Zurich Open Repository and Archive University of Zurich Main Library Strickhofstrasse 39 CH-8057 Zurich www.zora.uzh.ch

Year: 2016

Bottom-up Visual Integration in the Medial

Pflugshaupt, T ; Nösberger, M ; Gutbrod, K ; Weber, K P ; Linnebank, M ;Brugger,P

Abstract: Largely based on findings from functional neuroimaging studies, the medial parietal lobeis known to contribute to internally directed cognitive processes such as visual imagery or episodic mem- ory. Here, we present 2 patients with behavioral impairments that extend this view. Both had chronic unilateral lesions of nearly the entire medial parietal lobe, but in opposite hemispheres. Routine neuropsy- chological examination conducted >4 years after the onset of damage showed little deficits of minor severity. In contrast, both patients reported persistent unusual visual impairment. A comprehensive series of tachistoscopic experiments with lateralized stimulus presentation and comparison with healthy participants revealed partial visual hemiagnosia for stimuli presented to their contralesional hemifield, applying inferential single-case statistics to evaluate deficits and dissociations. Double dissociations were found in 4 experiments during which participants had to integrate more than one visual element, either through comparison or formation of a global gestalt. Against the background of recent neuroimaging findings, we conclude that of all medial parietal structures, the is the most likely candidate for a crucial involvement in such bottom-up visual integration.

DOI: https://doi.org/10.1093/cercor/bhu256

Posted at the Zurich Open Repository and Archive, University of Zurich ZORA URL: https://doi.org/10.5167/uzh-100459 Journal Article

Originally published at: Pflugshaupt, T; Nösberger, M; Gutbrod, K; Weber, K P; Linnebank, M; Brugger, P (2016). Bottom-up Visual Integration in the Medial Parietal Lobe. , 26(3):943-949. DOI: https://doi.org/10.1093/cercor/bhu256 Bottom-up visual integration in the medial parietal lobe

Running title

Vision in the medial parietal lobe

Authors

Tobias Pflugshaupt1,2, PhD

Myriam Nösberger1, MSc

Klemens Gutbrod3, PhD

Konrad P. Weber1,4, MD

Michael Linnebank1, MD

Peter Brugger1, PhD

Affiliations

1Department of Neurology, University Hospital Zürich, Switzerland

2Center of Neurology and Neurorehabilitation, State Hospital Luzern, Switzerland

3Department of Neurology, Inselspital, University Hospital Bern, Switzerland

4Department of Ophthalmology, University Hospital Zürich, Switzerland

Correspondence

Tobias Pflugshaupt, PhD

Center of Neurology and Neurorehabilitation

Luzerner Kantonsspital/State Hospital

6000 Luzern 16; Switzerland

Email: [email protected]; Tel: 0041 41 205 5053; Fax: 0041 41 205 2441

Page 1 of 29 Abstract

Largely based on findings from functional neuroimaging studies, the medial parietal lobe is known to contribute to internally-directed cognitive processes such as visual imagery or episodic memory.

Here we present two patients with behavioural impairments that extend this view. Both had chronic unilateral lesions of nearly the entire medial parietal lobe, but in opposite hemispheres. Routine neuropsychological examination conducted more than four years after the onset of brain damage showed little deficits of minor severity. In contrast, both patients reported persistent unusual visual impairment. A comprehensive series of tachistoscopic experiments with lateralised stimulus presen- tation and comparison with healthy participants revealed partial visual hemiagnosia for stimuli pre- sented to their contralesional hemifield, applying inferential single case statistics to evaluate deficits and dissociations. Double dissociations were found in four experiments during which participants had to integrate more than one visual element, either through comparison or formation of a global gestalt. Against the background of recent neuroimaging findings, we conclude that of all medial parietal structures, the precuneus is the most likely candidate for a crucial involvement in such bot- tom-up visual integration.

Keywords

Double dissociation; hemiagnosia; neurovisual disorder; precuneus; tachistoscopic presentation

Page 2 of 29 The medial parietal lobe of the brain extends in a rostral-caudal direction from the pars mar- ginalis of the cingulate to the occipito-parietal sulcus. It consists of the precuneus, the subpa- rietal sulcus, the posterior part of the cingulate , and the retrosplenial region. As these associa- tion areas are rarely affected by isolated focal damage (Cavanna and Trimble 2006; Leech and

Sharp 2014), their function is less known than that of their counterparts on the lateral or inferior surfaces of the . Contributing to this relative lack of knowledge, their hidden location in the longitudinal hinders experimental techniques such as transcranial magnetic stimulation.

What we know about the function of these regions in the is therefore predominantly based on functional neuroimaging studies conducted in healthy participants.

Reviewing 34 such studies, Cavanna and Trimble (Cavanna & Trimble 2006) propose that the bilateral precuneus is involved in three main areas of higher : visuo-spatial imagery, retrieval from episodic memory and - through interconnections with medial prefrontal regions - various aspects of self-processing such as first-person perspective taking or perceived agency. Fur- ther evidence from functional neuroimaging studies suggests that bilateral posterior cingulate

(Leech et al. 2012) and retrosplenial regions (Vann et al. 2009) form an essential node of the brain's default network, a system that is active during passive moments or during tasks that require partici- pants to remember past events or imagine upcoming events (Buckner 2013). More specifically, the ventral part of the posterior may be relevant for internally-directed cognition such as memory or planning, while its dorsal part seems to be implicated in the control of attentional focus (Leech and Sharp 2014). According to rare case studies summarised by Vann and colleagues

(Vann et al. 2009), damage to the typically produces amnesia (e.g., Valenstein et al. 1987; McDonald et al. 2001) and/or problems in spatial navigation (e.g., Takahashi et al.

1997; Ino et al. 2007). Finally, the subparietal sulcus has hardly ever been mentioned in clinical- anatomical correlation studies. Richer and colleagues (Richer et al. 1993) propose a role in proprio- ception, since intracerebral stimulation around this sulcus produced feelings of levitation in phar- maco-resistant patients undergoing presurgical investigation.

Page 3 of 29 Here we present two patients with chronic unilateral damage of nearly the entire medial pa- rietal lobe in opposite hemispheres. During standard neuropsychological assessment conducted more than four years after the onset of brain damage, both patients described persisting unusual visual impairment for contralesionally located stimuli. In the following, this impairment will be analysed in detail, with a series of tachistoscopic experiments forming the core of the investigation.

Materials and Methods

Patients

Demographic and clinical background

At the time of the experimental investigation, SS was a 23-year-old, right-handed, female psycholo- gy student with Swiss German as her first language. Four and a half years earlier, she had suffered left parietal intracerebral bleeding due to a ruptured arteriovenous malformation that was surgically embolised on the following day. Neurological examination two years later revealed no pathological findings. Immediately prior to our experimental investigation, uncorrected near visual acuity meas- ured at 30 cm was 1.25 and thus normal for both eyes. Visual fields were also unaffected as shown by automated static perimetry (Supplementary Fig. 1). Routine MRI of the head performed five months before the experimental investigation revealed parenchymal damage extending over large parts of the left medial parietal lobe, whereas primary and secondary visual areas in the as well as the lateral parietal cortex were spared (Fig. 1). Notably, her lesion also included the entire callosal splenium, indicated by the mere absence of the latter on sagittal slices.

[Insert Fig. 1 here]

The second patient was LK, a 43-year-old, right-handed, native German speaking male lec- turer who made a PhD in biology. Five years before the experimental investigation, an anaplastic astrocytoma was resected from his right parietal lobe. According to repeated clinical examination

Page 4 of 29 over the following years, a mild sensorimotor impairment of the left leg was the only persisting neurological symptom. Identical to SS, he displayed normal near visual acuity measured at 30 cm

(i.e. 1.25 for both eyes) and no visual field defect within the area relevant for the tachistoscopic experiments described below in automated static perimetry at the time of the experimental investi- gation (Supplementary Fig. 1). As a side remark, LK showed a 'pie on the floor' field defect in the periphery of the left lower quadrant, consistent with his right parietal lesion. Similar to SS, follow- up MRI of the head revealed brain damage affecting the entire medial parietal lobe and the callosal splenium - again sparing occipital as well as lateral parietal areas - with two major differences: In his case, the right but not the left parietal lobe was damaged, and LK's lesion dorsally extended more towards the , explaining the sensorimotor impairment of his left leg (Fig. 1).

Neuropsychological assessment

Prior to the experimental investigation described below, both patients underwent routine neuropsy- chological examination at the neurological outpatient clinic of the University Hospital Zürich, aimed at evaluating potential cognitive residua. Concerning subjective complaints, SS spontaneous- ly described fatigue, occasional headaches, slowed reading and mental arithmetic as well as im- paired recognition of right-sided visual stimuli ("I see right-sided things but have difficulties identi- fying them") as residual impairments. In contrast, LK mentioned the sensorimotor impairment of his left leg, spatial navigation problems in novel environments, impaired of duration, slowed reading and writing, rare eye hand-coordination problems when typing and impaired recog- nition of left-sided visual stimuli ("When the left-sided visual environment lacks structure, every- thing turns into mush").

A battery of standardised, norm-referenced tests examining attentional, mnestic, visuocon- structive and executive functions revealed minor impairments only. SS displayed a mild-to- moderate deficit in free recall from verbal long-term memory, LK a mildly reduced visuo-spatial span, and both patients showed a mild deficit in nonverbal response inhibition (Supplementary Ta-

Page 5 of 29 ble 1). According to additional screening tests, neither SS nor LK presented visual agnosia - tested under normal free-viewing conditions - visual hemineglect, optic ataxia, aphasia, acalculia or ideo- motor apraxia (Supplementary Table 2).

Language laterality was examined with the 'Bergen Dichotic Listening Test' (Hugdahl

1995). In the non-forced condition, SS and LK showed laterality indices/LI of +85 and +92, respec- tively (range from left to right: -100 to +100). In other words, both patients displayed a pronounced right-ear advantage for spoken syllables, which indicates left-hemisphere language dominance

(Hugdahl 2011). Moreover, handedness laterality quotients/LQ (range from left to right: -100 to

+100) based on the questionnaire developed by Salmaso and Longoni (Salmaso & Longoni 1985) were +79 in SS and +100 in SS. Both values represent strong right-handedness and are associated with a probability for left-hemisphere language dominance of 93% and 95%, respectively, applying the formula of Knecht and co-workers (Knecht et al. 2000).

Tachistoscopic experiments

In order to investigate the subjectively reported visual impairment for contralesional stimuli, a se- ries of 15 tachistoscopic experiments was conducted. During these experiments, SS and LK sat 50 cm in front of a 14'' notebook screen (31 x 24°), with a chin rest and a forehead stop ensuring con- stant distance. Basically they had to name, discriminate, or localise a large variety of visual stimuli

(e.g., colours, fragmented forms, objects, Navon letters, words, faces) briefly presented at 6.5°

(stimulus centre) to the left or right of a central fixation cross/CFC in the horizontal plane. Every experiment included four practise trials and 20 or 40 test trials that were uniformly distributed across both sides and presented in pseudo-randomised order. Each trial began with the CFC shown for 1000 ms, followed by the lateral stimulus (size range: 1-6° x 1-9°) displayed in addition to the

CFC. Then a mask of visual noise appeared for 500 ms to suppress afterimages, succeeded by an untimed response window during which patients gave oral responses. A second experimenter sitting opposite the patient carefully controlled central fixation during the presentation of the lateral stimu-

Page 6 of 29 lus. Task-specific presentation durations of the lateral stimulus (50, 100, 150, or 200 ms) were mul- tiples of the screen's refresh rate and based on task difficulty as derived from a pilot study with five healthy participants, assigning the shortest duration to the easiest tasks. Detailed description of all

15 experiments that were programmed and run with E-Prime 2.0TM (Psychology Software Tools

Inc., Pittsburgh, PA, USA) is given in Supplementary Table 3.

The results of SS and LK from these experiments were compared to those of 12 healthy con- trol participants (6 females; mean age: 32 years, age range: 21-53) with an academic background

(students, holders of a BSc, MSc or PhD). Ten of them (83%) showed a right-ear advantage (mean

LI: +17, LI range: -18 to +45) in the non-forced condition of the 'Bergen Dichotic Listening Test'

(Hugdahl 1995), and all of them displayed strong right-handedness (mean LQ: +84, LQ range: +40 to +100) in the questionnaire of Salmaso and Longoni (Salmaso & Longoni 1985). For the statisti- cal analysis of impairments (patient vs. controls), single dissociations between hemifields (left vs. right, within-patient), and double dissociations between patients, we applied inferential single case methods and statistically-based criteria for classical and strong dissociations (Crawford & Garth- waite 2005; Crawford et al. 2010). For example, the criteria for a classical double dissociation would be fulfilled if 1) SS showed - relative to controls - a significant deficit on task X (e.g. left- sided colour naming) but not on task Y (e.g. right-sided colour naming), 2) LK displayed the oppo- site pattern, and 3) the difference between task X and task Y significantly differed from correspond- ing values of the control group in both SS and LK. Moreover, both patients and all controls gave written informed consent prior to participation.

Results

Four of the 15 tachistoscopic experiments revealed double dissociations (Table 1). When comparing colours or letters to a central reference stimulus and when discriminating positions, SS displayed significant deficits only for right-sided stimuli while LK showed the opposite pattern. The findings from each of these three experiments fulfilled the criteria for a classical double dissociation. When

Page 7 of 29 identifying the global shape of Navon letters, SS again showed significant impairment for right- sided stimuli only, whereas LK had problems with both left- and right-sided global shapes. Howev- er, his performance for left-sided stimuli was significantly worse than that for right-sided ones, ful- filling the criteria for a strong dissociation. As schematically illustrated in Fig. 2, the overall pattern of double dissociations is consistent: Both SS and LK exclusively or predominantly showed im- pairment for stimuli presented to their contralesional visual hemifield.

[Insert Table 1 here]

[Insert Fig. 2 here]

In addition, single dissociations were found in SS's performance during the colour naming, size comparison, orientation comparison, object comparison, and form completion tasks. She again exclusively or predominantly displayed deficits for stimuli presented to her contralesional right vis- ual hemifield (Table 2). Conversely, LK had problems when localising visual stimuli presented to his contralesional left hemifield, in contrast to flawless performance for right-sided stimuli. Neither

SS nor LK showed pathological dissociations between hemifields when they had to name objects, read letters, identify the local shape of Navon letters, evaluate the lexical status of letter strings, read words, or recognise faces (Supplementary Table 4).

[Insert Table 2 here]

Discussion

We have introduced two patients with medial parietal brain damage in opposite hemispheres who display partial impairment when processing visual stimuli presented to their contralesional hemi- field. In the absence of task-relevant homonymous visual field defects, their difficulties are proba- bly best categorised as an incomplete variant of visual hemiagnosia, a disorder previously reported

Page 8 of 29 in rare case studies. For example, Mazzucchi and colleagues (Mazzucchi et al. 1985) described a patient with right temporo-occipital brain damage who demonstrated pronounced difficulties recog- nising visual stimuli such as colours, letters, words, objects, or faces presented to his left hemifield.

Complicating both the experimental investigation and the interpretation as an agnostic disorder to some degree, Goldmann perimetry revealed left upper homonymous quadrantanopia extending over the horizontal meridian in both of his eyes. A purer case of visual hemiagnosia was later introduced by Charnallet and co-workers (Charnallet et al. 1988). Their patient had left occipital brain damage

- apparently sparing the primary as confirmed by full visual fields in Goldmann pe- rimetry - and showed severe difficulties identifying a similarly wide variety of visual stimuli pre- sented to his right hemifield.

Compared to these cases, the visual hemiagnosia of SS and LK appears more subtle, as evi- dent by unimpaired performances during the letter reading or object naming experiment, for exam- ple. The four tasks that revealed double dissociations between our two patients - i.e. colour compar- ison, letter comparison, position discrimination, and identification of global Navon letter shapes - have one particular feature in common: They require participants to integrate more than one visual element, either through comparison or formation of a global gestalt. This also applies for the size comparison, orientation comparison, object comparison, and form completion tasks, which pro- voked single dissociations between hemifields in SS.

An alternative interpretation of the main findings in SS and LK would be that both may suf- fer from a deficit in attending to contralesional stimuli, rather than in visually integrating them.

Support for this notion comes from studies linking the precuneus to spatial attention in healthy par- ticipants (Mahayana et al. 2014) or to spatial hemineglect in brain-damaged patients (Molenberghs et al. 2012). We cannot completely rule out an influence of spatial inattention on our findings. But such influence seems marginal at most, as neither SS nor LK showed any sign of a contralesional hemineglect in standard cancellation or line bisection tasks (cf. Supplementary Table 2).

Page 9 of 29 Given the evidence pointing towards a role of the precuneus (Cavanna & Trimble 2006) and the retrosplenial cortex (Vann et al. 2009) in mnestic processing, it may surprise that SS and LK showed minor memory deficits only. However, both patients underwent routine neuropsychological examination more than four years after the onset of brain damage, a time period long enough to allow at least partial recovery. Impaired learning of new routes and spatial disorientation due to unilateral retrosplenial damage, for example, is known to resolve within a few months in the majori- ty of reported cases (Maguire 2001; Epstein 2008). All the more impressive is the observation that the visual integration impairment SS and LK show for contralesionally presented stimuli seems very persistent.

Since both SS and LK had a fully disconnected callosal splenium in addition to unilateral damage of the medial parietal lobe, it is important to discuss whether their impairment should be interpreted topologically - thereby assigning bottom-up visual integration function to the medial parietal lobe - or hodologically in terms of a disconnection syndrome. We have strong arguments in favour of the topological approach. First, crossed behavioural findings in the two patients suggest an association with their crossed lesion portion - i.e. the medial parietal damage - rather than with the uniform lesion of the unpaired splenium. Second, complete callosotomy in patients with left- hemisphere language dominance typically leads to anomia in speech and writing for visual stimuli presented to the left hemifield, whereas correctly matching or outlining the very same stimuli is still possible (Sperry et al. 1969). SS and LK who both show strong evidence for left-hemisphere lan- guage dominance differ from this pattern in at least two regards: SS displayed difficulties mainly with right-sided stimuli, and the visual impairment of both patients included basic perceptual matching.

Further support for a topological interpretation comes from functional neuroimaging find- ings indicating that the most prominent part of the medial parietal lobe, the precuneus, may contrib- ute to bottom-up visual integration. Himmelbach and colleagues (Himmelbach et al. 2009) - apply- ing an event-related design - found bilateral activity differences in the precuneus and the primary

Page 10 of 29 intermediate sulcus of a patient with subtotal simultanagnosia due to posterior cortical atrophy, when they contrasted successful with failed identification of global Navon letter shapes. Huberle and Karnath (Huberle & Karnath 2012) compared intact with disturbed perception of global square or circle shapes in healthy participants and showed bilateral activity differences in the temporo- parietal junction, the precuneus (predominantly left-sided) and the anterior cingulate cortex. Tan- skanen and co-workers (Tanskanen et al. 2008) recorded cortical responses of healthy participants to contour versus no-contour visual stimuli and found the most prominent differences in the bilat- eral precuneus and occipito-parietal sulcus. Moreover, the precuneus seems involved in complex forms of visual integration such as the perception of biological motion in point-light animations

(Ptito et al. 2003; Saygin & Sereno 2008).

Together with our results, these findings suggest that the bilateral precuneus contributes not only to internally-directed cognition such as visual imagery, retrieval from episodic memory or self- processing (Cavanna & Trimble 2006) but also to externally-directed, bottom-up integration of vis- ual elements. The double dissociations observed in SS and LK further suggest a retinotopical organ- isation insofar as the left precuneus might be involved in the integration of visual elements in the right hemifield and vice versa. Given its rich connectivity with numerous cortical and subcortical areas (Cavanna & Trimble 2006) and its contribution to integrating functions such as awareness or consciousness (Cavanna 2007), the bilateral precuneus might also be involved in bridging external- ly-directed and internally-directed cognition, together with other multimodal association areas such as the lateral (Mesulam 1998; Dixon et al. 2014).

With regard to limitations of the present study, an intrinsic constraint of any case report is first mentioned, i.e. the debatable generalisability of the findings. Furthermore, the arguments men- tioned above favouring a topological over a hodological interpretation refer to a comparison of the medial parietal lesion with the splenial disconnection. It is, however, possible that within the medial parietal lobe, not only cortical damage but also damage to the proximal white matter might contrib- ute to the unusual visual impairment observed in SS and LK. Finally, visual integration is likely

Page 11 of 29 based upon a network of brain regions. Here we propose that the bilateral precuneus may be an im- portant node of this network. Others have highlighted, for example, the bilateral temporo-parietal junction (Fink et al. 1997; Huberle & Karnath 2012; Rennig et al. 2013) or the bilateral primary intermediate sulcus located between the angular and the (Himmelbach et al.

2009).

Acknowledgments

We thank SS, LK, and all control participants for their time and effort as well as Ursula Gautschy for acquiring the visual fields.

References

Buckner RL. 2013. The brain's default network: origins and implications for the study of psychosis.

Dialogues Clin Neurosci. 15(3):351-358.

Cavanna AE, Trimble MR. 2006. The precuneus: a review of its functional anatomy and behaviour-

al correlates. Brain. 129(3):564-583.

Cavanna AE. 2007. The precuneus and consciousness. CNS Spectr. 12(7):545-552.

Charnallet A, Carbonnel S, Pellat J. 1988. Right visual hemiagnosia: a single case report. Cortex.

24(2):347-355.

Crawford JR, Garthwaite PH. 2005. Testing for suspected impairments and dissociations in single-

case studies in neuropsychology: evaluation of alternatives using Monte Carlo simulations

and revised tests for dissociations. Neuropsychology. 19:318-331.

Crawford JR, Garthwaite PH, Porter S. 2010. Point and interval estimates of effect sizes for the

case-controls design in neuropsychology: Rationale, methods, implementations, and pro-

posed reporting standards. Cogn Neuropsychology. 27:245-260.

Dixon ML, Fox KC, Christoff K. 2014. A framework for understanding the relationship between

externally and internally directed cognition. Neuropsychologia. [Epub ahead of print].

Page 12 of 29 Epstein RA. 2008. Parahippocampal and retrosplenial contributions to human spatial navigation.

Trends Cogn Sci. 12:388-396.

Fink GR, Halligan PW, Marshall JC, Frith CD, Frackowiak RS, Dolan RJ. 1997. Neural mecha-

nisms involved in the processing of global and local aspects of hierarchically organized vis-

ual stimuli. Brain. 120:1779-1791.

Himmelbach M, Erb M, Klockgether T, Moskau S, Karnath HO. 2009. fMRI of global visual per-

ception in simultanagnosia. Neuropsychologia. 47(4):1173-1177.

Huberle E, Karnath HO. 2012. The role of temporo-parietal junction (TPJ) in global Gestalt percep-

tion. Brain Struct Funct. 217(3):735-746.

Hugdahl K. 1995. Dichotic listening: probing functional integrity. In: Davidson RJ,

Hugdahl K, editors. Brain asymmetry. Cambridge (MA): MIT Press. p. 123-156.

Hugdahl K. 2011. Fifty years of dichotic listening research - still going and going and…. Brain

Cogn. 76(2):211-213.

Ino T, Doi T, Hirose S, Kimura T, Ito J, Fukuyama H. 2007. Directional disorientation following

left retrosplenial haemorrhage: a case report with fMRI studies. Cortex. 43(2):248-254.

Knecht S, Dräger B, Deppe M, Bobe L, Lohmann H, Flöel A, Ringelstein EB, Henningsen H. 2000.

Handedness and hemispheric language dominance in healthy . Brain. 12:2512-2518.

Leech R, Braga R, Sharp DJ. 2012. Echoes of the brain within the posterior cingulate cortex. J Neu-

rosci. 32(1):215-222.

Leech R, Sharp DJ. 2014. The role of the posterior cingulate cortex in cognition and disease. Brain.

137(1):12-32.

Maguire EA. 2001. The retrosplenial contribution to human navigation: a review of lesion and neu-

roimaging findings. Scand J Psychol. 42:225-238.

Mahayana IT, Tcheang L, Chen CY, Juan CH, Muggleton NG. 2014. The precuneus and visuospa-

tial attention in near and far space: a transcranial magnetic stimulation study. Brain Stimul.

[Epub ahead of print].

Page 13 of 29 Mazzucchi A, Posteraro L, Nuzzi G, Parma M. 1985. Unilateral visual agnosia. Cortex. 21(2):309-

316.

McDonald CR, Crosson B, Valenstein E, Bowers D. 2001. Verbal encoding deficits in a patient

with a left retrosplenial lesion. Neurocase. 7:407-417.

Mesulam MM. 1998. From sensation to cognition. Brain. 121(6):1013-1052.

Molenberghs P, Sale MV, Mattingley JB. 2012. Is there a critical lesion site for unilateral spatial

neglect? A meta-analysis using activation likelihood estimation. Front Hum Neurosci. 6:78.

Ptito M, Faubert J, Gjedde A, Kupers R. 2003. Separate neural pathways for contour and biological-

motion cues in motion-defined animal shapes. Neuroimage. 19(2):246-252.

Rennig J, Bilalić M, Huberle E, Karnath HO, Himmelbach M. 2013. The temporo-parietal junction

contributes to global gestalt perception-evidence from studies in chess experts. Front Hum

Neurosci. 7:513.

Richer F, Martinez M, Robert M, Bouvier G, Saint-Hilaire JM. 1993. Stimulation of human soma-

tosensory cortex: tactile and body displacement in medial regions. Exp Brain

Res. 93(1):173-176.

Rorden C, Brett M. 2000. Stereotaxic display of brain lesions. Behav Neurol. 12:191-200.

Salmaso D, Longoni AM. 1985. Problems in the assessment of hand preference. Cortex. 21(4):533-

549.

Saygin AP, Sereno MI. 2008. Retinotopy and attention in human occipital, temporal, parietal, and

frontal cortex. Cereb Cortex. 18(9):2158-2168.

Sperry RW, Gazzaniga MS, Bogen JE. 1969. Interhemispheric relationships: the neocortical com-

missures; syndromes of hemisphere disconnection. Handb Clin Neurol. 4:273-290.

Takahashi N, Kawamura M, Shiota J, Kasahata N, Hirayama K. 1997. Pure topographic disorienta-

tion due to right retrosplenial lesion. Neurology. 49(2):464-469.

Tanskanen T, Saarinen J, Parkkonen L, Hari R. 2008. From local to global: Cortical dynamics of

contour integration. J Vis. 8(7):15,1-12.

Page 14 of 29 Valenstein E, Bowers D, Verfaellie M, Heilman KM, Day A, Watson RT. 1987. Retrosplenial am-

nesia. Brain. 110(6):1631-1646.

Vann SD, Aggleton JP, Maguire EA. 2009. What does the retrosplenial cortex do? Nat Rev Neuro-

sci. 10(11):792-802.

Page 15 of 29 Tables

Table 1. Double dissociations found in the tachistoscopic experiments with lateralised stimulus presentation.

Experiment Patient Side Controls (N=12) Singlims_ESa RSDT_ESa Dissocs_ESa

b b Left Right Mean (SD) t (df=11) pone-tailed zCC t (df=11) pone-tailed zDCC Dissociation

Colour comparison SS 19 19.17 (0.83) -0.197 0.424 -0.205

15 19.67 (0.49) -9.157 <0.001*** -9.531 5.698 <0.001*** 6.363 Classical

LK 16 19.17 (0.83) -3.669 0.002** -3.819

20 19.67 (0.49) 0.647 0.265 0.673 2.799 0.009** -3.065 Classical

Letter comparison SS 19 19.50 (0.52) -0.924 0.188 -0.962

16 19.33 (0.78) -4.102 0.001** -4.269 2.387 0.018* 2.655 Classical

LK 15 19.50 (0.52) -8.314 <0.001*** -8.654

20 19.33 (0.78) 0.825 0.213 0.859 6.544 <0.001*** -7.636 Classical

Navon letter reading SS 10 9.75 (0.62) 0.387 0.353 0.403

(global letter) 6 9.75 (0.45) -8.006 <0.001*** -8.333 5.061 <0.001*** 5.541 Classical

LK 5 9.75 (0.62) -7.361 <0.001*** -7.661

8 9.75 (0.45) -3.736 0.002** -3.889 2.210 0.025* -2.393 Strong

Position discrimination SS 19 17.50 (1.93) 1.550 0.075 1.613

13 18.67 (1.67) -3.262 0.004** -3.395 4.460 <0.001*** 5.199 Classical

LK 10 17.50 (1.93) -7.748 <0.001*** -8.065

19 18.67 (1.67) 0.190 0.426 0.198 6.978 <0.001*** -8.577 Classical

Page 16 of 29 aSingle case methods developed by Crawford and colleagues (Crawford et al. 2010). Singlims_ES compares the test score of a patient with those of a control group, RSDT_ES compares the difference between two test scores (here left vs. right) of a patient against corresponding differences of a control group, and Dissocs_ES tests whether the pattern of a patient in two tasks meets the criteria for a classical or strong dissociation bPoint estimates of the effect size measured in standard deviation units

* Significant at α=0.05

** Significant at α=0.01

*** Significant at α=0.001

Page 17 of 29 Table 2. Single dissociations found in the tachistoscopic experiments with lateralised stimulus presentation.

Experiment Patient Side Controls (N=12) Singlims_ESa RSDT_ESa Dissocs_ESa

b b Left Right Mean (SD) t (df=11) pone-tailed zCC t (df=11) pone-tailed zDCC Dissociation

Colour naming SS 8 8.75 (0.97) -0.743 0.237 -0.773

6 9.17 (0.72) -4.230 <0.001*** -4.403 3.478 0.003** 4.008 Classical

LK 7 8.75 (0.97) -1.733 0.056 -1.804

8 9.17 (0.72) -1.561 0.073 -1.625 0.176 0.432 -0.198 None

Size comparison SS 16 18.42 (1.31) -1.775 0.052 -1.847

13 18.75 (0.87) -6.350 <0.001*** -6.609 2.733 0.010** 2.953 Classical

LK 19 18.42 (1.31) 0.425 0.340 0.443

19 18.75 (0.87 0.276 0.394 0.287 0.089 0.465 0.096 None

Orientation comparison SS 19 18.50 (1.00) 0.480 0.320 0.500

12 18.83 (0.72) -9.114 <0.001*** -9.486 6.555 <0.001*** 7.557 Classical

LK 19 18.50 (1.00) 0.480 0.320 0.500

18 18.83 (0.72) -1.108 0.146 -1.153 1.137 0.140 1.251 None

Object comparison SS 18 19.75 (0.45) -3.736 0.002** -3.889

16 19.92 (0.29) -12.987 <0.001*** -13.517 7.854 <0.001*** 9.847 Strong

LK 19 19.75 (0.45) -1.601 0.069 -1.667

20 19.92 (0.29) 0.265 0.398 0.276 1.764 0.053 -1.987 None

Form completion SS 9 9.08 (0.90) -0.085 0.467 -0.089

6 8.67 (1.15) -2.231 0.024* -2.322 1.922 0.040* 2.161 Classical

Page 18 of 29 LK 6 9.08 (0.90) -3.288 0.004** -3.422

7 8.67 (1.15) -1.395 0.095 -1.452 1.699 0.059 -1.906 None

Localising SS 10 9.75 (0.45) 0.534 0.302 0.556

10 10.00 (0.00) no datac

LK 6 9.75 (0.45) -8.006 <0.001*** -8.333

10 10.00 (0.00) no datac

aSingle case methods developed by Crawford and colleagues (Crawford et al. 2010) bPoint estimates of the effect size measured in standard deviation units cAll controls flawlessly localised right-sided stimuli. The resulting lack of statistical spread prevented the use of Singlims_ES for this variable, and that of RSDT_ES and Dis- socs_ES for this experiment

* Significant at α=0.05

** Significant at α=0.01

*** Significant at α=0.001

Page 19 of 29 Figures

Fig. 1. Brain lesions of SS (left) and LK (right) as depicted with standard clinical MRI. A) T1-weighted sagittal slices through the lesioned hemi- sphere, covering the entire lesion extent from medial to lateral. B) T1-weighted coronal slices, covering the entire lesion extent from rostral to cau- dal. C) T2- (SS) or T1-weighted (LK) axial slices, covering the entire lesion extent from dorsal to ventral. Lesions appear as cerebrospinal fluid- isointense areas. Multislices were created with MRIcron (Rorden & Brett 2000) (http://www.mccauslandcenter.sc.edu/mricro/mricron/). Single slic- es are oriented according to the neurological convention, i.e. the left side of the image corresponds to the left side of the brain.

Page 20 of 29

Fig. 2. Schematic illustration of brain lesions and double dissociations found in tachistoscopic experiments with lateralised stimulus presentation.

Medial parietal lesions are highlighted in red (SS) or blue (LK), the lesion of the callosal splenium in dark grey. Both patients showed classical or strong dissociations between hemifields in the colour comparison, letter comparison, Navon global letter reading, and position discrimination tasks.

Stimulus examples are shown in the hemifield where the worse performance was observed. Source note: The sagittal brain view was downloaded from http://www.healthline.com/human-body-maps/brain.

Page 21 of 29

Supplementary Fig. 1. Automated static perimetry. Greyscale plots based on the programme 'G2 standard' of the Octopus 900™ (Haag-Streit AG,

Köniz, Switzerland), displaying the central ±30° (inner circle) and peripheral ±60° (outer circle) of the visual field for both eyes and both patients.

In LK's plots, the dimensions of the notebook screen (31 x 24°) used for the tachistoscopic experiments are shown as a superimposed blue rectangle.

Page 22 of 29 Supplementary Tables

Supplementary Table 1. Results of standardised, norm-referenced neuropsychological tests.

SS LK Domain Subdomain Test Variable Raw score T valuea Raw score T valuea Attention Processing speed ZVT Mean duration (s) 79 42 90 44 Tonic alertness TAP; Alertness Median (ms) 213 49 273 42 SD (ms) 16 >60 51 44 Phasic alertness TAP; Alertness Median (ms) 213 48 259 42 SD (ms) 22 56 35 51 Selective attention TAP; Go/NoGo Median (ms) 424 >60 543 48 Omissions 0 >34 0 >34 False positives 0 >48 2 40 Memory Verbal span WMS-R; Digits Correct forwardb 7 45 6 41 Correct backwardb 7 49 6 45 Visuo-spatial span WMS-R; Blocks Correct forwardb 8 43 7 39 Correct backwardb 10 59 7 43 Verbal episodic memory HVLT-R Sum of trials 1-3 32 58 32 58 Delayed recall 8 35 12 59 Percent retained 73 29 100 55 Discrimination 10 44 12 57 Visual episodic memory BVMT-R Sum of trials 1-3 32 58 30 59 Delayed recall 12 59 12 >60 Percent retained 100 >40 109 >60 Discrimination 6 >40 6 >40 Visuoconstruction RCFT; Copy Points 35 >40 35 >40 Duration (s) 161 >40 173 >40 Executive functions Verbal fluency RWT; S-words Correct words 28 52 16 41 Design fluency 5PT Correct designs 45 60 30 43 Verbal response inhibition D-KEFS; CWIT, cond. 3 Duration (s) 44 55 49 54 Errors 3 43 0 56 Nonverbal response inhibition TAP; Incompatibility Median (ms) 513 37 515 46 Errors 4 46 10 37 Verbal task switching D-KEFS; CWIT, cond. 4 Duration (s) 65 43 58 53 Errors 0 56 2 50 aCorrected for age and - if available - for education. Based on international consensus (Heaton et al. 2004), T values between 35 and 39 represent mild, those between 25 and 29 moderate, and those below 20 severe deficits. bThe number of correct trials is analysed here, not the absolute span Note: Values representing a deficit are highlighted in bold and by a light grey background

Page 23 of 29 Abbreviations: ZVT = Zahlen-Verbindungs-Test (Oswald & Roth 1987), a processing speed test similar to the Trail Making Test A; TAP = Tests of Attentional Performance (Zimmermann & Fimm 2007), a battery of computer-based tests for attentional performance; WMS-R = Wechsler Memory Scales-Revised (Härting et al. 2000); HVLT = Hop- kins Verbal Learning Test - Revised (Brandt & Benedict 2001); BVMT-R = Brief Visuospatial Memory Test - Revised (Benedict 1997); RCFT = Rey Complex Figure Test (Meyers & Meyers 1995); RWT = Regensburger Wortflüssigkeits-Test (Aschenbrenner et al. 2000), a phonemic word fluency test; 5PT = Five-Point Test (Haid et al. 2002), a design fluency test; D-KEFS, CWIT = Delis–Kaplan Executive Function System, Color-Word Interference Test (Delis et al. 2001).

Supplementary References - Aschenbrenner S, Tucha O, Lange KW. 2000. Regensburger Wortflüssigkeits-Test (RWT). Göttingen: Hogrefe. - Benedict RHB. 1997. Brief Visuospatial Memory Test - Revised. Lutz (FL): Psychological Assessment Resources. - Brandt J, Benedict RHB. 2001. Hopkins Verbal Learning Test - Revised. Lutz (FL): Psychological Assessment Resources. - Delis DC, Kaplan E, Kramer JH. 2001. Delis Kaplan Executive Function System (D-KEFS). San Antonio (TX): The Psychological Corporation. - Härting C, Markowitsch HJ, Neufeld H, Calabrese P, Deisinger K, Kessler J (editors). 2000. Wechsler Gedächtnistests - revidierte Fassung. Bern: Hans Huber. - Haid T, Martl C, Schubert F, Wenzl M, Kofler M, Saltuari L. 2002. Der HAMASCH 5 Punkt Test. Erste Normierungsergebnisse. Z Neuropsychol. 13(3): 233. - Heaton RK, Miller SW, Taylor MJ, Grant I. 2004. Revised comprehensive norms for an expanded Halstead-Reitan Battery: Demographically adjusted norms for African American and Caucasian adults. Lutz (FL): Psychological Assessment Resources. - Meyers JE, Meyers KR. 1995. Rey Complex Figure Test and Recognition Trial. Lutz (FL): Psychological Assessment Resources. - Oswald WO, Roth E. 1987. Der Zahlen-Verbindungs-Test (ZVT). 2., überarbeitete und erweiterte Auflage. Göttingen: Hogrefe. - Zimmermann P, Fimm B. 2007. Testbatterie zur Aufmerksamkeitsprüfung (TAP). Version 2.1. Herzogenrath: Psytest.

Page 24 of 29 Supplementary Table 2. Results of neuropsychological screening tests.

SS LK Domain Subdomain Test Variable Raw score Statistic Raw score Statistic Visual perception Figure-ground perception VOSP; Shape detection Correct/maximum 20/20 >5%-cut-off 20/20 >5%-cut-off Form completion VOSP; Incomplete letters Correct/maximum 20/20 >5%-cut-off 19/20 >5%-cut-off Object perception VOSP; Silhouettes Correct/maximum 28/30 >5%-cut-off 21/30 >5%-cut-off Localisation VOSP; Number Location Correct/maximum 10/10 >5%-cut-off 9/10 >5%-cut-off Spatial perception VOSP; Cube analysis Correct/maximum 10/10 >5%-cut-off 10/10 >5%-cut-off Spatial attention Visual exploration Bells Test Left omissions 0 normala 0 normala Right omissions 0 normala 2 normala L/R difference 0 normalb 2 normalb Line Bisectionc Left omissions 0 0 Right omissions 0 0 Mean deviation (%) 1.22 (right) normald -1.72 (left) normald Eye-hand coordination Object grasping In-house screeninge Left correct/maximum 6/6 6/6 Right correct/maximum 6/6 6/6 Language Oral repetition Aphasia Screeningf Correct/maximum 10/10 10/10 Reading aloud Correct/maximum 6/6 6/6 Colour naming Correct/maximum 8/8 8/8 Object naming Correct/maximum 10/10 10/10 Writing upon dictation Correct/maximum 6/6 6/6 Oral comprehension Correct/maximum 8/8 8/8 Reading comprehension Correct/maximum 8/8 8/8 Calculation In-house screening Correct/Max. 12/12 12/12 Praxis Ideomotor, upper limb In-house screening Left correct/maximum 9/9 9/9 Right correct/maximum 9/9 9/9 aMore than three omissions on one or the other side indicate spatial inattention (Gauthier et al. 1989) bAn absolute difference between left and right omissions of five or more is considered pathologic (Rousseaux et al. 2002) cWith reference to Schenkenberg et al. (Schenkenberg et al. 1980), 18 lines of different lengths (i.e. 10, 12, 14, 16, 18, or 20 cm) were presented at three different positions (i.e. left, central, right) dMore than 14% displacement of the bisection mark in relation to the entire length is considered abnormal (Ferber & Karnath 2001) ePatients reached for peripherally presented targets under central fixation with their ipsilateral hand (e.g., left hand reached for left-sided targets) fWeniger 2006

Abbreviations: VOSP = The Visual Object and Space Perception Battery (Warrington & James 1991).

Supplementary References - Ferber S, Karnath HO. 2001. How to assess spatial neglect - line bisection or cancellation tasks? J Clin Exp Neuropsychol. 23(5): 599-607. - Gauthier L, Dehaut F, Joanette Y. 1989. The bells test: a quantitative and qualitative test for visual neglect. Int J Clin Neuropsychol. 11: 49-54.

Page 25 of 29 - Rousseaux M, Beis JM, Pradat-Diehl P, Martin Y, Bartolomeo P, Bernati T, Chokron S, Leclercq M, Louis-Dreyfus A, Marchal F et al. 2001. Présentation d’une batterie de dépistage de la négligence spatiale. Rev Neurol (Paris). 157:1385-1400. - Schenkenberg T, Bradford DC, Ajax ET. 1980. Line bisection and unilateral visual neglect in patients with neurologic impairment. Neurology. 30: 509-17. - Warrington EK, James M. 1991. The Visual Object and Space Perception Battery. Bury St Edmunds, England: Thames Valley Test Company. - Weniger D. 2006. Aphasie. In: Karnath H-O, Hartje W, Ziegler W, editors. Kognitive Neurologie. Stuttgart: Thieme, p. 48-64.

Page 26 of 29 Supplementary Table 3. Description of all tachistoscopic experiments.

No. Experiment name Task instruction Number of Presentation Mean stimulus Stimulus description test trials duration (ms) size (°) 1 Colour naming Name the colour of the circle 20 50 3.1 x 3.1 Filled coloured circles 2 Colour comparison Compare the colour of the lateral circle to 40 50 lateral: 3.1 x 3.1 Filled coloured circles that are identical (20 trials) or differ- that of the central one replacing the central central: 1.5 x 1.5 ent from a simultaneously shown smaller central circle fixation cross (CFC). Same or different? 3 Localising Decide which letter of this grid is in exact- 20 100 1.1 x 1.1 Single black dots. During the response window, a grid with ly the same position as the black dot 28 unique upper-case letters evenly distributed over the shown before entire screen is shown 4 Orientation comparison Decide whether the two lines are parallel 40 100 1.4 x 5.2 Two roughly vertical lines that are parallel (20 trials) or or not form an angle of 3-13° 5 Size comparison Decide which of the two circles is bigger. 40 100 2.3 x 7 Two filled black circles, identical (20 trials) or different Upper or lower? (0.4-0.9° diameter difference) in size 6 Letter reading Read the letter 20 50 1.2 x 1 Upper-case letters (Courier new), with letter confusabilitya balanced across sides 7 Letter comparison Compare the big lateral letter to the small 40 100 lateral: 1.8 x 1.5 Upper-case letters (Courier new) that are identical (20 trials) central one replacing the CFC. Same or central: 1 x 0.9 or different from a simultaneously shown smaller central different? letter 8 Object naming Name the object 20 100 2.8 x 2.8 Greyscale images of everyday objectsb 9 Object comparison Compare the big lateral object to the small 40 100 lateral: 2.6 x 2.6 Greyscale images of everyday objectsb that are identical (20 central one replacing the CFC. Same or central: 1.3 x 1.3 trials) or different from a simultaneously shown smaller different? central object 10 Form completion Name the number 20 100 2.2 x 3.4 Fragmented numbers 11 Position discrimination Decide which of the two squares contains 40 200 3.7 x 8.9 Two vertically arranged squares, one of them containing a a black dot in a central position. Upper or central black dot, the other the same dot in a non-central lower? (0.1-0.9° deviation) position 12 Navon letter reading Identify both the global as well as the 20 100 5.9 x 8.8 Hierarchical letters (Courier new), with many identical local letter information small letters forming a non-matching large letter 13 Lexical decision Decide if the letter string is a word or a 40 150 5.8 x 1.3 Five-letter strings (Courier new); nouns (20 trials; word nonword frequency balanced across sidesc) or nonwords formed by replacing internal letters of nouns 14 Word reading Read the word 20 150 5.8 x 1.3 Five-letter nouns (Courier new), with word frequency bal- anced across sidesc 15 Face recognition Identify the face 20 150 3.7 x 8.3 Portrait images of well-known people such as politicians, musicians or actors/actresses

aBased on the confusability indices presented by Townsend (Townsend 1971) bFrom Rossion & Pourtois (Rossion & Pourtois 2004) cBased on Hager and Hasselhorn (Hager and Hasselhorn 1994)

Page 27 of 29 Supplementary References - Hager W, Hasselhorn M. 1994. Handbuch deutschsprachiger Wortnormen. Göttingen: Hogrefe Verlag. - Rossion B, Pourtois G. 2004. Revisiting Snodgrass and Vanderwart's object pictorial set: the role of surface detail in basic-level object recognition. Perception. 33(2):217- 236. - Townsend JT. 1971. Theoretical analysis of an alphabetic confusion matrix. Percept Psychophys. 9(1): 40-50.

Page 28 of 29 Supplementary Table 4. Further results from the tachistoscopic experiments with lateralised stimulus presentation.

Experiment Patient Side Controls (N=12) Singlims_ESa RSDT_ESa Dissocs_ESa b b Left Right Mean (SD) t (df=11) pone-tailed zCC t (df=11) pone-tailed zDCC Dissociation Object naming SS 7 7.17 (1.03) -0.159 0.438 -0.165 8 9.08 (1.00) -1.038 0.161 -1.080 0.547 0.298 0.593 none LK 7 7.17 (1.03) -0.159 0.438 -0.165 10 9.08 (1.00) 0.884 0.198 0.920 0.649 0.265 -0.703 none Letter reading SS 10 10.00 (0.00) no datac 10 9.83 (0.39) 0.419 0.342 0.436 LK 10 10.00 (0.00) no datac 10 9.83 (0.39) 0.419 0.342 0.436 Navon letter reading SS 6 8.92 (1.51) -1.858 0.045* -1.934 (local letter) 5 8.83 (1.90) -1.937 0.039* -2.016 0.074 0.471 0.082 none LK 7 8.92 (1.51) -1.222 0.124 -1.272 10 8.83 (1.90) 0.592 0.283 0.616 1.685 0.060 -1.895 none Lexical decision SS 18 16.08 (2.02) 0.913 0.190 0.950 16 17.75 (1.82) -0.924 0.188 -0.962 1.482 0.083 1.649 none LK 14 16.08 (2.02) -0.989 0.172 -1.030 17 17.75 (1.82) -0.396 0.350 -0.412 0.480 0.320 -0.533 none Word reading SS 2 6.00 (1.95) -1.971 0.037* -2.051 5 8.17 (1.95) -1.562 0.073 -1.626 0.435 0.336 -0.490 none LK 2 6.00 (1.95) -1.971 0.037* -2.051 7 8.17 (1.95) -0.576 0.288 -0.600 1.477 0.084 -1.669 none Face recognition SS 6 7.17 (1.40) -0.803 0.220 -0.836 3 7.08 (1.56) -2.513 0.014* -2.615 1.217 0.124 1.339 none LK 7 7.17 (1.40) -0.117 0.455 -0.121 5 7.08 (1.56) -1.281 0.113 -1.333 0.830 0.212 0.912 none aSingle case methods developed by Crawford and colleagues (Crawford et al. 2010). Singlims_ES compares the test score of a patient with those of a control group, RSDT_ES compares the difference between two test scores (here left vs. right) of a patient against corresponding differences of a control group, and Dissocs_ES tests whether the pattern of a patient in two tasks meets the criteria for a classical or strong dissociation bPoint estimates of the effect size measured in standard deviation units cAll controls flawlessly read left-sided letters. The resulting lack of statistical spread prevented the use of Singlims_ES for this variable, and that of RSDT_ES and Dissocs_ES for this experiment

* Significant at α=0.05

Page 29 of 29