Abstract

We examined a stroke patient (HWS) with a unilateral lesion of the right medial ventral visual stream, involving the right fusiform and parahippocampal gyri. In a number of object recognition tests with lateralized presentations of target stimuli, HWS showed significant symptoms of hemiagnosia with contralesional recognition deficits for everyday objects. We further explored the patient's capacities of visual expertise that were acquired before the current perceptual impairment became effective. We confronted him with objects he was an expert for already before stroke onset and compared this performance with the recognition of familiar everyday objects. HWS was able to identify significantly more of the specific (“expert”) than of the everyday objects on the affected contralesional side. This observation of better expert object recognition in visual hemiagnosia allows for several interpretations. The results may be caused by enhanced information processing for expert objects in the ventral system in the affected or the intact hemisphere. Expert knowledge could trigger top–down mechanisms supporting object recognition despite of impaired basic functions of object processing. More importantly, the current work demonstrates that top–down mechanisms of visual expertise influence object recognition at an early stage, probably before visual object information propagates to modules of higher object recognition. Because HWS showed a lesion to the fusiform gyrus and spared capacities of expert object recognition, the current study emphasizes possible contributions of areas outside the ventral stream to visual expertise.

INTRODUCTION

Human lesion and neuroimaging studies (Karnath, Rüter, Mandler, & Himmelbach, 2009; Grill-Spector et al., 1999; Malach et al., 1995; Goodale, Milner, Jakobson, & Carey, 1991) as well as electrophysiology in monkeys (Logothetis, Pauls, & Poggio, 1995) demonstrated a robust association of the inferior temporal cortex with visual object perception. Interestingly, these processes appear to have a spatiotopic organization. Functional neuroimaging in humans (Niemeier, Goltz, Kuchinad, Tweed, & Vilis, 2005; Malach, Levy, & Hasson, 2002; Levy, Hasson, Avidan, Hendler, & Malach, 2001) and single-cell recordings in monkeys (Hung, Kreiman, Poggio, & DiCarlo, 2005; DiCarlo & Maunsell, 2003) showed that early object sensitive areas along the ventral object perception stream, like the lateral occipital complex or inferior temporal areas/neurons, show a stronger preference for contralateral visual object stimuli than for ipsilateral presentations. Comparable spatial preferences have also been reported for higher object sensitive areas like the fusiform face area (FFA) or parahippocampal structures (Golomb & Kanwisher, 2012; Kravitz, Kriegeskorte, & Baker, 2010; Hemond, Kanwisher, & Op de Beeck, 2007). These results obtained from fMRI in healthy participants were confirmed by a recent study of stroke patients with ventral cortex lesions (Rennig, Karnath, Cornelsen, Wilhelm, & Himmelbach, 2017). Despite preserved primary visual functions, these patients were significantly impaired in recognizing objects presented to their contralesional hemifield. This deficit contrasted with normal recognition performance of the same stimuli if presented in their ipsilesional hemifield.

This impairment of visual hemiagnosia (e.g., Rennig et al., 2017; Mazzucchi, Posteraro, Nuzzi, & Parma, 1985) after unilateral ventral cortex lesions as well as the related neuropsychological deficit termed visual agnosia (e.g., Farah, 2004) are characterized through an inability to recognize visually presented objects despite of preserved primary visual function and conceptual object knowledge. A similar impairment is prosopagnosia—it describes the inability to recognize faces despite of spared primary visual functions and preserved capacities of object recognition (e.g., Damasio, Damasio, & Van Hoesen, 1982). Interestingly, it has been observed that such agnosic deficits can be overcome by visual expertise. For example, a study with a patient suffering from prosopagnosia after a unilateral lesion to the ventral cortex (Bukach, Bub, Gauthier, & Tarr, 2006) showed that specific subfunctions of face perception that were considered perceptual expert functions were still preserved despite of impaired abilities of face recognition. A behavioral training study (Behrmann, Marotta, Gauthier, Tarr, & McKeeff, 2005) demonstrated that a visual agnosia patient with a lesion to the posterior temporal cortex was able to increase recognition rate for artificial objects with behavioral training. This training effect was even transferred to untrained versions of artificial objects as well as everyday objects.

This has led to the assumption that visual expertise as a specific quality of object perception might have a characteristic representation in the brain different from “regular” object processing. In line with this notion, it has been demonstrated that experts show specific behavioral (Ericsson & Lehmann, 1996) and neuronal (Harel, Kravitz, & Baker, 2013) characteristics while observing objects after frequent exposure. In functional neuroimaging studies, participants with specific expert knowledge, like car and bird experts (Gauthier, Skudlarski, Gore, & Anderson, 2000), radiologists (Bilalić, Grottenthaler, Nägele, & Lindig, 2016), or chess players (Bilalić, Turella, Campitelli, Erb, & Grodd, 2012; Bilalić, Langner, Ulrich, & Grodd, 2011; Krawczyk, Boggan, McClelland, & Bartlett, 2011), showed characteristic brain activations compared with nonexperts. Commonly, the FFA (Kanwisher, McDermott, & Chun, 1997) represents the brain region that is associated with such expert object processing (Gauthier, Curran, Curby, & Collins, 2003). However, differences between experts and novices were also discovered outside the FFA in posterior temporoparietal brain areas (Rennig, Himmelbach, Huberle, & Karnath, 2015; Rennig, Bilalić, Huberle, Karnath, & Himmelbach, 2013; van der Linden, van Turennout, & Indefrey, 2010; Moore, Cohen, & Ranganath, 2006).

Considering the significant effects of visual expertise on cases with visual agnosia and prosopagnosia (Bukach et al., 2006; Behrmann et al., 2005), we asked if visual expertise acquired before the onset of a perceptual impairment can also compensate recognition deficits. Unlike in other studies where patients with perceptual deficits were trained after the impairment had occurred (Bukach et al., 2006; Behrmann et al., 2005), we here were able to test capacities of visual expertise that were acquired before the respective impairment became effective. With this approach, we examined if long-term visual expertise helps to preserve capacities of object recognition despite of cardinal deficits. To test our question, we examined a stroke patient with visual hemiagnosia suffering from a lesion to the medial ventral stream affecting the fusiform and parahippocampal gyri of the right hemisphere. We examined this patient with object recognition tests that contained “regular” everyday objects as well as “expert” objects from his field of expertise. Based on the previous findings (Bukach et al., 2006; Behrmann et al., 2005), we hypothesized that “expert” objects compared with “regular” ones will elicit less pronounced deficits in the patient's contralesional hemifield.

METHODS

Participants

Expert Participant HWS

HWS is a 56-year-old man who suffered from a right hemispheric stroke. A high-resolution MR scan showed involvement of the fusiform and parahippocampal gyri, sparing large parts of the lingual gyrus (Figure 1). Demographic and clinical data of HWS are shown in Table 1. HWS is a trained car mechanic who has worked in this job since age 17 years and has run his own business for more than two decades. Based on this professional experience from about 38 years, we classified him as an expert for car-/garage-related objects, like certain tools (wrench, screwdriver, car jack, etc.), car parts (e.g., steering wheel, tire), or motor vehicles (cars, motorcycles). In agreement with his ventral stream lesions, he displayed unimpaired visually guided reaching but impaired reaching to memorized targets (Cornelsen, Rennig, & Himmelbach, 2016).

Figure 1. 

Neuroimaging data from HWS and the control stroke patients. (A) A high-resolution T1 image demonstrates a right hemispheric lesion in the inferior temporal cortex of HWS. (B) Normalized lesion of HWS superimposed on the single-subject T1 MNI152 template. (C) Lesion overlap for the ventral control patients. (D) Lesion overlap for the nonventral control patients. In each patient (including HWS), the lesion boundary was delineated directly on the individual scan for every single transverse slice using MRIcron software (www.mccauslandcenter.sc.edu/mricro/mricron). Both the MRI/CT scan and the lesion shape were then mapped into stereotaxic space using the normalization algorithm provided by the Clinical Toolbox (Rorden et al., 2012) based on SPM8 (fil.ion.ucl.ac.uk/spm). The lesion maps of all control patients were superimposed on the single-subject T1 MNI152 template. The figure shows the vertical z coordinate for each slice of standardized MNI space. For each voxel, the number of patients with a lesion at that location is color coded.

Figure 1. 

Neuroimaging data from HWS and the control stroke patients. (A) A high-resolution T1 image demonstrates a right hemispheric lesion in the inferior temporal cortex of HWS. (B) Normalized lesion of HWS superimposed on the single-subject T1 MNI152 template. (C) Lesion overlap for the ventral control patients. (D) Lesion overlap for the nonventral control patients. In each patient (including HWS), the lesion boundary was delineated directly on the individual scan for every single transverse slice using MRIcron software (www.mccauslandcenter.sc.edu/mricro/mricron). Both the MRI/CT scan and the lesion shape were then mapped into stereotaxic space using the normalization algorithm provided by the Clinical Toolbox (Rorden et al., 2012) based on SPM8 (fil.ion.ucl.ac.uk/spm). The lesion maps of all control patients were superimposed on the single-subject T1 MNI152 template. The figure shows the vertical z coordinate for each slice of standardized MNI space. For each voxel, the number of patients with a lesion at that location is color coded.

Table 1. 

Demographic and Clinical Data of HWS, All Ventral and Nonventral Control Patients as well as Healthy Control Participants

HWSVentral Control PatientsNon-ventral Control PatientsHealthy Controls
Number – 19 20 
Sex (male/female) Male 5/4 7/12 4/15 
Mean age (years) (SD56 60.88 (18.03) 56.95 (11.71) 63.35 (7.64) 
Etiology Infarct 8 infarct, 1 hemorrhage 18 infarct, 1 hemorrhage – 
Hemisphere (left/right) Right 5/4 6/13 – 
VOSP Object Recognition    – 
 Incomplete Letters Normal 9/9 normal 19/19 normal – 
 Silhouettes Normal 8/9 normal 19/19 normal – 
 Shape Decision Normal 7/9 normal 18/19 normal – 
 Progressive Silhouettes Normal 8/9 normal 19/19 normal – 
Time since lesion (days poststroke) (SDAcute patients (n = 4) Acute patients (n = 6) – 
6.00 (7.01) 4.75 (2.06) – 
Chronic patients (n = 5) Chronic patients (n = 13) – 
999.20 (348.91) 1109.14 (410.07) – 
HWSVentral Control PatientsNon-ventral Control PatientsHealthy Controls
Number – 19 20 
Sex (male/female) Male 5/4 7/12 4/15 
Mean age (years) (SD56 60.88 (18.03) 56.95 (11.71) 63.35 (7.64) 
Etiology Infarct 8 infarct, 1 hemorrhage 18 infarct, 1 hemorrhage – 
Hemisphere (left/right) Right 5/4 6/13 – 
VOSP Object Recognition    – 
 Incomplete Letters Normal 9/9 normal 19/19 normal – 
 Silhouettes Normal 8/9 normal 19/19 normal – 
 Shape Decision Normal 7/9 normal 18/19 normal – 
 Progressive Silhouettes Normal 8/9 normal 19/19 normal – 
Time since lesion (days poststroke) (SDAcute patients (n = 4) Acute patients (n = 6) – 
6.00 (7.01) 4.75 (2.06) – 
Chronic patients (n = 5) Chronic patients (n = 13) – 
999.20 (348.91) 1109.14 (410.07) – 

Group mean values and standard deviations. In the VOSP Object Recognition subtests, an abnormal performance was diagnosed if participant's score fell below the age-corrected cutoff. The numbers demonstrate how many patients out of the respective group reached a score classified as “normal.”

Control Participants

Acute unilateral stroke patients consecutively admitted to the Centre of Neurology at the University of Tübingen as well as chronic unilateral stroke patients from an in-house database were recruited as control patients. HWS as well as all control patients and healthy control participants were also tested for previous study (Rennig et al., 2017). All patients were tested for visual field defects, spatial neglect, visual extinction, subclinical spatial attention deficits, visual/auditory object identification, or severe aphasia (see Neuropsychological Screening section). Only patients without such deficits were included in this study. Twenty-eight control stroke patients could be recruited (10 acute, 18 chronic). These participants were subdivided into two control groups. One subgroup consisted of nine patients who suffered from lesions to the ventral stream (“ventral” controls); the other subgroup of 19 patients had lesions outside the ventral stream (“nonventral” controls). Figure 1 gives an overview of brain lesions from the two control groups; Table 1 gives an overview of the demographic and clinical data. Another 11 patients were tested but were later excluded from further analyses because of the presence of hemianopia, attentional impairments, spatial neglect, or visual extinction (see Neuropsychological Screening section).

In addition, we tested 20 elderly healthy control participants. All participants had normal or corrected-to-normal vision and gave their informed consent to participate in the study, which was performed in accordance with the ethical standards of the Declaration of Helsinki as revised in 2013. We used Crawford's single case statistic (Crawford & Garthwaite, 2002) to test HWS's age against the three control groups and did not observe a significant result (healthy controls: t test: t(19) = 0.93, p = .18; nonventral control patients: t test: t(18) = 0.05, p = .48; ventral control patients: t test: t(8) = 0.16, p = .44).

Stimuli and Procedure

Neuropsychological Screening

The presence of visual field defects was tested with the neurological confrontation method. In addition, a custom in-house perimetry and attention screening test was conducted with the same computer running the unilateral object identification tests described below (see Unilateral Object Recognition Tests section). The stimuli were presented on a laptop (screen size: 37.5° × 30°) using Matlab (MathWorks, Natick, MA) and the Psychophysics Toolbox (Brainard, 1997); a distance of 57 cm between the participant and the screen was kept constant. Participants focused a central gray fixation cross. The examiner pressed a button to start the trial: As a cue for the upcoming stimulus presentation, the cross turned black. After a randomized interval between 1 and 2 sec, a single dot flashed for 20 msec. In total, 16 possible locations covered left and right visual field areas, corresponding to a visual field of ±16° in the horizontal and ±12° in the vertical dimension. Thirty-two test trials and four additional catch trials without target presentation after the cue were presented to each patient. Participants responded verbally as soon as they detected a dot; RTs were recorded via voice key with a microphone placed right in front of the participants' mouth not covering the computer screen. Hits/misses were recorded by the experimenter. Misses from this test were used for the diagnosis of visual field defects. We identified three patients with hemianopia that were excluded from the study. Three patients showed symptoms of quadrantanopia. These patients were not excluded, but we adapted stimulus position such that all stimuli were shifted into the patient's spared quadrant (for a detailed description, see below). RTs from this test were used to diagnose subclinical attentional deficits. We compared RTs from the left and right visual hemifield directly against each other based on the performance of the healthy control group. For every healthy participant, we calculated the difference between the performance in the left and the right (or right and left) hemifield for RT. The direction of the subtraction (L − R, R − L) was assigned randomly to the healthy participants, providing equal frequencies for both directions (10 participants L − R; 10 participants R − L). Then mean and standard deviation over these difference values were calculated and used as normative statistics (RT difference: mean = 4 msec, SD = 124 msec). For all patients, we subtracted RT values from the potentially affected visual field from those of the unaffected side. All control patients whose RT difference values were 2 SDs below the standard mean value of the healthy control group were considered as abnormal and excluded from the later analysis. Based on this criterion, two patients were excluded due to longer RTs for stimuli presented in the contralesional hemifield.

The Letter Cancellation Test (Weintraub & Mesulam, 1985), the Bells Test (Gauthier, Dehaut, & Joanette, 1989), a copying task (Johannsen & Karnath, 2004), and a line bisection task (Ferber & Karnath, 2001) were used to test for the presence of spatial neglect. For the two cancellation and the copying tasks, cutoff criteria were used as applied in previous studies (Karnath, Rennig, Johannsen, & Rorden, 2011; Rorden & Karnath, 2010); a deviation of more than 14% from the true midpoint was considered abnormal for the bisection task (Ferber & Karnath, 2001). Visual extinction was assessed with the neurological confrontation technique (Becker & Karnath, 2007). Ten bilateral and 10 unilateral left- or right-sided visual stimuli were presented in random order. Patients were classified as showing visual extinction when they reported at least 90% of the left or right stimuli on each side correctly but failed to indicate the contralesional stimulus during bilateral stimulation in 50% of the trials. Six participants with at least one positive neglect test or with symptoms of extinction were identified and excluded from this study.

To also exclude any symptoms of deficits of object naming, we conducted an auditory object identification task. Object sounds from an online library (Marcell, Borella, Greene, Kerr, & Rogers, 2000) were presented via headphones; participants simply had to name the object corresponding to the respective sound. Twenty trials were conducted. Healthy controls reached a mean naming accuracy of 98%, with a standard deviation of 4.2%. A threshold of the mean accuracy of the healthy control group minus twice the standard deviation was used as threshold for all patients to be included into the study. With this method, also patients with severe aphasic naming problems were identified and excluded. In this test, all patients included into the study were able to identify at least 90% of the presented objects, indicating that none of them suffered from auditory associative agnosia or severe aphasic naming problems.

Additional Computerized Perimetry

In patients with ventral lesions where we could expect visual field defects due to lesions to the primary visual cortex, an additional computerized suprathreshold perimetry with 191 test locations within the central 30° visual field (Octopus 101, Haag-Streit International, Köniz, Switzerland) was performed. HWS and all but two participants from the ventral patient control group underwent this examination. In some patients, kinetic perimetry using at least four isopters was used. Light stimuli close to age-dependent perceptual thresholds were presented with increasing brightness if they were not initially detected to reveal relative and absolute visual field defects. Background illuminance was 10 cd/m2, and maximum stimulus illuminance was 1000 cd/m2. This screening did not reveal further impairments of primary vision that were not already detected with our in-house perimetry test.

Visual Object Recognition under Unrestricted Viewing Conditions

Visual object recognition was examined with the VOSP test battery (Warrington & James, 1991). In this test, advanced functions of object recognition are tested in four subtests with alienated depictions of everyday objects. Pictures were presented without temporal restrictions, and the patients were allowed to visually explore the stimuli. Performance in the VOSP was not part of the inclusion/exclusion criteria.

Unilateral Object Recognition Tests

Performance in these tests also were not part of the inclusion/exclusion criteria. Test stimuli were presented on a laptop (screen size: 37.5° × 30°) using Matlab (MathWorks) and the Psychophysics Toolbox (Brainard, 1997). For all unilateral object recognition tests, a distance of 57 cm between the participant and the screen was kept constant. Participants had to fixate a central gray fixation cross throughout the tests; fixation was visually controlled by the examiner who was always placed left/right to the participant and was able to observe the participant's eye movements. The examiner pressed a button to start the trial: As a cue for the upcoming stimulus presentation, the cross turned black, and after a randomized interval (1–2 sec), a visual stimulus appeared. The center of mass of all object stimuli was always located 11° left or right from the central black fixation cross. The fixation cross and the stimuli were presented at the same height for all participants but those with quadrantanopia. For patients with quadrantanopia (see above), stimulus presentation in both visual half-fields was shifted upward/downward on the screen for 7.5° relative to the position of the fixation cross on the y-axis, such that all stimuli were projected within the patient's spared lower/upper visual field. Presentation time in all tests but the Fribble recognition was 60 msec. Because of the high perceptual complexity of the Fribble tasks (see below), these stimuli were shown for 80 msec. Participants had to respond verbally as soon as a stimulus was presented. RTs were recorded with a microphone placed right in front of the participants' mouth not covering the computer screen. Accuracy (ACC) of the responses was recorded by the experimenter. The stimuli were presented in a pseudorandomized order in all tasks with the same number of presentations in the left and right visual hemifield. Each object identification test started with two preparation trials that were discarded from later data analysis to familiarize the participants with the respective task.

The whole object recognition experiment encompassed two parts: first, we conducted the Unilateral Object Recognition Test Battery (Rennig et al., 2017), which was then followed by a specific test with expert objects. The Unilateral Object Recognition Test Battery presented three different kinds of object stimuli (Efron elements, real objects, and Fribbles; for details, see below) in six different tests; it assessed general abilities of lateral object recognition. The seventh test was the Expert Object Recognition Test; it did not contain any objects from the Unilateral Object Recognition Test Battery.

Efron Test. The participants discriminated a true square from three rectangles of varying dimensions (Efron, 1968). All four Efron elements had an identical surface size (Figure 2A). The square stimulus had a side length of 5°. The three rectangle stimuli had a vertical side length of 5.2°, 5.4°, or 5.6° and a corresponding horizontal length of 4.8°, 4.6°, or 4.4°. Participants indicated whether they saw a square or a rectangle in 48 trials (=12 of each stimulus).

Figure 2. 

Visual stimuli from all unilateral object identification tests: (A) all four Efron elements (drawn to scale); (B) examples of object stimuli as shapes, black/white, and color versions (Leibe & Schiele, 2003); and (C) stimuli from three Fribbles categories (Williams & Simons, 2000). (D) Object stimuli classified as “regular” objects. (E) Object stimuli classified as “expert” objects.

Figure 2. 

Visual stimuli from all unilateral object identification tests: (A) all four Efron elements (drawn to scale); (B) examples of object stimuli as shapes, black/white, and color versions (Leibe & Schiele, 2003); and (C) stimuli from three Fribbles categories (Williams & Simons, 2000). (D) Object stimuli classified as “regular” objects. (E) Object stimuli classified as “expert” objects.

Real Object Recognition. Object images were taken from the ETH-80 stimulus set from the Max-Planck Institute for Computer Science, Saarbrücken, Germany (Leibe & Schiele, 2003). Seven different objects were selected: car, cup, horse, cow, dog, apple, and pear. All objects were presented in a side-faced viewing position and an average height/width (depending on the stimulus' natural orientation) of 6.5°. From the stimulus selection, three different object test blocks were derived. In the first block, objects were presented as silhouettes; in the second block, grayscale versions of the objects were shown; and the third block contained the original colored objects (Figure 2B). In all three blocks, participants named the objects verbally. Twenty-eight experimental trials (=4 of each stimulus, twice per presentation side) were presented in each object recognition block.

Fribble Recognition. Artificial 3-D object-like nonsense elements, so called Fribbles (Williams & Simons, 2000), were used to test recognition of complex 3-D objects. From an extensive set of stimuli, two different but very similar images were selected from three Fribble categories (Figure 2C). Participants were familiarized with the stimuli from the first Fribble category: one Fribble was named “A,” the second one was labeled as “B.” Participants had to memorize the labels for the two Fribbles and later name them after lateral presentation in the recognition tests. The same procedure was applied for the two other Fribble categories/pairs. Fribbles were presented with an average height/width (depending on the stimulus' natural orientation) of 6.8°. For later data analysis, behavioral scores were simply averaged across all three blocks. The sequence of Fribble categories was kept constant over all participants. Each block comprised 24 trials.

Fribble Discrimination. Two identical or different Fribbles from the same category were presented one above the other left or right to the fixation cross. Here, only Fribble stimuli were chosen that did not exceed a vertical size of 7° when presented above each other. Participants had to indicate if the Fribbles were different or identical. In this test, 48 (=16 of each category) trials were conducted. Identical and different stimulus pairs as well as the position of single stimuli (upper/lower presentation) were equally distributed over all test trials.

Expert Object Recognition Test. This test contained 30 different images of regular everyday and 30 images of car-/garage-related objects. Object images for both categories were found on the Internet and an in-house object image database (Belardinelli, Barabas, Himmelbach, & Butz, 2016). We also made sure that our 60 stimuli were as well separable into equally distributed groups of tools and nontools (30 tools/30 nontools). However, within the subgroups, the stimuli were only almost equally balanced. In the sample of regular objects, we used 16 tools and 14 nontools. The sample of expert object stimuli consisted of 14 tools and 16 nontools. Each object image was presented twice, once on each presentation side.

The sequence of the seven tests was kept constant for all participants: Efron shapes, real object recognition shapes, real object recognition grayscale, real object recognition color, Fribble recognition, Fribble discrimination, and Expert Object Recognition Test. With respect to the fact that the current work is a single-case study, we intended to make sure to measure behavioral capabilities of interest as thorough as possible and therefore repeated the Expert Object Recognition Test for our primary participant HWS. We performed the Unilateral Object Recognition Test Battery in HWS on the first day of testing in the standard sequence; on the second day, we conducted two runs of the Expert Object Recognition Test. For data analysis, we averaged HWS's performance over the two runs of the Expert Object Recognition Test.

MRI Data Acquisition and Analysis

All patients had circumscribed brain lesions due to ischemic stroke or hemorrhage demonstrated by MRI or CT. In the acute patients who underwent MRI scanning at admission, we used diffusion-weighted imaging within the first 48 hr poststroke and T2-weighted fluid-attenuated inversion recovery sequences when imaging was conducted 48 hr or later after stroke onset. Under both protocols, the initial scanning was optionally repeated during the following days until a firm diagnosis could be made and the infarcted area became clearly demarcated. The final scans were used for this study. In the acute patients, the mean time between stroke and imaging was 1.7 days (SD = 2.5). For chronic patients, the mean time between stroke and imaging was 796.9 days (SD = 1274.8). For all acute patients, the time between neuroimaging and behavioral testing ranged between 1 and 11 days (mean = 3.5, SD = 3.0). The majority of chronic patients were tested behaviorally within 1 year after chronic neuroimaging (mean = 334.6, SD = 342.3).

Lesion location was evaluated using MRIcron software (Rorden, Karnath, & Bonilha, 2007; www.mricro.com). The experimenter delineated the boundary of the lesion directly on the digital scans for every single transverse slice using MRIcron software. Both the MRI/CT scan and the lesion shape were then mapped into stereotaxic space using the normalization algorithm provided by the Clinical Toolbox (Rorden, Bonilha, Fridriksson, Bender, & Karnath, 2012) based on SPM8 (fil.ion.ucl.ac.uk/spm). This method uses MRI and CT normalization templates from aged brains, allowing valid normalization of lesioned brains from several imaging modalities. HWS's normalized lesion size (30,045 mm3) was not significantly different from nonventral control patients (mean lesion size = 23,001 mm3, SD = 40,075; t(18) = 0.17, p = .43) and ventral control patients (mean lesion size = 17,681 mm3, SD = 23,535; t(8) = 0.50, p = .32). Also, the two groups of control patients were not significantly different (t(26) = 0.37, p = .47).

RESULTS

Neuropsychological Screening

Performance in neuropsychological screening was part of the inclusion/exclusion criteria. Patient HWS as well as all control patients included into the study performed normal in tests addressing spatial neglect, visual extinction, subclinical spatial attention deficits, and auditory object identification.

Visual Object Recognition under Unrestricted Viewing Conditions

Participant performance is reported in Table 1. Patient HWS performed normal in all subtests. Three ventral controls and three nonventral control patients showed deficits in a single subtest of the VOSP object recognition battery. These patients were able to perform normal in all other subtests.

Unilateral Object Recognition Tests

We calculated an “overall object recognition score” by averaging percent correct values for each participant per presentation side over all six object recognition tests from the Unilateral Object Recognition Test Battery. Figure 3 illustrates this score for the contra- and ipsilesional hemifields of the three control groups and HWS. We compared HWS's performance in the ipsi- and contralesional hemifield with that of the three control groups by applying Crawford's adjusted single-impairment t statistic and the Revised Standardized Difference Test (RSDT; Crawford & Garthwaite, 2002, 2005). The RSDT method investigates if the performance difference between conditions in a single patient is significantly different from the respective differences in a control group. This is followed by individual comparisons of the patient's performance in either condition, with the respective performance of controls. In addition, we applied standard statistical routines for group comparisons to test for potential differences between the control groups. As dependent variable, we used the “overall object recognition score.” Because there is no lesion to refer to in healthy control participants, we assigned the labels “contralesional” and “ipsilesional” randomly to either the left or the right hemifield in healthy controls. We adjusted the significance threshold for the single-case comparisons with a Bonferroni correction for three statistical tests (HWS compared to each of three control groups), resulting in a p threshold of .016 for the dissociation tests and a p threshold of .025 for the two following single-impairments t tests. Comparing HWS's performance statistically against healthy controls (HC), nonventral control patients (NP), and ventral control patients (VP), we observed significant dissociations for HC and NP, but not for VP (HC: t(19) = 4.60, p < .001, r = 0.59, between contra- and ipsilateral stimulus presentations within this control group; NP: t(18) = 5.07, p < .001, r = 0.88; VP: t(8) = 1.75, p = .12, r = 0.51). Following single-impairment t tests revealed significant differences compared with HC and NP for the contralesional hemifield (HC: t(19) = 4.70, p < .001; NP: t(18) = 2.68, p < .001), but not for the ipsilesional hemifield (HC: t(19) = 0.24, p = .41; NP: t(18) = −.11, p = .45). To examine if the VP showed symptoms of visual hemiagnosia with respect to both other control groups (HC, NP), we calculated a 2 × 3 ANOVA with the within-subject factor Presentation side (contra- and ipsilesional) and the between-subject factor Group (HC, NP, VP). The ANOVA revealed a significant interaction (F(2, 45) = 8.13, p = .001) as well as significant main effects for both variables (Presentation side: F(1, 45) = 8.13, p < .001; Group: F(2, 45) = 5.51, p = .007). Subsequent t tests between the contra- and ipsilesional presentation side for every group revealed a significant result for VP, but not HC and NP (with a Bonferroni-corrected p threshold of .017; HC: t(19) = −1.24, p = .231; NP: t(18) = 0.93, p = .364; VP: t(8) = 3.06, p = .016). We also calculated a one-way ANOVA per presentation side over the three groups. For the contralesional presentation side, the ANOVA was significant (F(2, 45) = 10.25, p < .001) whereas no significant effect was observable for the ipsilesional side (F(2, 45) = 1.10, p = .342). Subsequent t tests (with a Bonferroni-corrected p threshold of .017) showed a significant difference between VP and HC (t(27) = 4.97, p < .001) as well as VP and NP (t(26) = 2.56, p = .016), but not between HC and NP (t(37) = 2.05, p = .049).

Figure 3. 

Behavioral results from the Unilateral Object Recognition Test Battery. Accuracy in percent correct is given for healthy controls (HC), nonventral control patients (NP), ventral control patients (VP), and HWS for the contralesional (CON) and ipsilesional (IPSI) presentation side. The graph shows the summarized object recognition score averaged over all tests of the Unilateral Object Recognition Test Battery. The error bars for the results of the control groups represent the SEM.

Figure 3. 

Behavioral results from the Unilateral Object Recognition Test Battery. Accuracy in percent correct is given for healthy controls (HC), nonventral control patients (NP), ventral control patients (VP), and HWS for the contralesional (CON) and ipsilesional (IPSI) presentation side. The graph shows the summarized object recognition score averaged over all tests of the Unilateral Object Recognition Test Battery. The error bars for the results of the control groups represent the SEM.

These results demonstrate that HWS's object recognition performance for stimuli presented in his contralesional hemifield was impaired whereas recognition performance for stimuli in his ipsilesional hemifield was comparable to the recognition performance of the healthy controls and nonventral control patients. Compared with nonventral control patients and healthy controls, the ventral control group also showed an inferior recognition performance for stimuli presented in the contralesional hemifield with a normal performance in the ipsilesional hemifield (Figure 3). In summary, these analyses demonstrated that HWS as well as the ventral control group suffered from visual hemiagnosia.

Expert Object Recognition Test

Data from the Expert Object Recognition Test are shown in Figure 4. The statistical test comprised two factors as well as two levels per analysis: the ipsi- and contralesional presentation side and the classification of objects into two groups (“expert” vs. “regular” or tool vs. nontool). Because the RSDT method (Crawford & Garthwaite, 2005) cannot handle a 2 × 2 design, we reduced the number of factors by subtracting the recognition performance of the ipsilesional from the contralesional performance per object category. This factor reduction was done for every participant to use these difference values to compare HWS's performance with the control groups using the RSDT method (Crawford & Garthwaite, 2005). The difference values should group closely around zero for all object categories in case of no particular preference of any object category on any presentation side. In case of, for example, a better recognition rate for “expert” objects on the contralateral presentation side, a negative value would emerge. In case of two negative values in both object categories (as expected in HWS), the lower (=more negative) value indicates a stronger perceptual deficit for this particular object category. Table 2 gives the descriptive data in percent correct for the Expert Object Recognition Test, both presentation sides and “expert” and “regular” objects as well as tools and nontools. The table also illustrates the construction of the difference value used for the statistical analysis.

Figure 4. 

Behavioral results from the Unilateral Expert Object Recognition Test. Accuracy in percent correct is given for healthy controls (HC), nonventral control patients (NP), ventral control patients (VP), and HWS for the contralesional (CON) and ipsilesional (IPSI) presentation side. The first panel (A) gives the results split up for “regular (Reg)” and “expert (Exp)” objects. The second panel (B) shows the results split up for “Nontools (NT)” and “Tools.” The error bars for the results of the control groups represent the SEM.

Figure 4. 

Behavioral results from the Unilateral Expert Object Recognition Test. Accuracy in percent correct is given for healthy controls (HC), nonventral control patients (NP), ventral control patients (VP), and HWS for the contralesional (CON) and ipsilesional (IPSI) presentation side. The first panel (A) gives the results split up for “regular (Reg)” and “expert (Exp)” objects. The second panel (B) shows the results split up for “Nontools (NT)” and “Tools.” The error bars for the results of the control groups represent the SEM.

Table 2. 

Percent Correct Recognition of the Different Stimulus Categories by HWS, the Healthy Control Group, the Nonventral Control Patients, and the Ventral Control Patients

HWSHealthy ControlsNonventral Control PatientsVentral Control Patients
ContraIpsiDiffContraIpsiDiffContraIpsiDiffContraIpsiDiff
Expert objects 57 85 −28 69 (13) 70 (15) −1 (11) 63 (19) 63 (18) 1 (11) 61 (18) 66 (12) −6 (17) 
Regular objects 32 80 −48 72 (14) 71 (13) 1 (8) 60 (14) 63 (14) −2 (7) 58 (18) 66 (17) −8 (15) 
Tools 42 85 −43 71 (10) 71 (13) 0 (9) 64 (13) 64 (15) 0 (10) 55 (22) 63 (10) −8 (21) 
Nontools 47 83 −36 71 (15) 70 (13) 0 (9) 59 (15) 62 (15) −2 (7) 48 (13) 59 (18) −11 (14) 
HWSHealthy ControlsNonventral Control PatientsVentral Control Patients
ContraIpsiDiffContraIpsiDiffContraIpsiDiffContraIpsiDiff
Expert objects 57 85 −28 69 (13) 70 (15) −1 (11) 63 (19) 63 (18) 1 (11) 61 (18) 66 (12) −6 (17) 
Regular objects 32 80 −48 72 (14) 71 (13) 1 (8) 60 (14) 63 (14) −2 (7) 58 (18) 66 (17) −8 (15) 
Tools 42 85 −43 71 (10) 71 (13) 0 (9) 64 (13) 64 (15) 0 (10) 55 (22) 63 (10) −8 (21) 
Nontools 47 83 −36 71 (15) 70 (13) 0 (9) 59 (15) 62 (15) −2 (7) 48 (13) 59 (18) −11 (14) 

Percent correct recognition is listed separately for the contra- and ipsilesional presentation sides (“Contra”; “Ipsi”). Moreover, a difference value (Diff; difference of contra- minus ipsilesional) for “expert” and “regular” objects as well as tools and nontools from the Expert Object Recognition Test for HWS and every control group is given. The mean value for HWS is his individual performance; the mean values for the healthy control group and the two control patient groups represent group mean values with standard deviation in brackets. The respective difference value between the ipsi- and contralateral presentation sides was used to statistically compare HWS's performance for “expert” and “regular” objects (as well as tools and nontools) against the three control groups.

The statistical results are given in Table 3. We performed two different statistical analyses: First, we compared “expert” to “regular” objects and then an additional analysis comparing tools to nontools. We adjusted the significance threshold per analysis with a Bonferroni correction for three statistical comparisons (for three control groups) for each comparison, resulting in a p threshold of .017 for the dissociation tests and a p threshold of .025 for the two following single-subject t tests.

Table 3. 

Statistical Results

tptptpr
ExpertRegularDissociation
HC −2.402 .013* −6.014 <.001* 2.621 .016* .11 
NP −2.552 .011* −6.610 <.001* 3.374 .003* .35 
VP −1.228 .129 −2.594 .018* 4.333 .004* .97 
 Tools Nontools Dissociation 
HC −3.653 .001* −4.012 <.001* 0.324 .749 .43 
NP −5.179 <.001* −4.617 <.001* −0.545 .593 .52 
VP −1.727 .064 −1.744 .062 0.033 .975 .91 
tptptpr
ExpertRegularDissociation
HC −2.402 .013* −6.014 <.001* 2.621 .016* .11 
NP −2.552 .011* −6.610 <.001* 3.374 .003* .35 
VP −1.228 .129 −2.594 .018* 4.333 .004* .97 
 Tools Nontools Dissociation 
HC −3.653 .001* −4.012 <.001* 0.324 .749 .43 
NP −5.179 <.001* −4.617 <.001* −0.545 .593 .52 
VP −1.727 .064 −1.744 .062 0.033 .975 .91 

Single-case analyses for HWS compared with healthy controls (HC; df = 19), nonventral control patients (NP; df = 18), as well as ventral control patients (VP; df = 8). We applied Crawford's RSDT and single-impairment t tests (Crawford & Garthwaite, 2002, 2005) for the comparison of the difference values between the ipsi- and contralateral presentation sides for “expert” and “regular” objects as well as for the same analysis for the comparison between tools and nontools. Bonferroni correction for three statistical tests (for three control groups per analysis) per comparison resulted in a p threshold of .017 for the dissociation tests and a p threshold of .025 for the two following single-subject t tests. The table shows t and p values of the respective tests as well as correlation coefficients r between contra- and ipsilateral stimulus presentations within the respective control group. Significant results are marked with an asterisk(*).

The analysis of the “expert” versus “regular” objects differences revealed significant dissociations in HWS relative to all three control groups (see Table 3). The single impairment t tests for either condition comparing HWS to the healthy and the nonventral control group showed significant differences for “regular” and “expert” objects. The comparison to ventral control patients showed no significant effects in the post hoc analysis. This result pattern indicates a significant perceptual advantage in HWS for “expert” objects over “regular” objects on the contralesional presentation side.

In the additional analysis comparing recognition of objects classified as tools and nontools, no significant dissociation in the comparisons of HWS to any of the three control groups emerged. We nevertheless conducted single-subject t tests for all conditions and found significant differences between HWS and healthy controls and the nonventral control patients, respectively, for tool and for nontools. No significant effect emerged for the ventral control patients. These results confirm the presence of visual hemiagnosia in HWS and ventral control patients without any category-specific effects.

DISCUSSION

In this study, we asked whether visual expertise acquired before the onset of a perceptual impairment can compensate unilateral object recognition deficits. We tested a patient with a lesion to the right medial ventrotemporal lobe and visual hemiagnosia using unilateral object recognition tests with “regular” object stimuli and “expert” objects. HWS showed a significantly better performance for “expert” compared with “regular” stimuli in his affected contralesional hemifield. The results of the Expert Object Recognition Test indicate that the difference value between ipsi- and contralateral presentations in HWS was significantly higher for “regular” objects than for “expert” objects. This means that there was a significantly stronger deficit for the recognition of “regular” objects than for “expert” objects on the contralesional presentation side. On the ipsilesional side, HWS recognized both object categories in a comparable manner. This result pattern was observable comparing HWS statistically against healthy controls, nonventral control patients not suffering from visual hemiagnosia, as well as ventral control patients that showed symptoms of visual hemiagnosia. The present results are in line with two previous patient studies that demonstrated that visual expertise can compensate for agnosic recognition deficits (Bukach et al., 2006; Behrmann et al., 2005). However, in contrast to these studies that trained patients after suffering a perceptual impairment, we were able to demonstrate that visual expertise acquired before the onset of a perceptual impairment can compensate perceptual deficits.

The result was not influenced by a disproportional presence of tools in the “expert” object category. Recent neuroimaging studies showed that recognition of tools compared with nontool objects activates different neuronal mechanisms with significant dorsal contributions (Mahon et al., 2007; Noppeney, Price, Penny, & Friston, 2006; Chao & Martin, 2000). In case of a disparity of tools and nontools, with more tools in the “expert” object category, contributions of intact dorsal areas in HWS could explain the present results. However, tools and nontools were sufficiently balanced in both object categories. Moreover, we did not observe a significant dissociation in an additional analysis comparing the recognition performance of tools against nontools on the ipsi- and contralateral presentation side. The absence of a significant dissociation shows that there was no specific perceptual advantage for one of the two categories, tools and nontools, on any presentation side.

The current results may be caused by generally enhanced information processing for expert objects that does not need a fully intact ventral system and/or a compensation in the intact hemisphere that is more effective for “expert” than for “regular” objects. These very broad explanations are backed by studies that showed a significantly higher recognition rate as well as a faster processing speed for expert objects (Gauthier, Williams, Tarr, & Tanaka, 1998; Gauthier & Tarr, 1997; Ericsson & Lehmann, 1996). Further support for an enhanced information processing along the ventral pathway for visual expert material comes from neuroimaging studies that found stronger neuronal signals for expert compared with nonexpert objects in early visual (Harel, Gilaie-Dotan, Malach, & Bentin, 2010) as well as early object-sensitive areas along the ventral pathway (Brants, Wagemans, & Op de Beeck, 2011; Harel et al., 2010; Wong, Palmeri, Rogers, Gore, & Gauthier, 2009; Op de Beeck, Baker, DiCarlo, & Kanwisher, 2006).

The present results suggest that top–down mechanisms influence object recognition at an early processing stage and emphasize the importance of areas outside the FFA in visual expertise. Patient HWS showed hemiagnosia for contralesionally presented everyday objects after an extensive right hemispheric lesion encompassing the fusiform and parahippocampal gyri. The observation of relatively better preserved recognition for highly familiar (“expert”) compared with “regular” stimuli in HWS allows to speculate that processing of visual expertise already happens at an early stage of perception, namely, before this information enters the (in HWS lesioned) modules of higher object processing, like the FFA (Kanwisher et al., 1997) or parahippocampal structures (Epstein & Kanwisher, 1998). The lateral occipital complex or even early visual areas might be differently tuned for perception in case of highly familiar (vs. “regular”) object stimuli. This assumption is supported by studies showing neuronal effects of visual object expertise in early visual and object-sensitive areas (Brants et al., 2011; Harel et al., 2010; Wong et al., 2009; Op de Beeck et al., 2006) as well as in posterior temporoparietal areas (Rennig et al., 2013, 2015; Bilalić et al., 2012; van der Linden et al., 2010; Moore et al., 2006). One possible mechanism contributing to expert object processing is holistic top–down vision (or the so-called Gestalt perception), which might be available to a bigger extent for processing of visual content of expertise. This explanation is derived from a recent neuroimaging study (Rennig et al., 2013) that demonstrated that areas previously associated with Gestalt perception (Huberle & Karnath, 2012) were significantly stronger involved in experts compared with novices while viewing their stimuli of expertise over a wide range of behavioral paradigms (Bilalić et al., 2012; Bilalić, Kiesel, Pohl, Erb, & Grodd, 2011; Bilalić, Langner, et al., 2011; Bilalić, Langner, Erb, & Grodd, 2010). In particular, it is possible to speculate that experts engage fundamentally different strategies in perception of expert objects taking advantage of a (learned) holistic approach that is different from a novice who might use more serial strategies of object processing. The holistic approach (driven by posterior temporoparietal brain areas) makes perception more efficient and less error-prone and could explain the compensation mechanisms for “expert” over “regular” objects in HWS.

The proposed mechanisms of Gestalt perception are comparable to a mechanism of object processing involving specific processing of low spatial frequency object components (Bar et al., 2006; Bar, 2003). In this theory on object processing, it is proposed that low spatial frequency components of objects are processed on a fast route through the dorsal stream where the gist of an object (i.e., its global form) is processed in the OFC and a first guess about the object's identity is established. Object details (coded in high spatial frequency components) are processed much slower through the ventral pathway. The low-frequency information is then integrated into object processing along the ventral stream to facilitate object recognition. This process could be even more efficient for expert objects where an initial guess about the object's identity is more likely to be correct due to the availability of specific templates for expert objects. HWS had no damage to brain areas along the dorsal stream; the behavioral results discovered in HWS thus would be in line with the hypothesis of Bar and colleagues (Bar et al., 2006; Bar, 2003). This explanation is also supported by the recent argumentation that several cortical regions beyond FFA interact in expert object processing, arguing against a role of FFA as an isolated hub of visual expertise (Harel, Kravitz, & Baker, 2014; Harel et al., 2013).

Several neuroimaging studies of visual expertise have shown that areas in the left hemisphere homologue of the right hemisphere FFA were significantly stronger involved in experts compared with novices when viewing objects from their respective domain of expertise (Bilalić et al., 2012, 2016; Harel et al., 2010; Gauthier, Tarr, Anderson, Skudlarski, & Gore, 1999). These findings might help in explaining the perceptual characteristics discovered in HWS. Despite of a significant lesion to his right hemispheric FFA, HWS was able to recognize significantly more “expert” than “regular” objects. It is possible that enhanced information processing capacities in his left hemispheric FFA homologue enabled him to compensate for “expert” objects in particular.

There are numerous other mechanisms that also could explain the current results, like conceptual expert knowledge or even emotional factors that might increase recognition for expert objects. On a behavioral level, it has been demonstrated that visual expertise is highly correlated with conceptual knowledge (Barton, Hanif, & Ashraf, 2009) and that conceptual/semantic knowledge in the respective domain of visual expertise is significantly associated with prefrontal brain areas (Gilaie-Dotan, Harel, Bentin, Kanai, & Rees, 2012). Because HWS did not show lesions outside the ventral stream, his conceptual and semantic knowledge about his objects of visual expertise and contributions from intact frontal areas could explain the results. Emotional/motivational factors might as well play a role in the expert object recognition abilities of HWS. Most experts have a long-lasting career in their field of expertise where knowledge is acquired with high enthusiasm and positive attribution toward the field of expertise. The same is true for HWS, who, based on his own reports and his achievements during his career, had a very positive attitude toward all car-/garage-related content. It has been demonstrated that affective object recognition is associated with structures in the OFC (Barrett & Bar, 2009). Because this area is spared in HWS, contributions from these regions might have compensated his perceptual abilities for expert objects on his affected contralesional hemifield.

This study shows that long-term visual expertise acquired before the onset of a perceptual impairment can compensate for deficits of visual object recognition. More importantly, the current work demonstrates that top–down mechanisms of visual expertise influence object recognition at an early processing stage, probably before visual object information propagates to modules of higher object recognition, like the FFA. Because HWS showed a lesion to the fusiform gyrus (comprising the FFA) and spared capacities of expert object processing, the current findings emphasize possible contributions of areas outside the ventral stream to visual expertise.

Acknowledgments

This work was supported by the DFG (Ka 1258/20-1, Ka 1258/23-1) and the European Union (ERC StG 211078). We want to thank Jonas Walter, Christoph Sperber, and Anna Bleyer for their help with data acquisition.

Reprint requests should be sent to Johannes Rennig, Division of Neuropsychology, Center of Neurology, University of Tübingen, D-72076 Tübingen, Germany, or via e-mail: johannes.rennig@uni-tuebingen.de.

REFERENCES

REFERENCES
Bar
,
M.
(
2003
).
A cortical mechanism for triggering top–down facilitation in visual object recognition
.
Journal of Cognitive Neuroscience
,
15
,
600
609
.
Bar
,
M.
,
Kassam
,
K. S.
,
Ghuman
,
A. S.
,
Boshyan
,
J.
,
Schmid
,
A. M.
,
Dale
,
A. M.
, et al
(
2006
).
Top–down facilitation of visual recognition
.
Proceedings of the National Academy of Sciences, U.S.A.
,
103
,
449
454
.
Barrett
,
L. F.
, &
Bar
,
M.
(
2009
).
See it with feeling: Affective predictions during object perception
.
Philosophical Transactions of the Royal Society of London, Series B: Biological Science
,
364
,
1325
1334
.
Barton
,
J. J. S.
,
Hanif
,
H.
, &
Ashraf
,
S.
(
2009
).
Relating visual to verbal semantic knowledge: The evaluation of object recognition in prosopagnosia
.
Brain
,
132
,
3456
3466
.
Becker
,
E.
, &
Karnath
,
H.-O.
(
2007
).
Incidence of visual extinction after left versus right hemisphere stroke
.
Stroke
,
38
,
3172
3174
.
Behrmann
,
M.
,
Marotta
,
J.
,
Gauthier
,
I.
,
Tarr
,
M. J.
, &
McKeeff
,
T. J.
(
2005
).
Behavioral change and its neural correlates in visual agnosia after expertise training
.
Journal of Cognitive Neuroscience
,
17
,
554
568
.
Belardinelli
,
A.
,
Barabas
,
M.
,
Himmelbach
,
M.
, &
Butz
,
M. V.
(
2016
).
Anticipatory eye fixations reveal tool knowledge for tool interaction
.
Experimental Brain Research
,
234
,
2415
2431
.
Bilalić
,
M.
,
Grottenthaler
,
T.
,
Nägele
,
T.
, &
Lindig
,
T.
(
2016
).
The faces in radiological images: Fusiform face area supports radiological expertise
.
Cerebral Cortex
,
26
,
1004
1014
.
Bilalić
,
M.
,
Kiesel
,
A.
,
Pohl
,
C.
,
Erb
,
M.
, &
Grodd
,
W.
(
2011
).
It takes two-skilled recognition of objects engages lateral areas in both hemispheres
.
PLoS One
,
6
,
e16202
.
Bilalić
,
M.
,
Langner
,
R.
,
Erb
,
M.
, &
Grodd
,
W.
(
2010
).
Mechanisms and neural basis of object and pattern recognition: A study with chess experts
.
Journal of Experimental Psychology: General
,
139
,
728
742
.
Bilalić
,
M.
,
Langner
,
R.
,
Ulrich
,
R.
, &
Grodd
,
W.
(
2011
).
Many faces of expertise: Fusiform face area in chess experts and novices
.
Journal of Neuroscience
,
31
,
10206
10214
.
Bilalić
,
M.
,
Turella
,
L.
,
Campitelli
,
G.
,
Erb
,
M.
, &
Grodd
,
W.
(
2012
).
Expertise modulates the neural basis of context dependent recognition of objects and their relations
.
Human Brain Mapping
,
33
,
2728
2740
.
Brainard
,
D. H.
(
1997
).
The psychophysics toolbox
.
Spatial Vision
,
10
,
433
436
.
Brants
,
M.
,
Wagemans
,
J.
, &
Op de Beeck
,
H. P.
(
2011
).
Activation of fusiform face area by Greebles is related to face similarity but not expertise
.
Journal of Cognitive Neuroscience
,
23
,
3949
3958
.
Bukach
,
C. M.
,
Bub
,
D. N.
,
Gauthier
,
I.
, &
Tarr
,
M. J.
(
2006
).
Perceptual expertise effects are not all or none: Spatially limited perceptual expertise for faces in a case of prosopagnosia
.
Journal of Cognitive Neuroscience
,
18
,
48
63
.
Chao
,
L. L.
, &
Martin
,
A.
(
2000
).
Representation of manipulable man-made objects in the dorsal stream
.
Neuroimage
,
12
,
478
484
.
Cornelsen
,
S.
,
Rennig
,
J.
, &
Himmelbach
,
M.
(
2016
).
Memory-guided reaching in a patient with visual hemiagnosia
.
Cortex
,
79
,
32
41
.
Crawford
,
J. R.
, &
Garthwaite
,
P. H.
(
2002
).
Investigation of the single case in neuropsychology: Confidence limits on the abnormality of test scores and test score differences
.
Neuropsychologia
,
40
,
1196
1208
.
Crawford
,
J. R.
, &
Garthwaite
, >
P. H.
(
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
.
Damasio
,
A. R.
,
Damasio
,
H.
, &
Van Hoesen
,
G. W.
(
1982
).
Prosopagnosia: Anatomic basis and behavioral mechanisms
.
Neurology
,
32
,
331
341
.
DiCarlo
,
J. J.
, &
Maunsell
,
J. H. R.
(
2003
).
Anterior inferotemporal neurons of monkeys engaged in object recognition can be highly sensitive to object retinal position
.
Journal of Neurophysiology
,
89
,
3264
3278
.
Efron
,
R.
(
1968
).
What is perception?
In
R.
Cohen
&
M.
Wartofsky
(Eds.),
Boston studies in the philosophy of science
(pp.
137
173
).
New York
:
Humanities
.
Epstein
,
R.
, &
Kanwisher
,
N.
(
1998
).
A cortical representation of the local visual environment
.
Nature
,
392
,
598
601
.
Ericsson
,
K. A.
, &
Lehmann
,
A. C.
(
1996
).
Expert and exceptional performance: Evidence of maximal adaptation to task constraints
.
Annual Review of Psychology
,
47
,
273
305
.
Farah
,
M.
(
2004
).
Visual agnosia
(2nd ed.).
Cambridge, MA
:
Bradford Books
.
Ferber
,
S.
, &
Karnath
,
H.-O.
(
2001
).
How to assess spatial neglect—line bisection or cancellation tasks?
Journal of Clinical and Experimental Neuropsychology
,
23
,
599
607
.
Gauthier
,
I.
,
Curran
,
T.
,
Curby
,
K. M.
, &
Collins
,
D.
(
2003
).
Perceptual interference supports a non-modular account of face processing
.
Nature Neuroscience
,
6
,
428
432
.
Gauthier
,
I.
,
Skudlarski
,
P.
,
Gore
,
J. C.
, &
Anderson
,
A. W.
(
2000
).
Expertise for cars and birds recruits brain areas involved in face recognition
.
Nature Neuroscience
,
3
,
191
197
.
Gauthier
,
I.
, &
Tarr
,
M. J.
(
1997
).
Becoming a “Greeble” expert: Exploring mechanisms for face recognition
.
Vision Research
,
37
,
1673
1682
.
Gauthier
,
I.
,
Tarr
,
M. J.
,
Anderson
,
A. W.
,
Skudlarski
,
P.
, &
Gore
,
J. C.
(
1999
).
Activation of the middle fusiform “face area” increases with expertise in recognizing novel objects
.
Nature Neuroscience
,
2
,
568
573
.
Gauthier
,
I.
,
Williams
,
P.
,
Tarr
,
M. J.
, &
Tanaka
,
J.
(
1998
).
Training “greeble” experts: A framework for studying expert object recognition processes
.
Vision Research
,
38
,
2401
2428
.
Gauthier
,
L.
,
Dehaut
,
F.
, &
Joanette
,
Y.
(
1989
).
The Bells Test: A quantitative and qualitative test for visual neglect
.
International Journal of Clinical Neuropsychology
,
11
,
49
54
.
Gilaie-Dotan
,
S.
,
Harel
,
A.
,
Bentin
,
S.
,
Kanai
,
R.
, &
Rees
,
G.
(
2012
).
Neuroanatomical correlates of visual car expertise
.
Neuroimage
,
62
,
147
153
.
Golomb
,
J. D.
, &
Kanwisher
,
N.
(
2012
).
Higher level visual cortex represents retinotopic, not spatiotopic, object location
.
Cerebral Cortex
,
22
,
2794
2810
.
Goodale
,
M. A.
,
Milner
,
A. D.
,
Jakobson
,
L. S.
, &
Carey
,
D. P.
(
1991
).
A neurological dissociation between perceiving objects and grasping them
.
Nature
,
349
,
154
156
.
Grill-Spector
,
K.
,
Kushnir
,
T.
,
Edelman
,
S.
,
Avidan
,
G.
,
Itzchak
,
Y.
, &
Malach
,
R.
(
1999
).
Differential processing of objects under various viewing conditions in the human lateral occipital complex
.
Neuron
,
24
,
187
203
.
Harel
,
A.
,
Gilaie-Dotan
,
S.
,
Malach
,
R.
, &
Bentin
,
S.
(
2010
).
Top–down engagement modulates the neural expressions of visual expertise
.
Cerebral Cortex
,
20
,
2304
2318
.
Harel
,
A.
,
Kravitz
,
D.
, &
Baker
,
C. I.
(
2013
).
Beyond perceptual expertise: Revisiting the neural substrates of expert object recognition
.
Frontiers in Human Neuroscience
,
7
,
885
.
Harel
,
A.
,
Kravitz
,
D. J.
, &
Baker
,
C. I.
(
2014
).
Holding a stick at both ends: On faces and expertise
.
Frontiers in Human Neuroscience
,
8
,
442
.
Hemond
,
C. C.
,
Kanwisher
,
N. G.
, &
Op de Beeck
,
H. P.
(
2007
).
A preference for contralateral stimuli in human object- and face-selective cortex
.
PLoS One
,
2
,
e574
.
Huberle
,
E.
, &
Karnath
,
H.-O.
(
2012
).
The role of temporo-parietal junction (TPJ) in global Gestalt perception
.
Brain Structure and Function
,
217
,
735
746
.
Hung
,
C. P.
,
Kreiman
,
G.
,
Poggio
,
T.
, &
DiCarlo
,
J. J.
(
2005
).
Fast readout of object identity from macaque inferior temporal cortex
.
Science
,
310
,
863
866
.
Johannsen
,
L.
, &
Karnath
,
H.-O.
(
2004
).
How efficient is a simple copying task to diagnose spatial neglect in its chronic phase?
Journal of Clinical and Experimental Neuropsychology
,
26
,
251
256
.
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
.
Karnath
,
H.-O.
,
Rennig
,
J.
,
Johannsen
,
L.
, &
Rorden
,
C.
(
2011
).
The anatomy underlying acute versus chronic spatial neglect: A longitudinal study
.
Brain
,
134
,
903
912
.
Karnath
,
H.-O.
,
Rüter
,
J.
,
Mandler
,
A.
, &
Himmelbach
,
M.
(
2009
).
The anatomy of object recognition--visual form agnosia caused by medial occipitotemporal stroke
.
Journal of Neuroscience
,
29
,
5854
5862
.
Kravitz
,
D. J.
,
Kriegeskorte
,
N.
, &
Baker
,
C. I.
(
2010
).
High-level visual object representations are constrained by position
.
Cerebral Cortex
,
20
,
2916
2925
.
Krawczyk
,
D. C.
,
Boggan
,
A. L.
,
McClelland
,
M. M.
, &
Bartlett
,
J. C.
(
2011
).
The neural organization of perception in chess experts
.
Neuroscience Letters
,
499
,
64
69
.
Leibe
,
B.
, &
Schiele
,
B.
(
2003
).
Analyzing appearance and contour based methods for object categorization
. In
2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003. Proceedings. IEEE Comput. Soc.
, p. >
II-409-15
.
Levy
,
I.
,
Hasson
,
U.
,
Avidan
,
G.
,
Hendler
,
T.
, &
Malach
,
R.
(
2001
).
Center-periphery organization of human object areas
.
Nature Neuroscience
,
4
,
533
539
.
Logothetis
,
N. K.
,
Pauls
,
J.
, &
Poggio
,
T.
(
1995
).
Shape representation in the inferior temporal cortex of monkeys
.
Current Biology
,
5
,
552
563
.
Mahon
,
B. Z.
,
Milleville
,
S. C.
,
Negri
,
G. A. L.
,
Rumiati
,
R. I.
,
Caramazza
,
A.
, &
Martin
,
A.
(
2007
).
Action-related properties shape object representations in the ventral stream
.
Neuron
,
55
,
507
520
.
Malach
,
R.
,
Levy
,
I.
, &
Hasson
,
U.
(
2002
).
The topography of high-order human object areas
.
Trends in Cognitive Sciences
,
6
,
176
184
.
Malach
,
R.
,
Reppas
,
J. B.
,
Benson
,
R. R.
,
Kwong
,
K. K.
,
Jiang
,
H.
,
Kennedy
,
W. A.
, et al
(
1995
).
Object-related activity revealed by functional magnetic resonance imaging in human occipital cortex
.
Proceedings of the National Academy of Sciences, U.S.A.
,
92
,
8135
8139
.
Marcell
,
M. M.
,
Borella
,
D.
,
Greene
,
M.
,
Kerr
,
E.
, &
Rogers
,
S.
(
2000
).
Confrontation naming of environmental sounds
.
Journal of Clinical and Experimental Neuropsychology
,
22
,
830
864
.
Mazzucchi
,
A.
,
Posteraro
,
L.
,
Nuzzi
,
G.
, &
Parma
,
M.
(
1985
).
Unilateral visual agnosia
.
Cortex
,
21
,
309
316
.
Moore
,
C. D.
,
Cohen
,
M. X.
, &
Ranganath
,
C.
(
2006
).
Neural mechanisms of expert skills in visual working memory
.
Journal of Neuroscience
,
26
,
11187
11196
.
Niemeier
,
M.
,
Goltz
,
H. C.
,
Kuchinad
,
A.
,
Tweed
,
D. B.
, &
Vilis
,
T.
(
2005
).
A contralateral preference in the lateral occipital area: Sensory and attentional mechanisms
.
Cerebral Cortex
,
15
,
325
331
.
Noppeney
,
U.
,
Price
,
C. J.
,
Penny
,
W. D.
, &
Friston
,
K. J.
(
2006
).
Two distinct neural mechanisms for category-selective responses
.
Cerebral Cortex
,
16
,
437
445
.
Op de Beeck
,
H. P.
,
Baker
,
C. I.
,
DiCarlo
,
J. J.
, &
Kanwisher
,
N. G.
(
2006
).
Discrimination training alters object representations in human extrastriate cortex
.
Journal of Neuroscience
,
26
,
13025
13036
.
Rennig
,
J.
,
Bilalić
,
M.
,
Huberle
,
E.
,
Karnath
,
H.-O.
, &
Himmelbach
,
M.
(
2013
).
The temporo-parietal junction contributes to global gestalt perception-evidence from studies in chess experts
.
Frontiers in Human Neuroscience
,
7
,
513
.
Rennig
,
J.
,
Himmelbach
,
M.
,
Huberle
,
E.
, &
Karnath
,
H.-O.
(
2015
).
Involvement of the TPJ area in processing of novel global forms
.
Journal of Cognitive Neuroscience
,
27
,
1587
1600
.
Rennig
,
J.
,
Karnath
,
H.-O.
,
Cornelsen
,
S.
,
Wilhelm
,
H.
, &
Himmelbach
,
M.
(
in press
).
Hemifield coding in ventral object-sensitive areas - Evidence from visual hemiagnosia
.
Cortex
.
doi: 10.1016/j.cortex.2017.06.011
.
Rorden
,
C.
,
Bonilha
,
L.
,
Fridriksson
,
J.
,
Bender
,
B.
, &
Karnath
,
H.-O.
(
2012
).
Age-specific CT and MRI templates for spatial normalization
.
Neuroimage
,
61
,
957
965
.
Rorden
,
C.
, &
Karnath
,
H.-O.
(
2010
).
A simple measure of neglect severity
.
Neuropsychologia
,
48
,
2758
2763
.
Rorden
,
C.
,
Karnath
,
H.-O.
, &
Bonilha
,
L.
(
2007
).
Improving lesion-symptom mapping
.
Journal of Cognitive Neuroscience
,
19
,
1081
1088
.
van der Linden
,
M.
,
van Turennout
,
M.
, &
Indefrey
,
P.
(
2010
).
Formation of category representations in superior temporal sulcus
.
Journal of Cognitive Neuroscience
,
22
,
1270
1282
.
Warrington
,
E. K.
, &
James
M.
(
1991
).
The visual object and space perception battery
.
Bury St Edmunds, Suffolk, UK
:
Thames Valley Test Company
.
Weintraub
,
S.
, &
Mesulam
,
M. M.
(
1985
).
Mental state assessment of young and elderly adults in behavioral neurology
. In
M. M.
Mesulam
(Ed.),
Principles of behavioral neurology
(pp.
71
123
).
Philadelphia
:
F. A. Davis Company
.
Williams
,
P.
, &
Simons
,
D. J.
(
2000
).
Detecting changes in novel, complex three-dimensional objects
.
Visual Cognition
,
7
,
297
322
.
Wong
,
A. C.-N.
,
Palmeri
,
T. J.
,
Rogers
,
B. P.
,
Gore
,
J. C.
, &
Gauthier
,
I.
(
2009
).
Beyond shape: How you learn about objects affects how they are represented in visual cortex
.
PLoS One
,
4
,
e8405
.