Neurophysiologie Clinique/Clinical Neurophysiology (2014) 44, 343353

Disponible en ligne sur

ScienceDirectwww.sciencedirect.com

REVIEW/MISE AU POINT

Neurophysiology of hypnosisNeurophysiologie de lhypnose

A. Vanhaudenhuysea, S. Laureysa, M.-E. Faymonvilleb,

a Coma Science Group, Cyclotron Research Centre, University of Lige, Sart Tilman B30, Alle du 6 Aotno 8, 4000 Lige, Belgiumb Department of Algology, University Hospital of Lige, Sart Tilman B35, 4000 Lige, Belgium

Received 3 July 2013; accepted 29 September 2013Available online 29 October 2013

KEYWORDSHypnosis;Default modenetwork;External network;Pain perception;Hysteria;Conversion disorders

Summary We here review behavioral, neuroimaging and electrophysiological studies of hyp-nosis as a state, as well as hypnosis as a tool to modulate brain responses to painful stimulations.Studies have shown that hypnotic processes modify internal (self awareness) as well as external(environmental awareness) brain networks. Brain mechanisms underlying the modulation of painperception under hypnotic conditions involve cortical as well as subcortical areas including ante-rior cingulate and prefrontal cortices, basal ganglia and thalami. Combined with local anesthesiaand conscious sedation in patients undergoing surgery, hypnosis is associated with improvedperi- and postoperative comfort of patients and surgeons. Finally, hypnosis can be consideredas a useful analogue for simulating conversion and dissociation symptoms in healthy subjects,permitting better characterization of these challenging disorders by producing clinically similarexperiences. 2013 Elsevier Masson SAS. All rights reserved.

MOTS CLSHypnose ;Rseau du mode pardfaut ;Rseau externe ;Perception de ladouleur ;

Rsum Nous proposons de discuter des tudes comportementales, lectrophysiologiques etde neuroimagerie investiguant lhypnose comme un processus de conscience ou comme unoutil pour moduler les rponses crbrales au repos ou lors de stimulations douloureuses. Dif-frentes tudes ont mis en vidence une modification de lactivit crbrale au niveau desrseaux interne (conscience de soi) et externe (conscience de lenvironnement). Par ailleurs,les mcanismes crbraux qui sous-tendent la modulation de la perception de la douleur sous-hypnose comprennent des rgions telles les cortex cingulaire antrieur et frontal, les ganglionsde la base et le thalamus. Combine une anesthsie locale et une sdation consciente chez lesHystrie ;Symptmes deconversion

patients subissant une chirurgie, lhypnose est galement associe une amlioration pri- etpostopratoire du confort des pacomme un outil utile pour crer sains, ce qui permet de mieux csimilaires. 2013 Elsevier Masson SAS. Tou

Corresponding author.E-mail addresses: avanhaudenhuyse@ulg.ac.be (A. Vanhaudenhuyse),

mfaymonville@chu.ulg.ac.be (M.-E. Faymonville).

0987-7053/$ see front matter 2013 Elsevier Masson SAS. All rights rehttp://dx.doi.org/10.1016/j.neucli.2013.09.006tients et des chirurgiens. Enfin, lhypnose peut tre considre

des symptmes de conversion et de dissociation chez des sujetsaractriser ces troubles en mimant des observations cliniques

s droits rservs.

steven.laureys@ulg.ac.be (S. Laureys),

served.

dx.doi.org/10.1016/j.neucli.2013.09.006http://www.sciencedirect.com/science/journal/09877053http://crossmark.crossref.org/dialog/?doi=10.1016/j.neucli.2013.09.006&domain=pdfmailto:avanhaudenhuyse@ulg.ac.bemailto:steven.laureys@ulg.ac.bemailto:mfaymonville@chu.ulg.ac.bedx.doi.org/10.1016/j.neucli.2013.09.006

3

nmatwbtwpcaaa

apscninflap

M

Wsmp

H

fIsiemumopcirinst

nspab[thicntwDaioimsnrjslcdeDuinsrDiacmiotr

rajcitansii

E

44

Numerous studies have highlighted the interest of hyp-otic procedures in various clinical situations, such as painanagement, treatment of phobia, depression, dissociativend psychotic disorders and so on. Some researchers believehat hypnosis is related to an altered state of consciousness,hile others assume that these phenomena can be explainedy psychological concepts such as clinician-patient expec-ations. Hypnosis can be defined as a procedure duringhich a health professional or researcher suggests that aatient or subject experience changes in sensations, per-eptions, thoughts, or behavior [60]. Hypnosis is seen as

state of focused attention involving focal concentration,nd inner absorption with a relative suspension of peripheralwareness and has three components [56]:

absorption: tendency to become fully involved in a per-ceptual, imaginative, or ideational experience;

dissociation: mental separation of components of experi-ence that would ordinarily be processed together;

suggestibility: responsiveness to social cues, leading to anenhanced tendency to comply with hypnotic instructions,representing a suspension of critical judgment.

We have shown that subjects in a hypnotic state reported phenomenology of an altered state of consciousness:articipants reported a higher degree of absorption and dis-ociation as compared to normal wakefulness and controlonditions [12]. Other studies have also shown that hyp-osis produces alterations in aspects of consciousness ands characterized by modulation of properties of the phe-omenal self-consciousness such as mental ease (i.e. easyow of thoughts), absorption, reduction in self-orientationnd automaticity (i.e. responses are experienced as beingroduced without deliberation and/or effort) [47].

easuring hypnosis in the brain

e here review neuroimaging and electrophysiological (EEG)tudies of hypnosis as a state, as well as hypnosis as a tool toodulate brain responses to stimulation such as, for exam-le, painful stimuli.

ypnosis in the brain at rest

MRI and PET studiesn a first study, regional cerebral blood flow (rCBF) washown to increase by 16% during hypnosis, with specificncrease in occipital and right temporal regions [62]. Sev-ral years later, Maquet et al. [38] explored the brainechanisms at rest underlying hypnosis in healthy vol-nteers and showed that hypnotic state was related to theetabolic activation of cortical areas involving left-sidedccipital, parietal, precentral, premotor, and ventrolateralrefrontal cortices and right-sided occipital and anterioringulate cortices, while a decrease of activity was observedn precuneus, bilateral temporal, medial prefrontal andight premotor cortices. In a functional magnetic resonance

maging (fMRI) study, we showed that self-related and exter-al brain networks were modified under hypnosis [12]. Theelf-related network coincides with midline brain struc-ures such as precuneus and mesio-frontal brain areas (also

Hapl

A. Vanhaudenhuyse et al.

amed default mode network DMN) and is involved inelf-related processes, while the external network encom-asses lateral fronto-parietal regions routinely exhibitingctivity increases during attention-demanding tasks and haseen linked to cognitive processes of external sensory input22]. In the normal awake state, we identified a nega-ive correlation between external and internal awareness inealthy volunteers: explicit subjective reports for increasedntensity of internal awareness were related to increasedonnectivity in the DMN, whereas increased external aware-ess scores were associated with increased connectivity inhe external network [64]. Under hypnosis, the external net-ork exhibited reduced functional connectivity, whereas theMN showed reduced connectivity in its posterior midlinend parahippocampal structures but increased connectivityn its lateral parietal and middle frontal areas [12], whilether works showed opposite results with increased activityn posterior regions of the DMN as compared to decreasedetabolic activity in anterior DMN areas [36,46]. Other fMRItudies have shown a hypnosis-related reduction in DMN con-ectivity [10,39], and increased activity in lateral prefrontalegions (involved in attentional/extrinsic systems) [10]. Sub-ects with high compared to low hypnotizability scores werehown to have greater functional connectivity between theeft dorsolateral prefrontal cortex (involved in executiveontrol processing) and the salience network (involved inetecting, integrating, and filtering relevant somatic andmotional information) [26]. The observed reduction in theMN activity might reflect a decreased degree of contin-ous information being retrieved from the external worldn terms of its relation to oneself [12]. The decreased con-ectivity observed by Demertzi et al. [12] in the extrinsicystem might reflect a blockage of the sensory systems toeceive stimuli as a result of hypnotic suggestion, whileeeley et al. [10] suggested that neural activity in DMN isnversely associated with attentional absorption and directlyssociated with spontaneous or stimulus-independent con-eptual thought. Divergent finding obtained by these studiesay be explained by distinct suggestion instructions used to

nduce hypnosis (e.g. pure/neutral hypnosis vs. experiencef pleasant autobiographical memories) or the experimen-al fMRI designs used (e.g. block vs. continuous eyes-closedesting state design).Finally, in a structural MRI study, Horton et al. [28]

eported differences in brain structure size between lownd highly hypnotizable subjects: highly hypnotizable sub-ects demonstrated a larger (32%) rostrum of the corpusallosum than subjects with low hypnotizability. This areas known to be involved in the allocation of attention andransfer of information between prefrontal cortices. Theuthors suggested that these results provide support for theeuropsychophysiological model that highly hypnotizableubjects have more effective frontal attentional systemsmplementing control, monitoring performance and inhib-ting unwanted stimuli from conscious awareness (Table 1).

lectroencephalography

ighly hypnotizable subjects (as compared to mediumnd low) have been shown to demonstrate different EEGhase synchronization rhythms: high subjects demonstratedess phase synchronization in frontal brain areas [4]. In

Neurophysiology

of hypnosis

345

Table 1 Results of PET and fMRI studies on neural correlates of hypnosis.

Reference Technique Number of subjects anddegree of hypnotizability

Paradigm Results

Maquet et al. [38] PET 9 high Listening to autobiographicaleventRevivification of autobiographicalevent under hypnosisHypnosis with color hallucination

Hypnosis > normal state: increased activity in bilateraloccipital cortex, L inferior parietal, precentral and prefrontalcortices, R ACC, R cerebellumDecreased activity in bilateral temporal, prefrontal and Rpremotor cortices, precuneus, R cerebellum

McGeown et al. [39] fMRI 10 high/7 low Resting stateResting state under hypnosis

Hypnosis > normal alertness in:scoring high: decreased activity in ACC, medial superiorfrontal gyri, L inferior and middle frontal gyriscoring low: decreased activity in PCC, thalamus, caudatenucleus, insula bilaterally; L inferior frontal gyrus, claustrum,lentiform nucleus, R subthalamic nucleusHigh > low: lower level of activity in L inferior, middle,superior and medial frontal gyri

Demertzi et al. [12] fMRI 12 high Resting stateRevivification of pleasantautobiographical event underhypnosisMental imagery ofautobiographical memories

Hypnosis > mental imagery: DMN (intrinsic system): increasedconnectivity in middle frontal and bilateral angular gyri anddecreased connectivity in PCC and bilateral parahippocampalareasExtrinsic system: decreased connectivity in the Rsupramarginal and L superior temporal areas

Deleey et al. [10] fMRI 8 medium to high Passive visual fixation inpre-hypnotic stateHypnosis statePost-hypnosis state

Hypnosis: decreased activation in DMN regions: L medialfrontal, R ACC, bilateral posterior cingulate, and bilateralparahippocampal gyri andIncreased activation in R middle frontal, inferior frontal andbilateral precentral gyri

Lipari et al. [36] fMRI 1 high Resting stateResting state under hypnosis

Hypnosis > normal alertness: increased activity inposteromedial cortex, bilateral occipital areas, superior andinferior parietal lobule, bilateral angular gyri, frontal areas,ACC, R parahippocampalDecreased activity in medial and middle prefrontal cortex

Hoeft et al. [26] fMRI 12 high/12 low Resting stateResting state under hypnosis

High > low: increased functional connectivity between ACC andL dorsolateral prefrontal cortex

fMRI: functional magnetic resonance imaging; DMN: default mode network; R: right; L: left; ACC: anterior cingulate cortex; PCC: posterior cingulate cortex.

3

airss[dhstshawp

trIsatmtataopaa

pwhEiihcHl[

obbtaontipEss[ebhd

asoisb

bSdtconnistatb

H

HscoidtlrhsrruisiH

46

ddition, studies suggested a predominance of alpha activ-ty specifically in the left hemisphere [43] and over anterioregions of cortex [69] in highly hypnotizable subjects. Iteems that hypnotizability rather than being in a hypnotictate was more associated with the generated alpha activity56]. Recently, highly hypnotizable subjects were shown toisplay lower fronto-parietal alpha synchronization duringypnosis as compared to low subjects, suggesting that highubjects have more labile fronto-parietal networks and arehereby more responsive to hypnotic induction [58]. Othertudies found that the most relevant EEG changes in highlyypnotizable subjects consisted in a decrease of alpha bandctivities (localized over motor and visual areas) [36], asell as the fact that mean alpha power seems not to be aredictor of hypnotic susceptibility [54].Regarding other EEG rhythms, several studies have shown

hat level of theta activity (associated with inhibition ofesponses) increased as a function of hypnotizability [46].t appears that highly hypnotizable subjects generated sub-tantially more theta power than did low subjects, with

larger difference in frontal locations [3,33,54]. In addi-ion, hypnosis (in low and high subjects) seems to increaseean theta power, suggesting an intensification of atten-ional processes [54]. However, other studies failed to show

significant relationship between theta activity and hypno-izability [9]. Williams et al. [69] have suggested that thetactivity is more a relaxation index rather than a trait indexf hypnotic susceptibility. Indeed, they showed that thetaower was higher during the hypnosis condition in both highnd low hypnotizable subjects and did not allow differenti-tion between these two types of subjects.Gamma activity also seems to be involved in hypnosis

rocessing, but the interest of this in separating subjectsith high from low hypnotizability is still debated. Highlyypnotizable subjects demonstrate a reduction of gammaEG density, in contrast to subjects with low hypnotizabil-ty [9,21,33]. However, other studies found opposite results,.e. greater 40 Hz spectral amplitudes in high versus lowypnotizable subjects [9,55]. Beta activity was found to dis-riminate between high and low hypnotizable subjects [9].ighly hypnotizable subjects seem to be characterized by aower beta range as compared to low hypnotizable subjects33].

In a study conducted on one virtuoso case studyf hypnosis, Fingelkurts et al. [21] showed a differenceetween hypnosis and rest conditions in delta, theta, alpha,eta and gamma frequency bands. The authors suggestedhat decreased size and stability of delta, beta and gammactivity could be indicative of an increased independencef brain processes, translating an effort to maintain alert-ess in the hypnotic condition, while increased alpha andheta may indicate that subject was more relaxed and facil-tated to process information. Recently, a non-subjectiveattern recognition study based on fractal dimension of theEG signal (i.e. topological and temporal dimensions of theignal) have been reported as the best features for clas-ification of subjects in low, medium or high hypnotizable3]. By studying coherence of the EEG signal, Kirenskaya

t al. [33] showed that baseline EEG differed in coherenceetween subjects with high and low hypnotizability. Indeed,ighly hypnotizable subjects were characterized by higheristributed brain regions coherence within delta, theta, and

stre

A. Vanhaudenhuyse et al.

lpha bands. A study conducted by Hinterberger et al. [25]howed the different states of consciousness that can bebserved during a complete hypnotic procedure (relaxation-nduction-suggestion-waking up) in one highly hypnotizableubject. The dominant pattern highlighted in this study cane summarized as follows:

closed-eyes condition may be associated with increasedbilateral parietal and occipital alpha, parietal sensory-motor and beta activities;

hypnotic state seems characterized by increased frontalalpha, decreased central, frontal and parietal gammabilaterally and increased occipital gamma;

deep hypnotic state is characterized by distributed rein-forcement of activity in all frequency bands;

the awake state showed reduced activity on all frequencybands in central, frontal and parietal areas, while gammaincreased in temporal and prefrontal areas (this last pat-tern is attributed by the authors to a highly relaxed butmindful wake state).

Finally, event-related potential (ERP) studies have alsoeen used to decode cognitive processes under hypnosis.ome studies demonstrated a reduction in the ERP responseuring hypnotic procedure, suggesting diminution of percep-ion of stimuli, whereas other studies failed to detect suchhanges in ERP responses [9,56]. In studies conducted onne virtuoso of hypnosis [31] as well as in highly hyp-otizable subjects [29], the mismatch negativity (MMN egative component elicited by any change or mismatchn a sequence of monotonous auditory stimuli in inattentiveubjects) was found to be larger in hypnosis as compared tohe baseline condition. However, this increase of MMN waslso shown in subjects with low hypnotizability, suggestinghat although related to the hypnosis condition, it could note attributed to distinctive hypnotic processes per se [29].

ow hypnosis can modulate pain perception?

ypnosis combined with local anesthesia and consciousedation in patients undergoing surgery, a technique alsoalled hypnosedation, is associated with improved intra-perative comfort and reduced anxiety and pain, diminishedntraoperative requirements for anxiolytic and analgesicrugs, optimal surgical conditions and a faster recovery ofhe patient [19]. Indications for surgical procedures underocal anesthesia and hypnosedation are listed in Table 2. Aetrospective behavioral study on 337 patients showed thatypnosis as an adjunct procedure to conscious intravenousedation provides better perioperative pain and anxietyelief as compared to simple intravenous sedation orelaxation procedure [15]. A prospective study on patientsndergoing plastic surgery confirmed these observations,.e. decreased peri- and postoperative anxiety, pain andigns of discomfort as well as better surgical conditionsn hypnosis group as compared to a control group [16].ypnosis proposed to patients suffering chronic pain after

pinal cord injury, compared to direct current stimula-ion (tDCS), neurofeedback and meditation was shown toesult in decreased pain intensity perception (the sameffect also being observed for meditation), while tDCS

Neurophysiology of hypnosis

Table 2 Surgical interventions using hypnosedation as aroutine analgesic procedure.

Minor surgery Major surgery

Correction of scars Thyroid lobectomyExtraction of wisdom teeth Total thyroidectomyCorrection of prominent

earsParathyrodectomy

Turbinoplasty septoplasty Breath prothesisNose fracture Correction of mammary

ptosisBurn dressing changes Head and neck liftingBlepharoplasty Headneck cancer with

reconstructionLiposuction SeptorhinoplastyBreath adenomectomy Debridementskin graftingHysteroscopy Maxillofacial reconstructionRemoval of osteosynthesis

materialTubal ligation

Inferior limb varicose vein Vaginal hysterectomyUmbilical or inguinal hernia

htmuctwnrctTitrkepiItsectiotihsmctjoeesspatbipp

H

Hdpbichypothesized that hypnosis and hysteria are characterizedAdapted from [20].

and neurofeedback did not [30]. A recent meta-analysiscomprising a total of 2597 patients undergoing surgical ormedical procedures revealed effects of hypnosis on variouspre- and postoperative factors such as emotional distress,pain, medication consumption, physiological parameters,recovery, and surgical procedure time as compared to stan-dard care alone or an attention control [57]. By applyingthe hypnotic procedure we routinely used in surgery, weshowed that affective (unpleasantness) as well as sensory(perceived intensity) components of pain perception wereboth reduced as compared to control conditions [17,16,18].

Brain mechanisms underlying the modulation of pain per-ception under hypnotic condition have been investigated bya growing number of neuroimaging studies. In PET studies,the modulatory effect of hypnosis was shown to be medi-ated by the anterior cingulate cortex (ACC) [17,48,50,49].In addition, studies have demonstrated increased modu-lation of the ACC and a large cortical and subcorticalnetwork, encompassing prefrontal, insular, and pregenualcortices, pre-SMA, thalami, striatum and brainstem in thecontext of hypnosis [17,18,48,50,49]. In an fMRI study,painful stimulation in a normal alert state resulted inbrain activation within a network encompassing cortical andsubcortical brain areas (i.e. ACC, premotor, dorsolateral,prefrontal, primary somatosensory and bilateral insular cor-tices, thalamus, bilateral striatum and brainstem) while thesame stimuli perceived under hypnosis failed to elicit anycerebral activation [63]. We also demonstrated a hypnosis-related increase in functional connectivity between primarysomatosensory cortex (S1) and anterior insular and pre-frontal cortices [63]. These results are not limited to healthyvolunteers but are also observed in pathological states,such as patients suffering from fibromyalgia or chronic pain(Table 3). In a combined EEG and fMRI study, Rainville et al.

[49] reported a reduction of the hypnosis-related increasesin occipital and delta activity when subjects were painfullystimulated.

bta

347

According to theories of hypnosis, one characteristic ofypnotic procedures is the inhibition of afferent nociceptiveransmission. This inhibition can be explained by the dra-atically decreased activity in the thalamus that is observednder hypnosis [18]. The thalamus has also been shown toorrelate with pain perception threshold while activation ofhe midline area (i.e. posterior cingulate cortex) correlatesith intensity of the stimulation and ACC with unpleasant-ess of the stimulation [61]. It has been proposed that theeported increased functional connectivity between mid-ingulate cortex, thalamus and brainstem might be relatedo pain-relevant arousal or attention mechanisms [32,50].hese observations can lead to the hypothesis that hypnosisnvolves subcortical gating processes on cortical activationhat underlies the decreased subjective pain perceptioneported by subjects under hypnosis. The basal ganglia arenown to encode and initiate basic movement patternsxpressed through premotor pathways and have also beenroposed to support basic attentional mechanisms facilitat-ng the calling up of motor programs and thoughts [18,63].n accordance with the reported decreases in premotor cor-ex activation in hypnosis, results of the different studiesuggest that hypnosis may diminish anxiety, defensive andmotional reactions to pain by reducing activation of bothortical and subcortical areas [16]. The increased modula-ion of insular activity is in line with role of this structuren pain affect and pain intensity coding [49]. Modulationf frontal area activity may reflect disruption in cogni-ive attentional, appraisal and memory systems that cannfluence perception of environmental stimulation duringypnosis [18,63]. Finally, ACC is a brain area reported ineveral studies on executive attention, detection of errors,onitoring of conflict between competing cognitive pro-esses and was shown to correlate with the difficulty ofhe task performed as well as with relaxation state of sub-ects [50]. Rainville et al. [50] proposed that engagementf the cognitive and neurophysiologic processes implied inach of those accounts may be accompanied by subjectivexperience of increased mental absorption as reported byubjects under hypnosis. In addition, ACC has also been con-idered to be involved in the suffering component ofain and affective reactions associated with pain unpleas-ntness [65]. Its decreased activity during hypnosis reflectshe decreased unpleasantness of the stimulation reportedy subjects under hypnosis. Finally, the observed reductionn occipital and delta activity during painful stimulation wasroposed to reflect disruption of relaxation and/or imageryrocesses by pain during hypnosis [49].

ypnosis as a substitute for hysteria?

ysteria, now referred to as conversion disorder, isefined as loss or distortion of a neurological function (e.g.aralysis, anesthesia, blindness, etc.) that is not explainedy any organic neurological lesion or medical disease, arisingn relation to some psychological stress or conflict, but notonsciously produced or intentionally feigned [2]. Charcoty similar brain processes. According to Charcot, dysfunc-ional processes within the central nervous system affectingctivity of motor or sensory pathways without permanent

348

A. Vanhaudenhuyse

et al.

Table 3 Modulation of the pain cerebral network during hypnosis.

Reference Technique Number of subjects Results

Healthy volunteersRainville et al. [48] PET 8 Modulation of pain-related activity in ACC correlated with unpleasantness

Rainville et al. [49] PET/EEG 8 Reduction of hypnosis-related increases in occipital and delta activity during painful stimulation

Faymonville et al. [17] PET 11 Activation in the R extrastriate area, R ACC, and corpus callosum with painful stimulation duringhypnosis (as compared to control conditions)

Hofbauer et al. [27] PET 10 Pain-related activation in R contralateral S1, S2, ACC, and insular cortex during hypnosis (analogousto that observed in normal state). Hypnotic modulation of the intensity of pain sensation led tochanges in pain-evoked activity within S1

Rainville et al. [50] PET 10 Modulation of pain-related activity in ACC, thalamus and brainstem, correlated with mentalrelaxation and mental absorption

Faymonville et al. [18] PET 19 Greater functional modulation between midcingulate cortex and bilateral insula, pregenual anteriorcingulate cortex, pre-SMA, R prefrontal cortex and striatum, thalamus and brainstem duringhypnosis (as compared to normal state)

Derbyshire et al. [13] fMRI 8 Activation of thalamus, ACC, cerebellum, S2, insula, inferior parietal and prefrontal cortices inhypnotically-induced pain (same activation observed in real stimulation condition)Additional activity in midinsula, S1, and orbitofrontal cortex, and decreased activation in ACCduring hypnotically-induced pain

Rder et al. [51] fMRI 7 Reduced activation in controlateral somatosensory, parietal, and prefrontal cortices, putamen andipsilateral amygdala during hypnosis-induced depersonalization (characterized by reduced painperception)

Vanhaudenhuyse et al. [63] fMRI 13 Non-painful sensory stimulation: activation of R S1, bilateral insula and brainstem in normalwakefulness. No activation in hypnotic statePainful stimulation: brainstem, R thalamus, bilateral striatum, R S1, bilateral insula, ACC, R middlefrontal gyrus and R premotor cortex activation in normal wakefulness. No activation in hypnotic stateRegions activating more in normal as compared to hypnosis during painful as compared tonon-painful stimulation: thalamus, bilateral striatum and ACC

PatientsWik et al. [68] PET 8 fibromyalgia Hypnosis increased activation in bilateral subcallosial cingulate gyrus, R thalamus, L inferior parietal

cortex and decreased activity in bilateral PCC and posterior part of the anterior cingulate gyrus

Derbyshire et al. [14] fMRI 13 fibromyalgia Greater activation in cerebellum, anterior midcingulate cortex and anterior and posterior withhypnosis, correlating with reported changes in pain

Abrahamsen et al. [1] fMRI 19 temporomandibulardisorders

Hypnotic hypoalgesia: activation in the posterior insula

Hypnotic hyperalgesia: activation in the R posterior insula, middle frontal gyrus and L supramarginalgyrus

Nusbaum et al. [44] PET 14 chronic low-back pain Hypnoanalgesic suggestion associated with activations in L anterior insula and nucleus accumbens,lenticular and caudate nuclei bilaterally, and ACC. Deactivations appeared in L precuneus and R PCC

fMRI: functional magnetic resonance imaging; EEG: electroencephalographic; R: right; L: left; ACC: anterior cingulate cortex; S1: primary somatosensory cortex; S2: secondarysomatosensory cortex; SMA: supplementary motor area; PCC: posterior cingulate cortex.

Neurophysiology

of hypnosis

349

Table 4 Studies on hypnosis and hysteria symptoms in healthy volunteers and patients.

Reference Technique Number of subjects Tasks Results

Hypnotically suggested motor paralysis tasks in healthy subjectsHalligan et al. [24] PET 1 Attempt movement of the

suggested paralyzed left legActivation of R orbitofrontal cortex and ACC (activation not observedwhen subject attempted to move the right non paralyzed leg)

Ward et al. [67] EMG/PET 12 Attempt to move and restduring: hypnotically suggestedparalysis of the left leg normal left leg withinstruction to feign the sameparalysis

EMG: no muscle activity during any conditionPET: suggested paralysis: increased activation in bilateral putamen, Lthalamus, L SMA, L cerebellum, R posterior medial orbitofrontal feigned paralysis: increased activation in L prefrontal, L inferiorparietal, R parietal operculum, R SMA, R ventral premotor, bilateralcerebellar attempt movement vs. rest during suggested paralysis: decreasedactivation in the R middle occipital gyrus attempt movement vs. rest during feigned paralysis: no decrease

Cojan et al. [7] fMRI 18 Go/Nogo hand movement: normal hypnotically suggestedparalysis of hand simulation of hand paralysis

Motor preparation phase: activation of contralateral M1 in allconditions. Somatosensory activation during hypnosis and simulation forhand movement. Activation of precuneus during hypnosis for handmovement Motor execution: L M1 and cerebellum activation for R handmovement in all conditions

Deeley et al. [11] fMRI 8 Hands movements in: normal hypnotically suggestedparalysis of hand

Normal condition: activation in controlateral SMA and primarysensorimotor cortices and ipsilateral cerebellum Paralysis condition: activation in bilateral SMA, R precentral gyrus, RACC, R S1, R M1, bilateral cerebellum. Deactivation of L S1 Normal > paralysis: activation in R S1, R M1, L cerebellumParalysis > normal: R SMA, bilateral cingulate gyri

Cojan et al. [8] EEG 24 Go/Nogo hand movement tasksin: normal hypnotically suggestedparalysis of left hand simulation of left handparalysis

Similar preparatory activations in all conditions (indicating preservedmotor intentions)P3 peak with specific higher activity in R inferior frontal cortex duringhypnotically suggested paralysis

Hypnotizability and conversion disordersRoelofs et al. [52] Behavioral 9 high/8 low

hypnotizable healthyImplicit and explicit mentalhand-rotation of hypnoticallysuggested paralysis of arm

Implicit: hypnotic susceptibility does not affect reaction time ofmental imagery Explicit: inability to imagine right hand rotating in 11% of highhypnotizable vs. 3% of low hypnotizable. Slowing observed for themental rotations with the paralyzed arm in high hypnotizable

350

A. Vanhaudenhuyse

et al.

Table 4 (Continued)

Reference Technique Number of subjects Tasks Results

Terhune et al. [59] Behavioral 21 low/30 highsuggestible healthy

Assessment of dissociativetendencies according hypnoticsusceptibility

Highly suggestible subjects were more responsive to hallucinationsuggestions, involuntariness, impairment of working memory capacityand reported greater pathological dissociative and fantasy-pronenesssymptoms

Bliss et al. [6] Behavioral 17 patients with severehysteria/49 controls

Hypnotic susceptibility Significant higher rate of hypnotizability in patients as compared tocontrols

Kuyk et al. [35] Behavioral 17 epileptic/20pseudo-epilepticpatients

Hypnosis used to recall ictusmemories

Higher rate of hypnotizability in pseudo-epileptic vs. epileptic patients

Goldstein et al. [23] Behavioral 20 pseudo-epilepticpatients/20 healthycontrols

Hypnotic susceptibilitymeasured by dissociation andabsorption levels

Higher level of dissociation in pseudo-epileptic patients than in controlsHigher level of absorption in controls as compared to patients

Litwin et al. [37] Behavioral 31 epileptic/10pseudo-epileptic/4 nonclassified patients

Dissociation, hypnotizabilityand absorption measurement

No difference in hypnotizability and absorption between epileptic andpseudo-epileptic patients80% pseudo-epileptic patients presented dissociative symptoms, while45% of epileptic patients

Roelofs et al. [53] Behavioral 50 conversion disorderspatients50 affective disorderspatients

Hypnotic susceptibility Patients with a conversion disorder were more susceptible to hypnoticsuggestions than patients with an affective disorderHypnotic susceptibility was correlated with the number of conversionsymptoms: patients who are more susceptible to hypnotic suggestionsdisplay more conversion symptoms

Hypnosis as treatment of conversion disorder symptomsMoene et al. [40] Behavioral 7 motor conversion

disorders patientsComprehensive clinicaltreatment including hypnosis

Recovery from motor disorders in all patients. Relapse of symptoms in 3patients

Moene et al. [41] Behavioral 45 motor conversiondisorder patients: 24hypnosis/21 controls

Hypnosis vs. comprehensiveprogram

Significant symptom reduction independent of the treatment conditionin all patients

Moene et al. [42] Behavioral 44 conversion disorderpatients: 20 hypnosis/24controls

Hypnosis vs. waiting listcondition (control)

Significant effect of hypnosis-based treatment in reducing symptomsand impairments in domains of physical, daily-life and social activities

fMRI: functional magnetic resonance imaging; EEG: electroencephalography; EMG: Electromyography; PET: positon emission tomography; L: left; R: right; SMA: supplementary motorarea; M1: motor cortex; ACC: anterior cingulate cortex; S1: primary sensory cortex.

d[ttadticht

tdctioolspp[haiped

C

Wsgadlnrtalitdwdseccdti

Neurophysiology of hypnosis

damage can explain hysterical losses in motor or sensoryfunctions, and could be induced by particular ideas, sug-gestion, or psychological states [66]. Several authors havehypothesized that psychological processes, such as affectiveor motivational factors, might induce distortion of sensoryand motor inputs, resulting in their exclusion from consciousawareness [66]. Kirsch mentioned that hypnotized subjectsare asked to experience paralysis, amnesia, anesthesia,involuntary movements and hallucinations. Hypnotizabilityis measured as the number of conversion and dissociationsymptoms that the person is able to display [34]. Subjectsunder hypnosis are thus suggested to consciously demon-strate motor or sensory phenomena, while patients withhysteria have unconscious fixed ideas based on unconscioussuggestions or autosuggestions that remain isolated fromthe rest of their mind and are expressed through motor orsensory disturbances [5]. According to Janet, dissociationcan be seen as a narrowing of the field of conscious-ness resulting in the compartmentalization of normallyintegrated mental functions, while Freud proposed thatdissociative processes result from a psychological defensemechanism that converts emotional distress into physicalsymptoms [5].

Evidence from neuroimaging studies indicates that pri-mary sensory as well as motor cortices processing remainfunctionally intact in patients with hysterical sensory-motor disorders, suggesting that dissociation may resultfrom disturbance of executive regions modulating attention,response selection and inhibition such as for example theprefrontal cortex [5]. Hypnotic procedures can induce symp-toms similar to those seen in hysteria and may permit betterunderstanding of neural processes of conversion disorders(see Table 4). According to Halligan et al. [24] hystericaland hypnotic paralysis share common neural systems involv-ing contralateral prefrontal regions. The authors used PET toexplore hypnotically suggested left leg paralysis as an ana-logue for conversion paralysis in one subject and showedsimilar brain activity (i.e. right ACC and right orbitofrontalcortex) to that reported in clinical conversion patient withcomparable leg paralysis [24]. Later study on a larger cohortof subjects confirmed the involvement of orbitofrontal cor-tex [67], as well as of ACC [11] in hypnotically suggestedmotor paralysis. Both of these brain areas were proposed tobe involved in active inhibition of movement by disconnect-ing premotor/prefrontal areas from primary motor cortex[45]. A recent EEG study in healthy volunteers reported adistinctive EEG topographic activity pattern during hypnoticparalysis with a specific source in right inferior frontal cor-tex, as compared to the control condition, indicating the keyrole of this brain region in executive control mechanisms [8].

In healthy volunteers, behavioral studies have shownthat highly hypnotizable subjects, as compared to sub-jects with low hypnotizability, were more responsive tosuggested hysteria symptoms such as motor paralysis [66],pathological dissociative and fantasy-proneness symptoms[59]. Results of the hypnotizability of patients with dis-sociative symptoms were more heterogeneous (Table 4).The level of hypnotizability was shown to correlate with

the number of conversion symptoms presented by patients[53]. For example, pseudo-epileptic patients (defined by thedemonstration of paroxysmal involuntary behavior patternsmimicking epileptic events) were shown to present more

D

Tc

351

issociative symptoms as compared to epileptic patients23,37]. Some studies have reported higher rates of hypno-izability in patients with conversion disorders as comparedo controls or patients with other pathological states (e.g.ffective disorders) [6,35,53], while others have shown noifference in both hypnotizability and absorption (defined ashe tendency to become fully involved in a perceptual, imag-native, or ideational experience) between patients withonversion disorders and controls [37]. A last study showedigher level of absorption in control subjects as comparedo patients [23].Finally, few studies have tested whether hypnosis-based

reatment shows effect for patients with conversion disor-ers (Table 4). A first study has shown that comprehensivelinical management (i.e. consisting of explaining the symp-oms, psychotherapy, physiotherapy and group therapy)ncluding hypnosis can have positive effects in the reductionf conversion symptoms of patients [40]. The specific effectf hypnosis alone was not studied and thus cannot be strictlyinked to the observed reduction of symptoms. An additionaltudy has shown that symptom reduction was observed inde-endently of the treatment condition (i.e. comprehensiverogram vs hypnosis added to a comprehensive program)41]. Later, significant treatment results were reported forypnosis-based therapy in patients with conversion disorderss compared to a no-treatment group [42]. These stud-es showed that hypnosis did not show additional value asart of a comprehensive treatment program but can be anffective treatment by itself for patients with conversionisorders.

onclusion

e have discussed here the use of hypnosis in clinicalettings as well as in neuroscience research, with theoal of learning more about the nature of hypnosis itselfnd its impact on sensory perceptions and pathologicalisorders such as conversion symptoms. If neuronal corre-ates of hypnotic state are still not completely understood,euroimaging studies emphasize that hypnosis results ineduced activity of the extrinsic brain network involved inhe environment and sensory perception. Findings on painnd hypnosis reinforce the idea that not only pharmaco-ogical but also psychological strategies can modulate thenterconnected network of cortical and subcortical regionshat participate in the processing of noxious stimuli andecrease significantly pain sensation in subjects. Finally,hile we cannot clearly conclude that conversion disor-ers and hypnotically suggested conversion symptoms sharetrictly the same neuropsychological processes, studies nev-rtheless suggest many parallels between the features oflinical conversion symptoms and hypnosis. Hypnosis can beonsidered as a useful analogue for creating conversion andissociation symptoms in healthy subjects, by the genera-ion and easy termination of clinically similar experiencesn laboratory conditions [46].isclosure of interest

he authors declare that they have no conflicts of interestoncerning this article.

3

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

52

eferences

[1] Abrahamsen R, Dietz M, Lodahl S, Roepstorff A, Zachariae R,Ostergaard L, et al. Effect of hypnotic pain modulation on brainactivity in patients with temporomandibular disorder pain. Pain2010;151(3):82533.

[2] American Psychiatric Association. Diagnostic and statisticalmanual of mental disorders. Washington: APA; 1994.

[3] Baghdadi G, Nasrabadi AM. Comparison of different EEG fea-tures in estimation of hypnosis susceptibility level. Comput BiolMed 2012;42(5):5907.

[4] Baghdadi G, Nasrabadi AM. EEG phase synchronization duringhypnosis induction. J Med Eng Technol 2012;36(4):2229.

[5] Bell V, Oakley DA, Halligan PW, Deeley Q. Dissociation in hys-teria and hypnosis: evidence from cognitive neuroscience. JNeurol Neurosurg Psychiatry 2011;82(3):3329.

[6] Bliss EL. Hysteria and hypnosis. J Nerv Ment Dis 1984;172(4):2036.

[7] Cojan Y, Waber L, Schwartz S, Rossier L, Forster A, VuilleumierP. The brain under self-control: modulation of inhibitory andmonitoring cortical networks during hypnotic paralysis. Neuron2009;62(6):86275.

[8] Cojan Y, Archimi A, Cheseaux N, Waber L, Vuilleumier P.Time-course of motor inhibition during hypnotic paralysis:EEG topographical and source analysis. Cortex 2013;49(2):42336.

[9] de Pascalis V. Psychophysiological correlates of hypnosis andhypnotic susceptibility. Int J Clin Exp Hypn 1999;47(2):11743.

10] Deeley Q, Oakley DA, Toone B, Giampietro V, Brammer MJ,Williams SC, et al. Modulating the default mode network usinghypnosis. Int J Clin Exp Hypn 2012;60(2):20628.

11] Deeley Q, Oakley DA, Toone B, Bell V, Walsh E, Marquand AF,et al. The functional anatomy of suggested limb paralysis. Cor-tex 2013;49(2):41122.

12] Demertzi A, Soddu A, Faymonville M, Bahri M, Gosseries O, Van-haudenhuyse A, et al. Hypnotic modulation of resting state fMRIdefault mode and extrinsic network connectivity. Prog BrainRes 2011;193:30922.

13] Derbyshire SW, Whalley MG, Stenger VA, Oakley DA. Cerebralactivation during hypnotically induced and imagined pain. Neu-roimage 2004;23(1):392401.

14] Derbyshire SW, Whalley MG, Oakley DA. Fibromyalgia pain andits modulation by hypnotic and non-hypnotic suggestion: anfMRI analysis. Eur J Pain 2009;13(5):54250.

15] Faymonville ME, Fissette J, Mambourg PH, Roediger L, Joris J,Lamy M. Hypnosis as adjunct therapy in conscious sedation forplastic surgery. Reg Anesth 1995;20(2):14551.

16] Faymonville ME, Mambourg PH, Joris J, Vrijens B, FissetteJ, Albert A, et al. Psychological approaches during conscioussedation. Hypnosis versus stress reducing strategies: a prospec-tive randomized study. Pain 1997;73(3):3617.

17] Faymonville ME, Laureys S, Degueldre C, DelFiore G, Luxen A,Franck G, et al. Neural mechanisms of antinociceptive effectsof hypnosis. Anesthesiology 2000;92(5):125767.

18] Faymonville ME, Roediger L, Del Fiore G, Delgueldre C, PhillipsC, Lamy M, et al. Increased cerebral functional connectivityunderlying the antinociceptive effects of hypnosis. Brain ResCogn Brain Res 2003;17(2):25562.

19] Faymonville ME, Boly M, Laureys S. Functional neuroanatomyof the hypnotic state. J Physiol Paris 2006;99(46):4639.

20] Faymonville ME, Vogt BA, Maquet P, Laureys S. Hypnosis andcingulate-mediated mechanisms of analgesia. In: Vogt BA,editor. Cingulate neurobiology & disease. Oxford: Oxford Uni-

versity Press; 2006. p. 381400.

21] Fingelkurts AA, Kallio S, Revonsuo A. Cortex functional con-nectivity as a neurophysiological correlate of hypnosis: an EEGcase study. Neuropsychologia 2007;45(7):145262.

[

A. Vanhaudenhuyse et al.

22] Fox MD, Snyder AZ, Vincent JL, Corbetta M, Van Essen DC,Raichle ME. The human brain is intrinsically organized intodynamic, anticorrelated functional networks. Proc Natl AcadSci U S A 2005;102(27):96738.

23] Goldstein LH, Drew C, Mellers J, Mitchell-OMalley S, OakleyDA. Dissociation, hypnotizability, coping styles and health locusof control: characteristics of pseudoseizure patients. Seizure2000;9(5):31422.

24] Halligan PW, Athwal BS, Oakley DA, Frackowiak RS. Imaginghypnotic paralysis: implications for conversion hysteria. Lancet2000;355(9208):9867.

25] Hinterberger T, Schoner J, Halsband U. Analysis of electrophys-iological state patterns and changes during hypnosis induction.Int J Clin Exp Hypn 2011;59(2):16579.

26] Hoeft F, Gabrieli JD, Whitfield-Gabrieli S, Haas BW, Bammer R,Menon V, et al. Functional brain basis of hypnotizability. ArchGen Psychiatry 2012;69(10):106472.

27] Hofbauer RK, Rainville P, Duncan GH, Bushnell MC. Cortical rep-resentation of the sensory dimension of pain. J Neurophysiol2001;86(1):40211.

28] Horton JE, Crawford HJ, Harrington G, Downs 3rd JH. Increasedanterior corpus callosum size associated positively with hyp-notizability and the ability to control pain. Brain 2004;127(Pt8):17417.

29] Jamieson GA, Dwivedi P, Gruzelier JH. Changes in mismatchnegativity across pre-hypnosis, hypnosis and post-hypnosis con-ditions distinguish high from low hypnotic susceptibility groups.Brain Res Bull 2005;67(4):298303.

30] Jensen MP, Sherlin LH, Askew RL, Fregni F, Witkop G, Gianas A,et al. Effects of non-pharmacological pain treatments on brainstates. Clin Neurophysiol 2013;124(10):201624.

31] Kallio S, Revonsuo A, Lauerma H, Hamalainen H, Lang H. TheMMN amplitude increases in hypnosis: a case study. Neuroreport1999;10(17):357982.

32] Kinomura S, Larsson J, Gulyas B, Roland PE. Activation byattention of the human reticular formation and thalamicintralaminar nuclei. Science 1996;271(5248):5125.

33] Kirenskaya AV, Novototsky-Vlasov VY, Zvonikov VM. Wak-ing EEG spectral power and coherence differences betweenhigh and low hypnotizable subjects. Int J Clin Exp Hypn2011;59(4):44153.

34] Kirsch I. Changing expectations: a key to effective psychother-apy. California: Brooks; 1990.

35] Kuyk J, Spinhoven P, van Dyck R. Hypnotic recall: a positivecriterion in the differential diagnosis between epileptic andpseudoepileptic seizures. Epilepsia 1999;40(4):48591.

36] Lipari S, Baglio F, Griffanti L, Mendozzi L, Garegnani M, Motta A,et al. Altered and asymmetric default mode network activity ina hypnotic virtuoso: an fMRI and EEG study. Conscious Cogn2012;21(1):393400.

37] Litwin R, Cardena E. Demographic ans seizure variables,but not hypnotizability or dissociation, differentiated psy-chogenic from organic seizures. J Trauma Dissociation 2001;1:99122.

38] Maquet P, Faymonville ME, Degueldre C, Delfiore G, Franck G,Luxen A, et al. Functional neuroanatomy of hypnotic state. BiolPsychiatry 1999;45(3):32733.

39] McGeown WJ, Mazzoni G, Venneri A, Kirsch I. Hypnotic induc-tion decreases anterior default mode activity. Conscious Cogn2009;18(4):84855.

40] Moene FC, Hoogduin KA, Van Dyck R. The inpatienttreatment of patients suffering from (motor) conversion symp-toms: a description of eight cases. Int J Clin Exp Hypn1998;46(2):17190.41] Moene FC, Spinhoven P, Hoogduin KA, van Dyck R. A randomisedcontrolled clinical trial on the additional effect of hypnosisin a comprehensive treatment programme for in-patients with

http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0005http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0010http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0015http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0020http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0025http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0030http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0030http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0030http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0030http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0030http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0030http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0030http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0030http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0030http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0030http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0030http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0030http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0030http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0035http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0040http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0045http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0050http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0055http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0060http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0065http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0070http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0075http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0080http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0085http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0090http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0095http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0100http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0105http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevier.com/S0987-7053(13)00197-4/sbref0110http://refhub.elsevie