Neuropsychologia 61 (2014) 291–298

Contents lists available at ScienceDirect

Neuropsychologia journal homepage: www.elsevier.com/locate/neuropsychologia

Medial temporal lobe resection attenuates superior temporal sulcus response to faces Fredrik Åhs a,n, Jonas Engman a, Jonas Persson a, Elna-Marie Larsson b, Johan Wikström b, Eva Kumlien c, Mats Fredrikson a a

Department of Psychology, Uppsala University, Uppsala, Sweden Department of Radiology, Oncology and Radiation Science, Uppsala University, Uppsala, Sweden c Department of Neuroscience, Neurology, Uppsala University, Uppsala, Sweden b

art ic l e i nf o

a b s t r a c t

Article history: Received 2 December 2013 Received in revised form 22 May 2014 Accepted 26 June 2014 Available online 5 July 2014

Face perception depends on activation of a core face processing network including the fusiform face area, the occipital face area and the superior temporal sulcus (STS). The medial temporal lobe (MTL) is also involved in decoding facial expression and damage to the anterior MTL, including the amygdala, generally interferes with emotion recognition. The impairment in emotion recognition following anterior MTL injury can be a direct result from injured MTL circuitry, as well as an indirect result from decreased MTL modulation of areas in the core face network. To test whether the MTL modulates activity in the core face network, we used functional magnetic resonance imaging to investigate activation in the core face processing network in patients with right or left anterior temporal lobe resections (ATR) due to intractable epilepsy. We found reductions of face-related activation in the right STS after both right and left ATR together with impaired recognition of facial expressions. Reduced activity in the fusiform and the occipital face areas was also observed in patients after right ATR suggesting widespread effects on activity in the core face network in this group. The reduction in face-related STS activity after both right and left ATR suggests that MTL modulation of the STS may facilitate recognition of facial expression. & 2014 Elsevier Ltd. All rights reserved.

Keywords: Brain Amygdala Connectivity Fusiform face area Occipital face area

1. Introduction Recognition of facial emotional expressions is impaired following damage to the medial temporal lobe (MTL) (Meletti et al., 2003; Adolphs et al., 2005) indicating a crucial role for this region in the complex process of decoding facial emotional expressions. Neuroimaging studies have identified several areas within the MTL that contribute to face processing. The region most often reported to exhibit increased activity to facial stimuli in the MTL is the amygdala (Morris et al., 1996; Whalen et al., 2001; Hariri, Tessitore, Mattay, Fera, & Weinberger, 2002), but activations of the anterior inferiotemporal cortex are also frequently reported (Kriegeskorte, Formisano, Sorger, & Goebel, 2007; Axelrod & Yovel, 2013; Nestor, Vettel, & Tarr, 2008). The functions performed by face processing MTL regions are not by themselves sufficient for classification of facial expressions. Instead, they operate as nodes in a circuit specialized in processing facial information that includes the fusiform face area (FFA), the occipital face area (OFA) and the face selective region of the superior temporal sulcus

n Correspondence to: Department of Psychology Uppsala University, Box 1225, SE-752 43 Uppsala, Sweden. Tel.: þ 46 18 4715758; fax: þ46 18 4712400.

http://dx.doi.org/10.1016/j.neuropsychologia.2014.06.030 0028-3932/& 2014 Elsevier Ltd. All rights reserved.

(STS) (Haxby, Hoffman, & Gobbini, 2002, Fairhall & Ishai, 2007). The MTL sends and receives extensive projections to all these face processing areas (Amaral, Behniea, & Kelly, 2003) and is thought to modulate the core face processing network (Kanwisher, McDermott, & Chun, 1997; Morris et al., 1998). Previous neuroimaging studies in humans have associated the OFA, the FFA and the STS with different aspects of facial expression. The OFA is thought to feed face-related information to the FFA that processes invariant aspects of the face, such as identity (Sergent, Ohta, & Macdonald, 1992) and face familiarity (Cloutier, Kelley, & Heatherton, 2011). The STS, on the other hand, is known to process variable aspects of faces, such as changes in gaze direction (Puce, Allison, Bentin, Gore, & McCarthy, 1998) or emotional expression (Johnston, Mayes, Hughes, & Young, 2013). The interaction between MTL areas and core face processing areas may hence contribute to these previously identified aspects of face processing. To our knowledge, only one study in humans has directly tested the hypothesis that the anterior MTL regions modulate face selective areas in the visual cortex (Vuilleumier, Richardson, Armony, Driver, & Dolan, 2004). Vuilleumier et al. (2004) reported bilaterally reduced activation of the FFA and other occipital areas to fearful relative neutral faces in patients with left or bilateral

292

F. Åhs et al. / Neuropsychologia 61 (2014) 291–298

sclerosis in the amygdala and hippocampus. As the study by Vuillemier et al. (2004) did not include patients with selective right hemisphere damage, it is not known whether right hemispheric damage exert similar effects. We investigated nine patients with left and eight patients with right ATR while matching angry and fearful facial expressions. We predicted that resection of the MTL would attenuate emotional recognition without having any laterality hypothesis. We further predicted decreased activation to angry and fearful faces in the core face processing network in patients with ATR as compared to healthy controls.

2. Material and methods 2.1. Subjects Seventeen patients with epilepsy who had undergone unilateral anteromedial temporal lobe resection (ATR) and 19 healthy control subjects (mean age 7SD: 46.17 14.0 years, 8 women) were recruited. Nine patients were operated on the left side (Left MTL-group, mean age7 SD: 47.7 7 9.4 years, 7 women) and 8 patients on the right side (Right MTL-group, mean age7 SD: 44.8 7 12.5 years, 4 women). The resection included amygdala and the hippocampus (Spencer, Spencer, Mattson, Williamson, & Novelly, 1984) and was performed at the Department of Neurosurgery at Uppsala University Hospital. Histopathological analysis of the resected tissue showed mesial temporal sclerosis in 12 patients, mesial temporal sclerosis and cortical dysplasia in 1, focal cortical dysplasia in 3 and ganglioglioma and focal cortical dysplasia in one. Further characteristics of the patients are shown in Table 1. All participants were administered The Mini International Neuropsychiatric Interview by a trained psychologist. Exclusion criteria were substance abuse, ongoing anxiety disorder or depression, and DSM-IV (American Psychiatric Association, 1994) axis I disorders. Subjects refrained from tobacco and alcohol intake 12 h prior to MRI assessment. The local ethics committee approved the study and all subjects gave their signed informed consent.

2.3. Magnetic resonance imaging Scanning was performed at 3 T using a Philips Achieva scanner (Philips Medical Systems, Best, The Netherlands). Structural images for anatomical reference were acquired with a T1-weighted inversion recovery sequence (60 axial slices, 2 mm slice thickness without inter slice gap, TR/TE/TI ¼ 5700/15/400 ms, FOV¼230  230 mm2, voxel size¼1  1  2 mm3). The inversion recovery sequence was used instead of a 3D T1-weighted gradient echo sequence to obtain more detailed anatomical information regarding the temporal lobe resection. Functional images were acquired with an echoplanar T2n-weighted imaging sequence and covered 30 axial slices (3 mm thick, 1 mm gap), collected in an ascending order, that encompassed the entire cerebrum and the most of the cerebellum (TR/TE ¼ 3000/35 ms, FOV¼ 230  230 mm2, flip angle¼ 901, voxel size¼3  3  3 mm3). All scanning parameters were selected to optimize the BOLD signal and maximize brain coverage.

2.3.1. Pre-processing of imaging data Preprocessing of functional brain data was performed using SPM8 (www.fil.ion. ucl.ac.uk/spm/software/spm8/). Functional images were realigned and co-registered with the anatomical images. Anatomical T1 weighted images were segmented and warped to MNI space using the New_Segment function in SPM8. No cost function masking (Brett, Leff, Rorden, & Ashburner, 2001) was applied to the resected brains as this does not improve normalization when using the unified model for segmenting and normalizing brains (Ashburner & Friston, 2005; Crinion et al., 2007; Ripolles et al., 2012). Functional images were smoothed with an 8 mm isotropic Gaussian kernel.

2.3.2. Regions of interest The location and extent of the FFA, STS and the OFA were defined based on a previous fMRI study of face processing (Gschwind, Pourtois, Schwartz, de Ville, & Vuilleumier, 2012) (Table 2). The MNI-coordinates in Table 2 presents the range in activations within the FFA, STS and OFA found by Gschwind et al. (2012). The MNIcoordinates were used to create box-shaped regions of interest (ROIs) with the coordinates in Table 2 defining the ROI-limits. The use of ROIs based on an independent data-set (Gschwind et al., 2012) reduces bias in group comparisons between the control group and resected groups. The amygdala was defined using the Talairach Daemon library in the Wake forest university pick atlas software (http://fmri.wfubmc.edu/ software/PickAtlas). ROIs were created bilaterally.

2.2. Facial emotional expression task

2.4. Statistical analysis

The face matching paradigm used was described previously by Hariri et al. (2002). This paradigm robustly activates the MTL and core face processing areas (Sergerie, Chochol, & Armony, 2008; Sabatinelli et al., 2011). Participants were asked to match one of two simultaneously presented facial expressions with a target facial expression and a matching sensorimotor task was used as a control condition. All faces were taken from the Ekman and Friesen series and expressed either fear or anger, since previous studies have shown that patients with temporal lobe epilepsy are impaired at recognizing these emotions (Meletti et al., 2003). As a sensorimotor control task, the subjects were asked to match geometric shapes. All in all, the paradigm consisted of 9 blocks: 4 blocks of facial expressions and 5 blocks of geometrical shapes, each lasting 32 s for a total scan length of 288 s. Each block began with a brief (2 s) instruction statement, “Match Faces” or “Match Forms”, and consisted of 6 target images. For each face block, 3 images of each gender and target affect (angry or fearful) were presented. For each control block, 2  3 different geometric shapes were presented as targets. All images were shown sequentially, with no inter-stimulus interval, for a period of 5 s. During imaging, subjects responded by pressing one of two buttons with their left or right hand, to determine response accuracy.

Independent t-tests comparing group differences in accuracy in recognizing fearful and angry facial expressions were performed in SPSS (PASW Statistics for Windows, Version 18.0. Chicago). Statistical analysis of fMRI data was performed in SPM8 (www.fil.ion.ucl.ac.uk/spm/software/spm8/). The functional scans for each participant were subjected to a first level analysis comparing blocks of face matching to blocks of matching shapes. Contrast images were entered into a second level analysis to compute statistical maxima within each ROI for each group (Control group, Left MTL-group, Right MTL-group), as well as the difference between the control group and each of the resected groups. Calculations for spatial extent correction for multiple comparisons were done using the REST AlphaSim utility (www.restfmri.net; toolkit V1.3), which performs simulations in the same manner as AlphaSim implemented in the AFNI software (http://afni.nimh.nih.gov/ pub/dist/doc/manual/AlphaSim.pdf). Cluster size probability levels were computed within each a priori mask (Table 2) with 1000 Monte Carlo simulations, an individual voxel threshold probability of po 0.05, a cluster connection radius of 5, and a 10 mm full width half maximum smoothness. AlphaSim was also used to calculate cluster size probability levels over the whole brain for exploratory group comparisons.

Table 1 Demographic and clinical characteristics of patients and controls. Clinical status was evaluated using the Engel grading, with higher numbers indicating more pathology.

Table 2 Regions of interest (ROI's) in the fusiform face area (FFA), the occipital face area (OFA) and the superior temporal sulcus (STS). ROI

Mean age (SD) N Sex (Female/Male) Handedness (Right/Left) Mean age at onset (SD) Mean time since surgery (SD) Outcome (Engel class I/II/III)

Control

Right MTL

Left MTL

46.1 (14.0) 19 10/9 19/0 NA NA NA

47.7 (9.4) 8 4/4 7/1 13.1 (8.1) 6.7 (4.0) 5/2/1

44.8 (12.5) 9 7/2 7/2 13.6 (7.5) 13.4 (3.5) 5/3/1

FFA OFA STS

Hemisphere

Right Left Right Left Right Left

MNI coordinates x-range

y-range

34 to 48  44 to  34 30 to 50  52 to  34 48 to 68  66 to  42

 64  66  94  90  64  62

to to to to to to

z-range  36  38  68  62  36  26

 26 to  28 to  18 to  18 to 2 to 20 4 to 26

 12  14 4 2

F. Åhs et al. / Neuropsychologia 61 (2014) 291–298

2.5. Estimation of extent of MTL resection Resected areas were manually delineated on each patient's structural scan using ITK-snap (Yushkevich et al., 2006), and converted to MNI space using the same parameters as those for functional and anatomical scans following segmentation in SPM8. Resection volumes were compared using independent t-test in SPSS. Resections were smoothed using a 4 mm full width half maximum Gaussian kernel, summed using ImCalc within SPM8. The resection overlap was visualized overlayed on an anatomical template using MRicron (Rorden, Karnath, & Bonilha, 2007) (Fig. 1). To quantify the impact of the resection on face related activity on the resected side, the left and right MTL-groups were compared to the control group in SPM8 using the sum of the resections as an inclusive mask. The significance level was set to 10 contiguous voxels surviving an uncorrected probability level of p o 0.005.

3. Results 3.1. Discrimination of angry and fearful expressions Because there was no difference between right and left resected groups in face matching (p 40.67), all patients were pooled and compared to controls. The control group performed better at matching angry and fearful faces than patients [Mean accuracy (SD) controls: 90.3%(9.6); patients: 79.6%(19.0); t33 ¼2.16, p ¼0.04] confirming that MTL resections impaired recognition of facial expressions. 3.2. Extent of MTL resection There were no differences between the left and right MTL group in resection volume (Right MTL: 15.26 76.21 cm3; Left MTL: 14.267 6.25 cm3; p4 0.74) (see Fig. 1). However, two patients in the group with resections to the left MTL did not have any amygdala damage. Therefore, ROI-based group comparisons were performed both with and without these two participants. When compared to the control group, the right MTL-group exhibited less face related activation in the right MTL (see Fig. 2; MNI-coordinates: 32,  13,  15; Z ¼2.90, k ¼27) and the left MTL-

293

group in the left MTL (MNI-coordinates:  22,  7,  20; Z ¼2.89, k¼ 10). No difference in face reactivity was observed in the intact hemisphere of either patient group relative the control group suggesting that the effect of the resection on MTL activation was restricted to the resected side.

3.3. Reactivity to threatening faces in the core face network Face matching evoked greater activity than shape matching in all a priori core face regions (OFA, FFA, STS) in the control group (see Fig. 3 and Table 3 for within-group statistics). To test the central hypothesis of attenuated activity in core face regions following resection, we compared face related activity in these regions between resected and healthy individuals. We observed reduced activation in the right STS in both left and right MTLgroups when compared to controls (see Fig. 4 and Table 4 for between-group statistics). Underscoring this pattern of attenuated activity in the STS, the MTL-groups failed to activate STS also in the within subjects comparison (Table 3). Further, left sided OFA and FFA reactivity was attenuated in the right resected group relative to controls. No differences in OFA and FFA areas emerged when the left resected group was compared to the control group (Table 4). However, when the two patients without any amygdala damage in the left resected group were excluded from the analysis, the left resected patients with amygdala damage exhibited reduced face reactivity in the right FFA relative the control group (x,y,z ¼36,  39,  18; Z¼3.36; k ¼434; p o0.001). This result indicates that amygdala damage due to MTL resection may specifically contribute to reduced face-reactivity in the right FFA. The reduced facereactivity in the right STS as compared with the control group remained significant when the two patients without amygdala damage were removed from the analysis (x,y,z¼48,  48,13; Z¼2.18; k ¼260; p o0.01). We next compared patients with resections to the left MTL to patients with right-sided MTL-resection. We observed reduced face-reactivity in the left OFA, FFA and STS of right resected relative left resected patients (Table 4). No face-reactivity reductions in the left resected patients as compared to the right resected patients reached statistical significance (Table 4). However, when the two individuals without amygdala damage were removed from the analysis, we observed reduced face reactivity in the right FFA of left resected relative right resected patients (x,y,z ¼38,  37,  21; Z¼3.53; k ¼315; p ¼0.001). Reduced face-reactivity in the FFA was thus present on the side contra-lateral to the resection in both patient groups when only patients with damage to the amygdala were considered.

3.4. Whole brain group-comparisons

Fig. 1. Anatomical extent of the resection in the left and right hemispheres. Top panels show resection overlap overlaid on an anatomical image in the left MTLgroup. Lower panels display the resection overlap in the right MTL-group. MTL, medial temporal lobe; L, left; R, right.

To identify possible effects of MTL-resection outside of the core face areas, comparisons between groups were computed across all brain voxels. When the control group was compared to the right resected group of patients, reduced face reactivity was observed in two clusters of voxels with statistical maxima in the right lateral globus pallidus and the precuneus (Table 5). Reduced facereactivity in the left resected group relative the control group was found in the left precentral gyrus (Table 5). When resected groups were compared, we observed decreased face reactivity in the right precuneus and thalamus and left superior temporal gyrus in right resected patients (Table 5). No reductions in face-reactivity were observed in left resected as compared to right resected patients.

294

F. Åhs et al. / Neuropsychologia 61 (2014) 291–298

Fig. 2. Effect of uni-lateral MTL-resection on face related activity in the MTL. A) Coronal section showing impaired face related activity in the right MTL of right (blue) resected patients at the level of the hippocampus when compared to the control group. B) Left (red) resected patients exhibited reduced face related activity in the hippocampus bordering to the amygdala in the left MTL. C) Bars display group means and standard errors in the right MTL cluster. D) Bars display group means and standard errors in the right MTL cluster. MTL, medial temporal lobe. (For interpretation of the references to color in this figure legend,the reader is referred to the web version of this article).

4. Discussion We found attenuated responses to angry and fearful facial expressions in right STS following unilateral resection of the anterior portion of the left or right MTL. The reduction in face related STS activity suggests a modulatory effect of anterior MTL, including the amygdala, on STS activity (Kanwisher et al., 1997; Haxby et al., 2002). Because MTL-resection also impaired recognition of angry and fearful expressions, our results suggest that removal of the anterior MTL lead to reduced emotion recognition by eliminating functions performed by the anterior MTL in concert with the STS. Behavioral lesion studies that have demonstrated impaired recognition of facial emotions in patients with temporal lobe epilepsy (Meletti et al., 2003; Bonora et al., 2011) have traditionally ascribed behavioral performance decrements to functions localized in the damaged brain area. This assumption may be correct, but a complementary explanation for reduced emotion recognition is that lesions in the anterior MTL affect brain functions in intact face processing areas (Vuilleumier & Pourtois, 2007). Interactive brain mechanisms may be specifically important when decoding complex and dynamic environmental stimuli (Bassett et al., 2011), like emotional facial expressions, that may change rapidly and therefore needs to be processed in parallel by

multiple brain systems. The capacity of the STS to decode complex and rapidly changing visual characteristics may therefore depend on its interaction with other brain areas like the amygdala (Allison, Puce, & McCarthy, 2000), with which it shares rich connections (Stefanacci & Amaral, 2000). Indeed, effective connectivity analysis suggests that the amygdala modulates STS activity during encoding of facial information (Furl, Henson, Friston, & Calder, 2013). MTL resection, including parts of the amygdala, might thus reduce afferent input to the STS, which could explain the here observed reduction in face related activity in the STS. Reduced face reactivity in the STS was observed in the right hemisphere after both left and right resections. The finding of a reduction of STS reactivity in the right hemisphere independent of resection side suggests both ipsilateral and contralateral modulation of STS activity by the anterior MTL. This observation is in line with the bilateral functional connectivity pattern of MTL regions such as the amygdala observed in resting state fMRI studies (Roy et al., 2009), suggesting that MTL resection on either side can produce bilateral activity alterations in areas modulated by the MTL. A recent report investigating the connectivity pattern between areas in the core face network also suggests stronger functional connectivity between the left and right core face processing areas than between core face processing areas within each hemisphere (Davies-Thompson & Andrews, 2012). The strong inter-hemispheric connectivity of core

F. Åhs et al. / Neuropsychologia 61 (2014) 291–298

295

Fig. 3. Activation in the core face network to angry and fearful facial expressions as compared to geometrical shapes. A) The control group activated all regions of the core face network. B) Patients with resections to the right MTL showed activations in the FFA and the OFA but not the STS. C) Patients with left sided resections also activated the FFA and the OFA and failed to activate the STS. Gray shaded areas represent regions of interests. Statistical images are thresholded at p o 0.05 uncorrected. FFA, fusiform face area; MTL, medial temporal lobe; OFA, occipital face area; STS, superior temporal sulcus; L, left; R, right.

Table 3 Statistics for the contrast Faces 4Shapes within each group and region of interest in the core face network. Group

Control x,y,z (MNI) Z voxels (n) p Right MTL x,y,z (MNI) Z voxels (n) p Left MTL x,y,z (MNI) Z voxels (n) p

FFA

OFA

STS

Left

Right

Left

Right

Left

Right

 38  59  24 7.08 477 o 0.001nnn

38  57  18 7.50 441 o 0.001nnn

 33  81  6 7.35 696 o 0.001nnn

41  68  15 7.72 290 o 0.001nnn

 57  42 6 3.37 264 0.05

57  42 15 4.32 744 o 0.001nnn

 39  48  18 6.22 477 o 0.001nnn

38  52  18 6.01 432 o 0.001nnn

 36  75  15 5.39 605 o 0.001nnn

39  67  17 5.88 290 o 0.001nnn

 56  55 12 3.00 256 0.06

59  42 3 3.17 133 0.07

 41  48  20 6.21 477 o 0.001nnn

38  55  18 6.01 422 o 0.001nnn

 33  84  6 5.88 690 o 0.001nnn

32  75  17 6.21 290 o 0.001nnn

 44  46 27 2.03 39 0.57

51  63 4 2.28 28 0.38

FFA, fusiform face area; MNI, Montreal Neurological Institute; MTL, medial temporal lobe; OFA, occipital face area; STS, superior temporal sulcus.

face processing regions implies that reduced ipsilateral MTL modulation of core face areas indirectly could affect activity in the same areas on the contralateral side. This idea has been advocated previously to explain normal FFA activation to faces in a patient with a unilateral lesion in the OFA, which is thought to feed face information forward to the FFA (Rossion et al., 2003). Another possibility is that MTL modulation of the contralateral STS proceeds directly through commissural efferents or through a combination of

direct and indirect routes. The group differences in the STS were restricted to the right hemisphere which might be related to the often observed greater face-related responses in the right STS than the left STS. The right STS might therefore be more strongly modulated by the MTL than the left STS during face processing. Whereas impaired reactivity in the core face network was present in the right STS both in left and right MTL-groups, patients operated on the right side also exhibited reduced reactivity in the

296

F. Åhs et al. / Neuropsychologia 61 (2014) 291–298

Fig. 4. Effect of uni-lateral MTL-resection on face related activity in the core face network. A) Activation in the right STS of patients with both right (blue) and left (red) sided MTL resections was reduced relative to controls. B) Patients with right sided MTL resections further exhibited reduced activations in the left OFA relative to the control group. C) Patients with right sided MTL resections also displayed attenuated activation of the left FFA relative to the control group. Gray shaded areas represent the regions of interests as defined in Table 2. Statistical images are thresholded at p o0.05 uncorrected. FFA, fusiform face area; MTL, medial temporal lobe; OFA, occipital face area; STS, superior temporal sulcus; L, left; R, right. (For interpretation of the references to color in this figure legend,the reader is referred to the web version of this article). Table 4 Statistics for group comparisons within core face areas. FFA

Control Vs. Right x,y,z (MNI) Z voxels (n) p Control Vs. Left x,y,z (MNI) Z voxels (n) p Left Vs. Right x,y,z (MNI) Z voxels (n) p Right Vs. Left x,y,z (MNI) Z voxels (n) p

OFA

STS

Left

Right

Left

Right

Left

Right

 38  61  26 2.71 117 0.01n

45  63  12 2.77 62 0.08

 33  78  3 3.04 261 0.005nn

44  69  9 3.18 73 0.10

 51  25 7 2.34 188 0.11

59  43 15 2.80 183 0.04n

– – 0 –

– – 0 –

– – 0 –

– – 0 –

 60  39 18 2.55 191 0.11

57  40 16 2.47 201 0.03n

 39  66  25 2.55 92 0.02n

– – 0 –

 37  76  1 2.82 265 0.005nn

31  75  16 2.28 104 0.06

 57  25 7 3.13 406 0.01n

– – 0 –

– – 0 –

37  37  21 2.13 67 0.08

– – 0 –

– – 0 –

 61  40 19 2.30 158 0.15

– – 0 –

FFA, fusiform face area; MNI, Montreal Neurological Institute; OFA, occipital face area; STS, superior temporal sulcus.

Table 5 Whole brain group comparisons. Comparison

Area

x,y,z (MNI)

Z

voxels (n)

p

Control4 Right MTL

Lateral Globus pallidus Precuneus (BA31) Precentral gyrus (BA6) Precuneus (BA31) Superior temporal gyrus (BA 22) Thalamus

 15, 0, 7 19,  69, 21  27,  19, 64 22,  69, 18  55,  16, 1 27,  28, 6 –

4.19 3.61 3.12 3.79 3.92 3.57 –

32916 3930 4641 11915 5722 4114 –

o 0.01nn 0.05n 0.03n o 0.01nn 0.01n 0.03n –

Control4 Left MTL Left MTL4Right MTL

Right MTL 4Left MTL

BA, Brodmann area; MNI, Montreal Neurological Institute; MTL, Medial temporal lobe.

left FFA and OFA. These results suggests that resection to the right as compared to left MTL results in more widespread reactivity reductions in the core face processing network. Previous reports in patients with split brain has indicated right lateralized specialization for face perception (Gazzaniga, 2000) and neuroimaging

studies also suggest stronger reactivity to faces in the right as compared to the left FFA (Yovel, Tambini, & Brandman, 2008). However, at the level of the amygdala, no distinct pattern of hemispheric specialization has emerged when evaluating responsivity to emotional presentations (Sergerie et al., 2008). The lack of

F. Åhs et al. / Neuropsychologia 61 (2014) 291–298

clear lateralization in amygdala reactivity to faces does not rule out the possibility that the left and right amygdala can modulate core face processing areas differently, an idea that would fit the present observation. The observed group differences are also in line with a previous fMRI study on working memory in patients with ATR, which also found different activation patterns in left and right resected patients (Stretton et al., 2014). The more widespread reductions in neural reactivity in patients with right ATR did however not correspond to behavior, because the accuracy in face matching did not differ between patients with right and left sided resection. Future studies could aim at describing the functional significance of differences in modulation of core face areas by the right and left anterior MTL. In an exploratory whole brain analysis, we found reduced face reactivity in the striatum and the parietal cortex of right resected patients, whereas face reactivity was reduced in premotor areas of left resected patients. As in our ROI-analysis, the reductions in the right resected group were more widespread than in the left resected group. Previous neuroimaging studies of face processing have also found activations in the superior parietal cortex (Stretton et al., 2014), motor areas (Kilts, Egan, Gideon, Ely, & Hoffman, 2003) and the striatum (Fusar-Poli et al., 2009). Our findings suggest that face-related activity in these areas is modulated by the MTL. Anterior temporal lobe resections included not only the amygdala but also hippocampus and neocortical structures. Therefore, the observed reduction in STS reactivity could be explained by impaired modulation of STS activity by other structures than the amygdala. However, lesions to the amygdala are sufficient to block modulation of STS activity when viewing facial expressions of conspecifics in monkeys (Hadj-Bouziane et al., 2012). Further, when two left-resected individuals with intact amygdala were excluded from the comparison with the control group, we observed reduced face reactivity in the right FFA in left resected patients with amygdala damage in addition to the right STS. This observation suggests, but does not prove, that our finding of reduced face-activity in resected patients may be attributed to reduced amygdala modulation. It should be noted that we here only investigated neural responses to angry and fearful facial expressions. Our finding of reduced STS reactivity following MTL resection could therefore be limited to these two expressions. However, robust amygdala responses have previously been found to surprised and neutral (Ahs, Davies, Gorka, & Hariri, 2013) as well as happy and sad facial expressions (Fitzgerald, Angstadt, Jelsone, Nathan, & Phan, 2006). The amygdala hence seems to respond to a spectrum of facial expressions rather than being selective to one, or a few categories. Likewise, the STS is thought to process variable features of facial expressions that are present in multiple emotion categories, such as emotional expression (Johnston et al., 2013) or direction of gaze (Alison et al., 2000). It therefore seems likely that MTL modulation of STS activity spans multiple categories of facial expressions. Further, modulation of the OFA, FFA and STS by the MTL may be more pronounced during processing of facial information than during processing of other emotional material, as a previous report investigating neural responses to emotional scenes found preserved emotion-enhanced activation in visual areas following MTL-resesection (Edmiston et al., 2013). Our study was limited to the visual domain. The STS and the amygdala however process stimuli from multiple modalities (Mesulam, 1998; Phillips et al., 1998), and resections to the anterior MTL therefore could affect STS activation to emotional stimuli also from other sensory domains. Behaviorally, patients with temporal lobe epilepsy and hence compromised MTL functionality (Bonora et al., 2011) show impaired identification of emotional qualities in vocal stimuli. Given that our results can be

297

extrapolated to the auditory domain it could be predicted that the attenuated discrimination of vocal affect reported by Bonora et al. (2011) in part is due to reduced modulation of STS activity following damage to the anterior temporal lobe. In conclusion, we found that reduced modulation of STS activity following resections of the anterior MTL was paralleled by compromised emotion recognition. These results emphasize the importance of MTL-STS pathways for successful decoding of facial expressions. Of the resected MTL areas, we speculate that the amygdala may be especially important in modulating face-related STS activity. Our findings may be relevant for understanding aberrant MTL-connectivity in autism (Kleinhans et al., 2008) and social anxiety disorder (Frick, Howner, Fischer, Kristiansson, & Furmark, 2013).

Acknowledgments We thank Sara Pankowski, Victoria Trepp and Johanna Hoppe for help with data collection. This study was funded by the Swedish Research Council, the Swedish Brain Foundation and the Selander foundation.

References Adolphs, R., Gosselin, F., Buchanan, T. W., Tranel, D., Schyns, P., & Damasio, A. R. (2005). A mechanism for impaired fear recognition after amygdala damage. Nature, 433, 68–72. Ahs, F., Davies, C. F., Gorka, A. X., & Hariri, A. R. (2013). Feature-based representations of emotional facial expressions in the human amygdala. Social Cognitive and Affective Neuroscience, http://dx.doi.org/10.1093/scan/nst112 ([Epub ahead of print]). Allison, T., Puce, A., & McCarthy, G. (2000). Social perception from visual cues: role of the STS region. Trends in Cognitive Sciences, 4, 267–278. Amaral, D. G., Behniea, H., & Kelly, J. L. (2003). Topographic organization of projections from the amygdala to the visual cortex in the macaque monkey. Neuroscience, 118, 1099–1120. Ashburner, J., & Friston, K. J. (2005). Unified segmentation. Neuroimage, 26, 839–851. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders. Washington, DC. Axelrod, V., & Yovel, G. (2013). The challenge of localizing the anterior temporal face area: a possible solution. Neuroimage, 81, 371–380. Bassett, D. S., Wymbs, N. F., Porter, M. A., Mucha, P. J., Carlson, J. M., & Grafton, S. T. (2011). Dynamic reconfiguration of human brain networks during learning. Proceedings of the National Academy of Sciences United States of America, 108, 7641–7646. Bonora, A., Benuzzi, F., Monti, G., Mirandola, L., Pugnaghi, M., Nichelli, P., et al. (2011). Recognition of emotions from faces and voices in medial temporal lobe epilepsy. Epilepsy & Behavior, 20, 648–654. Brett, M., Leff, A. P., Rorden, C., & Ashburner, J. (2001). Spatial normalization of brain images with focal lesions using cost function masking. Neuroimage, 14, 486–500. Cloutier, J., Kelley, W. M., & Heatherton, T. F. (2011). The influence of perceptual and knowledge-based familiarity on the neural substrates of face perception. Social Neuroscience, 6, 63–75, http://dx.doi.org/10.1080/17470911003693622. Crinion, J., Ashbumer, J., Leff, A., Brett, M., Price, C., & Friston, K. (2007). Spatial normalization of lesioned brains: Performance evaluation and impact on fMRI analyses. Neuroimage, 37, 866–875. Davies-Thompson, J., & Andrews, T. J. (2012). Intra- and interhemispheric connectivity between face-selective regions in the human brain. Journal of Neurophysiology, 108, 3087–3095. Edmiston, E. K., McHugo, M., Dukic, M. S., Smith, S. D., Abou-Khalil, B., Eggers, E., et al. (2013). Enhanced visual cortical activation for emotional stimuli is preserved in patients with unilateral amygdala resection. Journal of Neuroscience, 33, 11023–11031, http://dx.doi.org/10.1523/jneurosci.0401-13.2013. Fairhall, S. L., & Ishai, A. (2007). Effective connectivity within the distributed cortical network for face perception. Cerebral Cortex, 17, 2400–2406. Fitzgerald, D. A., Angstadt, M., Jelsone, L. M., Nathan, P. J., & Phan, K. L. (2006). Beyond threat: amygdala reactivity across multiple expressions of facial affect. Neuroimage, 30, 1441–1448. Frick, A., Howner, K., Fischer, H., Kristiansson, M., & Furmark, T. (2013). Altered fusiform connectivity during processing of fearful faces in social anxiety disorder. Translational Psychiatry, 3, e312, http://dx.doi.org/10.1038/tp.2013.85. Furl, N., Henson, R. N., Friston, K. J., & Calder, A. J. (2013). Top-down control of visual responses to fear by the amygdala. Journal of Neuroscience, 33, 17435–17443, http://dx.doi.org/10.1523/jneurosci.2992-13.2013.

298

F. Åhs et al. / Neuropsychologia 61 (2014) 291–298

Fusar-Poli, P., Placentino, A., Carletti, F., Landi, P., Allen, P., Surguladze, S., et al. (2009). Functional atlas of emotional faces processing: a voxel-based metaanalysis of 105 functional magnetic resonance imaging studies. Journal of Psychiatry Neuroscience, 34, 418–432. Gazzaniga, M. S. (2000). Cerebral specialization and interhemispheric communication - Does the corpus callosum enable the human condition? Brain, 123, 1293–1326. Gschwind, M., Pourtois, G., Schwartz, S., de Ville, D. V., & Vuilleumier, P. (2012). White-matter connectivity between face-responsive regions in the human brain. Cerebral Cortex, 22, 1564–1576. Hadj-Bouziane, F., Liu, N., Bell, A. H., Gothard, K. M., Luh, W-M., Tootell, R. B. H., et al. (2012). Amygdala lesions disrupt modulation of functional MRI activity evoked by facial expression in the monkey inferior temporal cortex. Proceedings of the National Academy of Science USA, 109, E3640–E3648. Hariri, A. R., Tessitore, A., Mattay, V. S., Fera, F., & Weinberger, D. R. (2002). The amygdala response to emotional stimuli: a comparison of faces and scenes. Neuroimage, 17, 317–323. Haxby, J. V., Hoffman, E. A., & Gobbini, M. I. (2002). Human neural systems for face recognition and social communication. Biological Psychiatry, 51, 59–67. Johnston, P., Mayes, A., Hughes, M., & Young, A. W. (2013). Brain networks subserving the evaluation of static and dynamic facial expressions. Cortex, 49, 2462–2472, http://dx.doi.org/10.1016/j.cortex.2013.01.002. Kanwisher, N., McDermott, J., & Chun, M. M. (1997). The fusiform face area: a module in human extrastriate cortex specialized for face perception. Journal of Neuroscience, 17, 4302–4311. Kleinhans, N. M., Richards, T., Sterling, L., Stegbauer, K. C., Mahurin, R., Johnson, L. C., et al. (2008). Abnormal functional connectivity in autism spectrum disorders during face processing. Brain, 131, 1000–1012. Kilts, C. D., Egan, G., Gideon, D. A., Ely, T. D., & Hoffman, J. M. (2003). Dissociable neural pathways are involved in the recognition of emotion in static and dynamic facial expressions. Neuroimage, 18, 156–168. Kriegeskorte, N., Formisano, E., Sorger, B., & Goebel, R. (2007). Individual faces elicit distinct response patterns in human anterior temporal cortex. Proc Natl Acad Sci U S A, 104, 20600–20605. Meletti, S., Benuzzi, F., Rubboli, G., Cantalupo, G., Maserati, M. S., Nichelli, P., et al. (2003). Impaired facial emotion recognition in early-onset right mesial temporal lobe epilepsy. Neurology, 60, 426–431. Mesulam, M. M. (1998). From sensation to cognition. Brain, 121, 1013–1052. Morris, J. S., Frith, C. D., Perrett, D. I., Rowland, D., Young, A. W., Calder, A. J., et al. (1996). A differential neural response in the human amygdala to fearful and happy facial expressions. Nature, 383, 812–815. Morris, J. S., Friston, K. J., Buchel, C, Frith, C. D., Young, A. W., Calder, A. J., et al. (1998). A neuromodulatory role for the human amygdala in processing facial expressions. Brain, 121, 47–57. Nestor, A., Vettel, J. M., & Tarr, M. J. (2008). Task-specific codes for face recognition: how they shape the neural representation of features for detection and individuation. Plos One, 3, e3978. Phillips, ML, Young, AW, Scott, SK, Calder, AJ, Andrew, C, Giampietro, V, et al. (1998). Neural responses to facial and vocal expressions of fear and disgust. Proceedings of the Royal Society B-Biological Sciences, 265, 1809–1817.

Puce, A., Allison, T., Bentin, S., Gore, J. C., & McCarthy, G. (1998). Temporal cortex activation in humans viewing eye and mouth movements. Journal of Neuroscience, 18(6), 2188–2199. Ripolles, P., Marco-Pallares, J., de Diego-Balaguer, R., Miro, J., Falip, M., Juncadella, M., et al. (2012). Analysis of automated methods for spatial normalization of lesioned brains. Neuroimage, 60, 1296–1306. Rorden, C., Karnath, H.-O., & Bonilha, L. (2007). Improving lesion-symptom mapping. Journal of Cognitive Neuroscience, 19, 1081–1088. Rossion, B., Caldara, R., Seghier, M., Schuller, A. M., Lazeyras, F., & Mayer, E. (2003). A network of occipito-temporal face-sensitive areas besides the right middle fusiform gyrus is necessary for normal face processing. Brain, 126, 2381–2395. Roy, A. K., Shehzad, Z., Margulies, D. S., Kelly, A. M. C., Uddin, L. Q., Gotimer, K., et al. (2009). Functional connectivity of the human amygdala using resting state fMRI. Neuroimage, 45, 614–626. Sabatinelli, D., Fortune, E. E., Li, Q., Siddiqui, A., Krafft, C., Oliver, W. T., et al. (2011). Emotional perception: meta-analyses of face and natural scene processing. Neuroimage, 54, 2524–2533. Sergent, J., Ohta, S., & Macdonald, B. (1992). Functional neuroanatomy of face and object processing – a positron emission tomography study. Brain, 115, 15–36. Sergerie, K., Chochol, C., & Armony, J. L. (2008). The role of the amygdala in emotional processing: a quantitative meta-analysis of functional neuroimaging studies. Neuroscience and Biobehavioral Reviews, 32, 811–830. Spencer, D. D., Spencer, S. S., Mattson, R. H., Williamson, P. D., & Novelly, R. A. (1984). Access to the posterior medial temporal lobe structures in the surgical treatment of temporal lobe epilepsy. Neurosurgery, 15, 667–671. Stefanacci, L., & Amaral, D. G. (2000). Topographic organization of cortical inputs to the lateral nucleus of the macaque monkey amygdala: a retrograde tracing study. Journal of Comparative Neurology, 421, 52–79. Stretton, J., Sidhu, M. K., Winston, G. P., Bartlett, P., McEvoy, A. W., Symms, M. R., et al. (2014). Working memory network plasticity after anterior temporal lobe resection: a longitudinal functional magnetic resonance imaging study. Brain, http://dx.doi.org/10.1093/brain/awu061. Vuilleumier, P., & Pourtois, G. (2007). Distributed and interactive brain mechanisms during emotion face perception: evidence from functional neuroimaging. Neuropsychologia, 45, 174–194. Vuilleumier, P., Richardson, M. P., Armony, J. L., Driver, J., & Dolan, R. J. (2004). Distant influences of amygdala lesion on visual cortical activation during emotional face processing. Nature Neuroscience, 7, 1271–1278. Whalen, P. J., Shin, L. M., McInerney, S. C., Fischer, H., Wright, C. I., & Rauch, S. L. (2001). A functional MRI study of human amygdala responses to facial expressions of fear versus anger. Emotion, 1, 70–83. Yovel, G., Tambini, A., & Brandman, T. (2008). The asymmetry of the fusiform face area is a stable individual characteristic that underlies the left-visual-field superiority for faces. Neuropsychologia, 46, 3061–3068. Yushkevich, P.A, Piven, J., Hazlett, H. C., Smith, R. G., Ho, S., Gee, J. C., et al. (2006). User-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability. Neuroimage, 31, 1116–1128.