJNR Am. J. Neuroradiol. 31, 711–719. Jenkinson, M., Beckmann, C.F., Behrens, T.E., Woolrich, M.W., Smith, S.M., 2012. Fsl. Neuroimage 62, 782–790. Jones, D.K., Knösche, T.R., Turner, R., 2013. White matter integrity, fiber count, and other fallacies: the do’s and don’ts of diffusion MRI. Neuroimage 73, 239–254. Jones, D.K., 2004. The effect of gradient sampling schemes on measures derived from diffusion tensor MRI: a Monte Carlo study. Magn. Reson. Med. 51, 807–815. Kasper, B.S., Stefan, H., Buchfelder, M., Paulus, W., 1999. Temporal lobe microdysgenesis in epilepsy versus control brains. J. Neuropathol. Exp. Neurol. 58, 22–28. Keihaninejad, S., Zhang, H., Ryan, N.S., et al., 2013. An unbiased longitudinal analysis framework for tracking white matter changes using diffusion tensor imaging with application to Alzheimer’s disease. Neuroimage 72, 153–163. Mori, S., 2005. MRI Atlas of Human White Matter, 1 st ed. Elsevier, Amsterdam; Boston. Otte, W.M., van Eijsden, P., Sander, J.W., Duncan, J.S., Dijkhuizen, R.M., Braun, K.P., 2012. A metaanalysis of white matter changes in temporal lobe epilepsy as studied with diffusion tensor imaging. Epilepsia 53, 659–667. Rodriguez-Cruces, R., Concha, L., 2015. White matter in temporal lobe epilepsy: clinicopathological correlates of water diffusion abnormalities. Quant. Imaging Med. Surg. 5, 264–278. Rorden, C., Brett, M., 2000. Stereotaxic display of brain lesions. Behav. Neurol. 12, 191200. Scanlon, C., Mueller, S.G., Cheong, I., Hartig, M., Weiner, M.W., Laxer, K.D., 2013. Grey and white matter abnormalities in temporal lobe epilepsy with and without mesial temporal sclerosis. J. Neurol. 260, 2320–2329. Schwarz, C.G., Reid, R.I., Gunter, J.L., et al., 2014. Improved DTI registration allows voxel-based analysis that outperforms tract-based spatial statistics. Neuroimage 94, 65–78. Smith, S.M., Jenkinson, M., Johansen-Berg, H., et al., 2006. Tract-based spatial statistics: voxelwise analysis of multi-subject diffusion data. Neuroimage 31, 1487–1505. Smith, S.M., 2002. Fast robust automated brain extraction. Hum. Brain Mapp. 17, 143–155. Talairach, J., Tournoux, P., 1988. Co-planar Stereotaxic Atlas of the Human Brain : 3-dimensional Proportional System: an Approach to Cerebral Imaging. Thieme Medical Publishers, Stuttgart; New York: G. Thieme; New York. Wang, Y., Gupta, A., Liu, Z., et al., 2011. DTI registration in atlas based fiber analysis of infantile Krabbe disease. Neuroimage 55, 1577–1586. Winkler, A.M., Ridgway, G.R., Webster, M.A., Smith, S.M., Nichols, T.E., 2014. Permutation inference for the general linear model. Neuroimage 92, 381–397. Yasuda, C.L., Betting, L.E., Cendes, F., 2010. Voxel-based morphometry and epilepsy. Expert Rev. Neurother. 10, 975–984. Yushkevich, P.A., Zhang, H., Simon, T.J., Gee, J.C., 2008. Structure-specific statistical mapping of white matter tracts. Neuroimage 41, 448–461. Zhang, H., Yushkevich, P., Rueckert, D., Gee, J., 2010. A computational white matter atlas for aging with surface-based representation of fasciculi. In: Fischer, B., Dawant, B., Lorenz, C. (Eds.), Biomedical Image Registration. Springer, Berlin Heidelberg, pp. 83–90. Zhang, S., Peng, H., Dawe, R.J., Arfanakis, K., 2011. Enhanced ICBM diffusion tensor template of the human brain. Neuroimage 54, 974–984. http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0005 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0005 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0005 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0005 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0005 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0005 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0005 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0005 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0005 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0005 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0005 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0010 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0015 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0020 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0020 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0020 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0020 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0020 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0020 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0020 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0020 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0020 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0020 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0020 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0025 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0025 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0025 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0025 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0025 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0025 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0025 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0025 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0025 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0025 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0030 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0030 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0030 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0030 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0030 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0030 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0030 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0030 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0030 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0030 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0030 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0030 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0035 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0040 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0045 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0050 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0055 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0060 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0065 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0070 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0075 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0075 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0075 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0075 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0075 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0075 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0075 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0075 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0075 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0075 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0075 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0075 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0080 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0085 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0090 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0095 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0095 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0095 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0095 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0095 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0095 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0100 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0105 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0110 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0115 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0120 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0120 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0120 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0120 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0120 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0120 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0120 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0120 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0120 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0120 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0120 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0120 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0125 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0130 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0135 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0135 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0135 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0135 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0135 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0135 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0135 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0135 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0135 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0140 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0145 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0150 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0155 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0155 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0155 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0155 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0155 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0155 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0155 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0155 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0155 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0155 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0155 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0155 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0160 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0165 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0170 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0170 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0170 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0170 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0170 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0170 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0170 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0170 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0170 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0170 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0170 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0170 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0175 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0175 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0175 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0175 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0175 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0175 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0175 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0175 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0175 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0175 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0175 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0180 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0180 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0180 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0180 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0180 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0180 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0180 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0180 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0180 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0180 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0180 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0180 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0185 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 http://refhub.elsevier.com/S0920-1211(17)30165-1/sbref0190 Voxel-based analysis of diffusion tensor imaging in patients with mesial temporal lobe epilepsy 1 Introduction 2 Methods 2.1 Electroencephalogram (EEG) 2.2 Magnetic resonance imaging (MRI) 2.3 Hippocampal volumetry 2.4 Diffusion tensor imaging processing 2.5 Statistical analysis 3 Results 3.1 Clinical and EEG findings 3.2 MRI findings 3.3 Whole brain voxel-wise analysis 3.4 TBSS analysis 3.5 TSA analysis 4 Discussion 5 Conclusion Conflicts of interest Acknowledgement References