Abstract

Patients with unilateral spatial neglect (USN) often show impaired performance in spatial working memory tasks, apart from the difficulty retrieving “left-sided” spatial data from long-term memory, shown in the “piazza effect” by Bisiach and colleagues. This study's aim was to compare the effect of the spatial position of a visual object on immediate and delayed memory performance in USN patients. Specifically, immediate verbal recall performance, tested using a simultaneous presentation of four visual objects in four quadrants, was compared with memory in a later-provided recognition task, in which objects were individually shown at the screen center. Unlike healthy controls, USN patients showed a left-side disadvantage and a vertical bias in the immediate free recall task (69% vs. 42% recall for right- and left-sided objects, respectively). In the recognition task, the patients correctly recognized half of “old” items, and their correct rejection rate was 95.5%. Importantly, when the analysis focused on previously recalled items (in the immediate task), no statistically significant difference was found in the delayed recognition of objects according to their original quadrant of presentation. Furthermore, USN patients were able to recollect the correct original location of the recognized objects in 60% of the cases, well beyond chance level. This suggests that the memory trace formed in these cases was not only semantic but also contained a visuospatial tag. Finally, successful recognition of objects missed in recall trials points to formation of memory traces for neglected contralesional objects, which may become accessible to retrieval processes in explicit memory.

INTRODUCTION

Patients with unilateral spatial neglect (USN) following right hemisphere damage fail to orient, attend, and respond normally to left-sided visual stimuli, manifestations reflecting a strong bias of focused attention to the ipsilesional (right) side. Disrupted and redundant pattern of visual search in these patients led to the observation that they suffer also from a spatial working memory (WM) deficit, which is shown on both sides of space. Thus, USN patients are likely to forget they had already searched right-sided locations and tend to return there, leading to recursive search on the right and aggravation of left-sided neglect (Mannan et al., 2005; Husain et al., 2001; Wojciulik, Husain, Clarke, & Driver, 2001). In another study using a change detection task (Pisella, Berberovic, & Mattingley, 2004), parietal USN patients showed WM deficits in judgments of spatial location change, but not color or shape change. Impaired WM was found in right parietal neglect patients also on a vertical Corsi block task (Malhotra et al., 2005), pointing again to the existence of a nonlateralized WM deficit in USN.

Another type of memory disturbance found in USN is the difficulty shown at times by these patients in retrieval of “left-sided” information relative to an imagined viewing position, first shown in the famous “piazza effect” by Bisiach and Luzzatti (1978). This lateralized deficit in retrieval of data from long-term memory (LTM), called representational neglect or imaginal neglect, was found to improve by a mere shift of the head or gaze direction to the left hemispace (Meador, Loring, Bowers, & Heilman, 1987). By asking USN patients to name freely as many towns of France they can be remember or to name them while imagining an iconic map of France, it was found that the condition in which they imagined the map led to the omission of towns located to the left of the midline, suggesting that geographic information has to be spatialized to be neglected (Rode, Rossetti, Perenin, & Boisson, 2004; see Treccani, Cubelli, Sellaro, Umilta, & Della Sala, 2012, for an objection to this assumption).

A recent study (Bourlon, Oliviero, Wattiez, Pouget, & Bartolomeo, 2011) had healthy participants imagine or look at the map of France and report if a (heard) town was right (east) or left (west) of Paris while eye movements were recorded, unbeknown to participants. The authors found that participants typically make eye movements toward the town's location, both in the imaginal and the perceptual conditions of the task. Thus, it is likely that some of the mechanisms involved in spontaneous oculomotor behavior may be shared in the exploration of visuospatial mental images. Deficits of these common processes participating in the oculomotor exploration of a visual scene might contribute to the formation of representational neglect in brain-damaged patients.

The representational disorder of memory described above does not require the existence of perceptual neglect (Beschin, Cocchini, Della Sala, & Logie, 1997; Guariglia, Padovani, Pantano, & Pizzamiglio, 1993) and may occur not only during the retrieval of familiar visuospatial information stored in remote LTM (like the objects in different areas of the Piazza del Duomo in Milan) but also for visual information that has been perceived recently (Denis, Beschin, Logie, & Della Salla, 2002).

In the face of explicit failure to report on left-sided stimuli, implicit indications of immediate and delayed effects of information neglected in the learning phase are abundant (McGlinchey-Berroth, Milberg, Verfaellie, Alexander, & Kilduff, 1993; Berti & Rizzolatti, 1992; Volpe, LeDoux, & Gazzaniga, 1979). Evidence of implicit processing of neglected contralesional information was found at the neurophysiological level as well. For example, a study combining event-related evoked potentials and fMRI in a patient with USN secondary to a focal right parietal damage (intact visual pathways) found that, in conditions of bilateral simultaneous stimulation that elicited extinction, left faces that were not perceived still activated the right primary visual cortex and the visual association cortex in the inferior regions of the temporal lobe and evoked N1 potentials, with preserved face-specific negative potentials at 170 msec (Sagiv, Vuilleumier, & Swick, 2000). Another study (Deouell, Bentin, & Soroker, 2000) investigated the processing of lateralized auditory stimuli in USN, using the MMN (an ERP signaling the automatic recruitment of processing resources upon the occurrence of deviations from regularity in the acoustic environment). Unlike contralesional pitch and location deviants, which elicited reduced MMN, stimulus duration deviants evoked equal MMN from both sides.

An example of an implicit indication of delayed effects of neglected information in a recognition memory task is the longer RT of a “no” decision for a previously presented but neglected item, compared with a novel item (Forti & Humphreys, 2007). Another implicit way to demonstrate encoding and late effects of contralesional information in USN was used by Vuilleumier, Schwartz, Clarke, Husain, and Driver (2002). In this study, early exposure to objects that were eventually neglected and evoked no explicit delayed recollection facilitated the recognition of similar objects presented later in a fragmented/partial form, compared with novel objects. In just one study we are aware of, the possibility of delayed recognition of formerly neglected information was shown explicitly (Bisiach, Ricci, Silani, Cossa, & Crespi, 1999). In that recognition memory study, using an old/new judgment task, 7 of 14 USN patients correctly recognized part of the items, which were neglected in the learning phase (yet, the hit rate differed from the false alarm rate in only three patients, and the small number of items presented in the learning phase [15] precluded strong statistical inference).

The aim of this study was to assess the effect of the spatial position of simultaneously presented visual objects on immediate recall and delayed recognition memory in USN. To that end, 14 stroke patients with left USN and 11 healthy control participants viewed 60 pictures, each containing four daily objects, one in each quadrant of the computer screen. Participants performed an immediate verbal recall task, which was followed by a delayed recognition task in which the same objects, interleaved with novel items from the same categories, were individually shown at the screen center. Patients were asked to indicate for each object if it had been presented before, and if so, in which quadrant. Our aim was threefold: first, to evaluate if the spatial position of verbally recalled items would still have an effect on later recognition; second, to assess if items that were not recalled in the first task could still be explicitly retrieved by USN patients in the delayed recognition task; third, to compare the recollection accuracies for objects' identities and objects' locations and assess in this way the impact of spatial WM impairment on patients' ability to encode and later retrieve spatial versus semantic attributes of visually presented stimuli.

METHODS

Participants

Fourteen USN stroke patients were recruited for this study. Stroke was ischemic in 10 and hemorrhagic in 4. There were eight women and six men, at an age range of 30–75 years (mean = 61, SD = 12.0) and an educational level of 3–17 years of formal schooling (mean = 11.6, SD = 4.3). All patients were in the chronic stage, with time after stroke onset of 6–70 months (mean = 40, SD = 24.6). In the recruitment process, we reviewed patient files in recent years' registries of two rehabilitation departments in Israel (the Physical Medicine and Rehabilitation Department in Hadassah Hebrew University Medical Center, Jerusalem, and the Department for Neurological Rehabilitation B at the Loewenstein Hospital, Raanana) and identified potential participants on the basis of the following inclusion criteria: first event of ischemic or hemorrhagic stroke with damage confined to the right cerebral hemisphere, USN manifested during the hospitalization period (years 2005–2010), right handedness, no visual field defects, discharged to their homes. Telephone contact with potential participants was made during 2011. Patients were not recruited if they reported sustaining an additional stroke or significant health or functional deterioration since the time of discharge. Eleven age-matched healthy individuals with negative neurological history served as controls. The age of the control participants (mean = 62.4 years, SD = 11.1 years) was not statistically different from that of the patients (t test: t(23) = 0.392, p = .697). The study received institutional review board approval, and all participants provided informed consent.

Testing for Residual USN in the Chronic Stage

Upon admission to the study, all patients underwent a neurological examination to assure that they had no visual field defects and no oculomotor disturbances. The following tests were administered to assess the severity of residual USN in the chronic stage: (1) Line bisection: Participants had to mark the middle of 1-mm-wide horizontal lines of three different lengths (36 mm, 90 mm, 180 mm) each presented separately, eight times in random order, on the middle of an horizontal A4 paper, with the line midpoint aligned to the participant's midsagittal plane. The mean signed displacement (signed [+] for right-sided deviation and [−] for left-sided deviation) of the subjective midpoint from the true midpoint, in the 180-mm lines, served as a marker of neglect severity in this test; (2) Target cancellation: Two tests were used: the “Star Cancellation” subtest of the “Behavioral Inattention Test” (Wilson, Cockburn, & Halligan, 1987) and the Mesulam–Weintraub Cancellation Test (Mesulam, 2000); (3) Hit rate and RT for target stimuli in the left versus the right side of the screen in a computerized feature-search task: the “Starry Night” (Deouell, Sacher, & Soroker, 2005). All the patients showed signs of left-sided USN in at least two of the above tests (see Table 1 for demographic, clinical, and neglect test results for each patient).

Table 1. 

Demographic and Clinical Data of USN Patients

Patient
Sex
Age
Edu. (years)
TAO (months)
Etiology
UL Motor
LL Motor
Sensory
Report Left Neglect in ADL
Star Cancellation
MWCT
Line Bisection
Visual Search
Deviation from Midline
RT (msec)
Left (/27)
Right (/27)
Left (/30)
Right (/30)
(mm)
SD
Left
Right
HT 67 17 72 2/e 27 25 26 30 43.2 12.0 2051 840 
OZ 72 70 0/e 25 22 22 21 21.9 5.7 1340 1158 
NA 58 16 55 27 26 28 30 0.3 4.1 1033 641 
EW 64 16 50 25 23 28 30 8.1 5.8 1232 829 
MT 52 10 40 2/e 24 17 11 25 1.3 5.6 2019 1055 
SN 66 12 70 22 25 27 29 32.0 19.8 2609 1049 
YD 68 12 69 1/e 22 25 29 20 1.3 1.9 1128 756 
BA 75 14 16 26 27 28 30 1.9 3.9 897 589 
GS 30 14 10 1/e 27 27 NA NA 10.9 7.5 1231 682 
PI 71 18 2/e 24 26 24 21 −2.6 4.7 1087 970 
RP 69 10 42 0/e 22 23 NA NA 4.8 7.8 1927 1953 
RC 62 10 16 1/e 20 24 NA NA 32.9 4.6 1884 885 
RS 52 16 26 2/e 25 27 28 30 1.2 3.3 1279 1085 
MA 50 1/e 20 24 22 26 4.6 4.3 1923 1160 
Patient
Sex
Age
Edu. (years)
TAO (months)
Etiology
UL Motor
LL Motor
Sensory
Report Left Neglect in ADL
Star Cancellation
MWCT
Line Bisection
Visual Search
Deviation from Midline
RT (msec)
Left (/27)
Right (/27)
Left (/30)
Right (/30)
(mm)
SD
Left
Right
HT 67 17 72 2/e 27 25 26 30 43.2 12.0 2051 840 
OZ 72 70 0/e 25 22 22 21 21.9 5.7 1340 1158 
NA 58 16 55 27 26 28 30 0.3 4.1 1033 641 
EW 64 16 50 25 23 28 30 8.1 5.8 1232 829 
MT 52 10 40 2/e 24 17 11 25 1.3 5.6 2019 1055 
SN 66 12 70 22 25 27 29 32.0 19.8 2609 1049 
YD 68 12 69 1/e 22 25 29 20 1.3 1.9 1128 756 
BA 75 14 16 26 27 28 30 1.9 3.9 897 589 
GS 30 14 10 1/e 27 27 NA NA 10.9 7.5 1231 682 
PI 71 18 2/e 24 26 24 21 −2.6 4.7 1087 970 
RP 69 10 42 0/e 22 23 NA NA 4.8 7.8 1927 1953 
RC 62 10 16 1/e 20 24 NA NA 32.9 4.6 1884 885 
RS 52 16 26 2/e 25 27 28 30 1.2 3.3 1279 1085 
MA 50 1/e 20 24 22 26 4.6 4.3 1923 1160 

Abbreviations: F = female; M = male; Edu = education; TAO = time after onset of stroke; Etiology: I = ischemic infarction, H = intracerebral hemorrhage; UL = upper limb; LL = lower limb; Motor Impairment: 0/1/2/3 = no impairment/mild impairment without functional significance/moderate impairment/severe, respectively; Sensory impairment in left limbs: e = tactile extinction; Report neglect in activities of daily living: 2 = very much, 1 = a bit, 0 = no; MWCT = Mesulam–Weintraub Cancellation Test; NA = not administered; Line Bisection: deviation of the subjective midline from the true midline, rightward (+), leftward (−); Visual Search: tested using the computerized “starry night” test (see Methods section), mean RT in msec for left and right stimuli.

Apparatus

Participants were seated at a 50-cm viewing distance from a 15-in. laptop screen (Sony Vaio VPCEA290X) with an active size of 23 × 17.2 cm, a resolution of 1366 × 768 pixels, 32-bit color depth, and a refresh rate of 60 Hz. A program written with the Experiment-Builder software (SR Research, Ontario, Canada) presented visual stimuli first for the recall task, then for the recognition task. All responses were verbal and recorded with a Sandisk Sansa Clip MP3 player.

Immediate Verbal Recall Task

In the first task, four color pictures of daily, semantically unrelated objects were shown simultaneously next to the edges of the four quadrants of a white “card” form (11 × 11 cm, equivalent to 12.6° × 12.6° of visual angle) having its center aligned with the middle of the computer screen. The objects were realistic pictures from several categories—animals, food, plants, furniture, garments, vehicles, tools, musical instruments, and toys. The pictures were symmetrical or almost symmetrical on the vertical axis, making object recognition possible on the basis of the right-sided half. A total number of 60 cards (240 objects) were presented in random order. The task sequence was as follows: First, a central fixation point appeared for 2 sec, followed by a four-object card displayed for 3 sec. Then a central fixation point was displayed for 2 sec ending with the appearance of a question mark. Participants had 8 sec for verbal immediate free recall of the objects in the card they had just seen. Then, another central fixation was displayed for 2 sec, followed by a new four-object card, and so forth. Participants were instructed to look at the screen and try to remember all four objects displayed, because they will be asked to verbally recall the items as soon as a question mark appears. A short training session, consisting of five cards was administered to familiarize the participants with the task.

Delayed Recognition Task

The recognition task started 8 min after the end of the recall task. Each object was presented centrally for 4 sec and then replaced by a 6-sec blank screen. Participants were asked to report if each item had been presented before, and if so, in which quadrant it had appeared. Participants were not informed before the recall task that a recognition task would ensue, neither that they would be asked about the spatial position of the items. A total of 96 objects were sequentially presented in a random order. Half of them were “new” objects, and half were previously presented objects, selected randomly of 48 cards, 12 objects from each quadrant, each from a different card. The recorded verbal responses were analyzed and hit, miss, false alarm, and correct rejection responses were used to compute discrimination scores (d′) and bias estimates (beta ratios). Statistical analyses were done using SPSS software 16.0 (SPSS, Inc., Chicago, IL). The memory tasks are depicted schematically in Figure 1.

Figure 1. 

Timeline of the memory tasks performed.

Figure 1. 

Timeline of the memory tasks performed.

Lesion Analysis

Follow-up CT scans of the patients were analyzed using the “Analysis of Brain Lesions” (ABLe) module implemented in MEDx software (Medical-Numerics, Sterling, VA). ABLe characterizes brain lesions in MRI/CT scans of the adult human brain by spatially normalizing the lesioned brain into Talairach space using the Montreal Neurological Institute template brain. It reports anatomical structures in the normalized brain by using an interface to the Talairach Daemon (San Antonio, TX) or the Automated Anatomical Labeling (AAL) atlas (Tzourio-Mazoyer et al., 2002) and quantifies the amount of lesioned tissue in those standard structures (Mah, Arnold, & Grafman, 2004). Lesions were manually outlined on the digitized CTs using the MEDx software. Registration accuracy of the scans to the Montreal Neurological Institute template was satisfactory (range = 92–96%; mean = 94.2%, SD = 0.91%). Using the output of the ABLe module, multiple Pearson's correlations were performed across participants, correlating the percentage of damage in AAL determined brain structures (AAL labels) and patients' recognition performance (d′ scores). Only structures with damage to more than 5% of the volume in at least four USN patients were included in the analysis. We report all areas for which lesion–behavior correlations were significant up to p < .1, to ensure detection of areas possibly implicated in recognition performance, even at a lenient criterion. In two patients (MA and MT), normalization procedure failed because of technical problems. The normalized lesions of 12 patients are presented in Figure 2 and Table 2.

Figure 2. 

Normalized lesions (white) of USN patients represented on 11 standard Damasio axial slices. The right hemisphere is represented on the left side.

Figure 2. 

Normalized lesions (white) of USN patients represented on 11 standard Damasio axial slices. The right hemisphere is represented on the left side.

Table 2. 

AAL Atlas Structures in Which Lesion Extent Showed a Significant Negative Correlation with Recognition Scores

AAL Label
Pearson's r
p
Negative Correlation with Discrimination Accuracy (d′) 
Angular gyrus −.84 .03 
Precentral gyrus −.87 .02 
Middle frontal gyrus −.82 .04 
Inferior frontal gyrus pars triangularis −.78 .06 
 
Negative Correlation with Bias (Beta Ratio) 
Hippocampus −.99 .0002 
Parahippocampal gyrus −.99 .073 
Amygdala −.77 .044 
Middle temporal gyrus −.67 .069 
AAL Label
Pearson's r
p
Negative Correlation with Discrimination Accuracy (d′) 
Angular gyrus −.84 .03 
Precentral gyrus −.87 .02 
Middle frontal gyrus −.82 .04 
Inferior frontal gyrus pars triangularis −.78 .06 
 
Negative Correlation with Bias (Beta Ratio) 
Hippocampus −.99 .0002 
Parahippocampal gyrus −.99 .073 
Amygdala −.77 .044 
Middle temporal gyrus −.67 .069 

RESULTS

Immediate Free Recall Task

Control participants performed significantly better than USN patients, with a mean recall rate of 86.5% compared with 55.4% for the USN patients. A two-way ANOVA with within-subject factor Quadrant Position and between-subject factor Neglect found significant effects for quadrant position, F(3, 69) = 21.9, p = .000, Neglect, F(1, 23) = 943.6, p = .000, and the interaction term Neglect × Quadrant Position, F(3, 693) = 15.4, p = .000. Control participants performed almost uniformly across the four quadrants of the screen, with a mean recall rate of 86.5% and no statistically significant differences between quadrants (repeated-measures one-way ANOVA with within-subject factor Quadrant Position, F(3, 30) = 0.881, p = .462. In contrast, USN patients correctly recalled on average 76%, 61%, 56%, and 28% of items on right-upper quadrant (RUQ), right-lower quadrant (RLQ), left-upper quadrant (LUQ), and left-lower quadrant (LLQ), respectively, showing both a right-side advantage and a vertical bias. Among USN patients, the mean percentage of recall was 42% for left-sided items and 69% for right-sided items. The medians were 40% and 72%, respectively. The worse performance was 33% for left-sided items and 34% for right-sided items, whereas the best performance was 74% for left-sided items and 88% for right-sided items. A repeated-measures one-way ANOVA with within-subject factor Quadrant Position revealed a significant effect of Quadrant Position in the USN patients group, sphericity assumed, F(3, 39) = 27.46, p < .0001. Post hoc pairwise comparisons using Bonferroni correction found significant differences between LLQ and LUQ (p = .001), LLQ and RLQ (p < .001), LLQ and RUQ (p < .001), and between LUQ and RUQ (p = .001), whereas the differences between LUQ and RUQ, and RLQ and RUQ were not significant. Participants' performance is presented in Figure 3.

Figure 3. 

Immediate free recall as a function of the screen quadrant in which an item was presented (averaged across subjects ± 1 SEM) in the control and the USN groups. A right-side advantage as well as a vertical bias is seen in USN patients.

Figure 3. 

Immediate free recall as a function of the screen quadrant in which an item was presented (averaged across subjects ± 1 SEM) in the control and the USN groups. A right-side advantage as well as a vertical bias is seen in USN patients.

Delayed Recognition Task

Recognition as a Function of Items' Spatial Location in the Learning Phase

Control participants correctly recognized on average 83.7 ± 9.5% of previously presented pictures, compared with a mean of 50.3 ± 14.9% among the USN patients (χ2(1, n = 1201) = 145.73, p = .000). Discrimination scores (d′) were 2.91 ± 0.48 in the control group and 1.79 ± 0.58 in the USN group. Beta values (bias estimates) were 4.31 ± 3.38 in the control group and 5.68 ± 3.30 in the USN group, pointing to an adoption of a strict criterion in both groups (indeed, none of the participants in either group exhibited a tendency to produce false alarms in catch trials [beta ratio < 1.0] and the groups' mean correct rejection rate was high: 96.4 ± 3.24% for controls and 95.5 ± 3.74% for USN patients). In both groups, there was a statistically significant interaction between spatial position and recognition (Control group: χ2(3, n = 528) = 9.46, p = .024; USN group: χ2(3, n = 672) = 71.43, p = .000), with a higher recognition rate for items previously presented in the RUQ than in the LLQ. Grouping the data to the left and right halves of the screen also yield statistically significant differences between the two sides, in both groups (Control group: χ2(1, n = 528) = 9.39, p = .002; USN group: χ2(1, n = 672) = 32.99, p = .000). Grouping the data to the upper and lower halves of the screen yielded a statistically significant difference only in the neglect group (Control group: χ2(1, n = 528) = 0.039, p = .844; USN group: χ2(1, n = 672) = 27.09, p = .000). Figure 4A shows the rate of recognition as a function of the quadrants in which pictures were presented in the recall task.

Figure 4. 

(A) Left: Correct recognition probability (hit rates) according to items' original positions. Significant quadrant effects are seen in both groups. Right: High correct rejection rate of new items in both groups. Error bars denote SEMs. (B) Correct recognition probability of previously recalled items according to their original quadrant positions. Quadrant effects are not significant anymore in the USN group, whereas a slight nonsignificant advantage for correct recognition of right-sided items is seen in the control group. Error bars denote SEMs. (C) Left: Correct recognition probability of not-recalled items according to their original quadrant position in the recall task. In the control group, more than half of the items were correctly recognized despite the fact that they were not recalled immediately after presentation. Quadrant effect was not significant in this group. In the USN group, a significant spatial position effect is noted, although a substantial proportion of items are still correctly recognized on the left. Right: False alarm rates for both groups. Note that this rate is much lower in both groups than the correct recognition of the not-recalled items (on the left). Error bars denote SEMs.

Figure 4. 

(A) Left: Correct recognition probability (hit rates) according to items' original positions. Significant quadrant effects are seen in both groups. Right: High correct rejection rate of new items in both groups. Error bars denote SEMs. (B) Correct recognition probability of previously recalled items according to their original quadrant positions. Quadrant effects are not significant anymore in the USN group, whereas a slight nonsignificant advantage for correct recognition of right-sided items is seen in the control group. Error bars denote SEMs. (C) Left: Correct recognition probability of not-recalled items according to their original quadrant position in the recall task. In the control group, more than half of the items were correctly recognized despite the fact that they were not recalled immediately after presentation. Quadrant effect was not significant in this group. In the USN group, a significant spatial position effect is noted, although a substantial proportion of items are still correctly recognized on the left. Right: False alarm rates for both groups. Note that this rate is much lower in both groups than the correct recognition of the not-recalled items (on the left). Error bars denote SEMs.

Delayed Recognition of Items That Were Verbally Reported in the Immediate Recall Task

As shown earlier, the average success rate in the immediate recall task was 86.5% in healthy controls (with no significant differences between quadrants) and 76%, 61%, 56%, and 28% in RUQ, RLQ, LUQ, and LLQ, respectively, in the USN patients. When the analysis of delayed recognition was confined only to items that were reported verbally in the immediate recall task, average recognition rate (i.e., “hits”) was increased to a mean of 87.2 ± 8.9% for controls and 72.1 ± 16.7% for USN patients, and all quadrant effects became not significant, not only in the control group but also in the USN group (Control group: χ2(3, n = 460) = 6.00, p = .111; Neglect patients: χ2(3, n = 358) = 1.63, p = .652). Grouping the data to the left and right halves of the screen showed no laterality effect in the USN group (χ2(1, n = 358) = 0.17, p = .679), whereas in the control group the recognition rate was slightly higher for right-sided items than for left-sided items (χ2(1, n = 460) = 5.66, p = .017). Grouping the data to the upper and lower halves of the screen yielded no statistically significant difference in either group (Control group: χ2(1, n = 460) = 0.35, p = .555; USN group: χ2(1, n = 358) = 0.20, p = .652). The results of this analysis are presented in Figure 4B.

Delayed Recognition of Items That Were Not Reported in the Immediate Recall Task

Analysis of delayed recognition accuracy confined to items that have not been reported verbally in the immediate recall task revealed a higher rate of recognition in the control group (60.3 ± 28.8%) compared with the USN group (25.6 ± 14.5%), with no significant differences between quadrants in the control group and a significant performance superiority for right-sided quadrants in the USN group (Control group χ2(3, n = 68) = 2.24, p = .523; USN group χ2(3, n = 314) = 19.93, p = .000). Grouping the data to the left and right halves of the screen showed no laterality effect in the control group (χ2(1, n = 68) = 1.4, p = .236), whereas in the USN group the recognition rate was significantly higher for right-sided compared with left-sided items (χ2(1, n = 314) = 11.75, p = .001). Grouping the data to the upper and lower halves of the screen showed no verticality effect for the control group (χ2(1, n = 68) = 0.911, p = .340), whereas in the USN group recognition performance was better for upper items compared with lower items (χ2(1, n = 314) = 7.46, p = .006). The results of this analysis are presented in Figure 4C.

Importantly, the rate of explicit delayed recognition of left-sided items that were not verbally reported in the immediate recall task was found in both groups to be significantly higher than the false alarm rate on Wilcoxon signed-rank tests (Control group: mean recognition 54.8 ± 35.3%, false alarm 3.6 ± 3.2%, Z = −2.669, p = .008; USN group: mean recognition 19.4 ± 11.2%, false alarm 4.5 ± 3.7%, Z = −2.983, p = .003).

Delayed Recognition in Relation to the Order of Verbal Report in the Immediate Recall Task

In both groups, no association was found between the order (1-2-3-4) of verbal report on items in the immediate recall task and the accuracy of their later recognition. No association was found between the order of recall and the accuracy of later recognition when applying a different designation of the order of recall (first-middle-last, where “middle” would be the second item (#2) in trials in which three items were recalled and #2 or #3 in trials with four items recalled; Controls: χ2(2, n = 460) = 0.08, p = .961; USN group: χ2(2, n = 358) = 0.58, p = .750).

Recollection of the Spatial Location of Items

Participants reported in which quadrant each object they recognized as “old” had first appeared (i.e., in the immediate recall task). The control participants recollected the correct location in 70.1 ± 11.8% of recognized items, compared with 59.7 ± 9.8% in the USN group (note that the USN patients' performance is still much higher than the chance level of 25%). Among the USN patients, the accuracy of spatial location recollection was higher for items they reported in the immediate recall task, as compared with items they failed to report on in the immediate recall task yet were correctly recognized by them as “old” (63.3% and 47.5%, respectively; χ2(1, n = 338) = 5.3, p = .021). In the control participants, no statistically significant differences were found between these two types of recognized items. In the USN group, the accuracy of recall of initial position varied with the quadrant position of an item. (Controls: χ2(3, n = 442) = 4.44, p = .217; Neglect patients: χ2(3, n = 338) = 11.06, p = .011). In USN patients, recollection of the spatial location of items was more accurate for items presented originally on the right than on the left (χ2(1, n = 338) = 4.08, p = .043). A tendency for increased accuracy for right-sided items was noted also in the control group, although it did not reach statistical significance (χ2(1, n = 442) = 3.42, p = .064). Grouping upper and lower items yielded no significant association between the vertical position of an item and the accuracy of position recall (Control group: χ2(1, n = 442) = 0.691, p = .406; USN group: χ2(1, n = 338) = 2.312, p = .128; see Figure 5).

Figure 5. 

Position recall accuracy of all recognized items according to their original quadrant position. USN patients correctly positioned more than half of the items to their initial quadrant, much above the 25% chance level indicated by the dotted line. Significant quadrant effects are seen in the USN group but not in control group. Error bars denote SEMs.

Figure 5. 

Position recall accuracy of all recognized items according to their original quadrant position. USN patients correctly positioned more than half of the items to their initial quadrant, much above the 25% chance level indicated by the dotted line. Significant quadrant effects are seen in the USN group but not in control group. Error bars denote SEMs.

The accuracy of objects' recognition did not correlate with the accuracy of spatial location recollection: Pearson's r(11) = 0.31 in controls, p = .357, and r(14) = 0.11 in neglect patients, p = .702. There was a tendency to report that objects had been presented on the right hemifield: The USN group reported on 38.8% of the items that they have been presented on the left, whereas the control group assigned the left side to 46.8% of the items. The most frequent mislocalization error among USN patients was erroneously positioning a left-sided item on the right. This occurred significantly more often than the inverse error (in 34.6% of recognized left-sided items vs. 23.2% of recognized right-sided items, χ2(1, n = 279) = 4.23, p = .040). No such difference was found in the control group (χ2(1, n = 389) = 0.38, p = .53). There was no significant difference between left- and right-sided items as for positioning errors on the vertical axis. Relative numbers of correct position answers and mislocalization errors are presented in Table 3.

Table 3. 

True Position and Reported Position of Correctly Recognized Items, Mean Percent Correct Values across Participants

graphic
 
graphic
 

The diagonal marked in purple designs accurate position recollection of an item to its initial quadrant, whereas the other cells indicate various mislocalization errors. Position answers are reported only for correctly recognized items, so that the total in each quadrant is the mean object recognition rate. The correct position recollection rate can be calculated by dividing the values in the diagonal by the values in the bottom row.

Lesion Analysis

As could be expected, most of the parenchymal damage shown in brain imaging was confined to the territory of the right middle cerebral artery (the graphic presentation of normalized lesion data is shown separately for each patient in Figure 2, using 11 representative axial CT slices, in accordance with the method of Damasio & Damasio, 1989). Total lesion volume ranged from 26 to 290 cm3 (mean = 92 cm3, SD = 78 cm3). As can be seen in Figure 2, lesion size and location differ markedly between patients. The following brain structures were affected in at least half of the patients: the insula (mean lesion extent, 47% of the structure's volume), the putamen (44%), the globus pallidus (32%), the transverse temporal gyrus of Heschl (53%), the superior temporal gyrus (46%), the temporal pole (38%), and the middle temporal gyrus (32%). The supramarginal gyrus was affected in seven patients, and the mean lesion extent was 38%. The angular gyrus was affected in six patients, and the mean lesion extent was 33%. The hippocampus and the parahippocampal gyrus were damaged only in three patients (EW, SN, and RS). We assessed the correlation between lesion extent in each region of the AAL atlas (expressed as the percentage of the region that was involved by the stroke) and recognition performance. Table 2 presents the structures in which damage was negatively correlated with d′ or beta ratio at a level of significance of up to p < .1.

We also assessed the correlation between lesion extent in each region and correct recognition (hit rate) performance, and the same areas were found to be correlated as in the analysis with d′. As can be seen in Table 2, lesion extent in regions of the frontal and the inferior parietal cortex affected mainly the discrimination accuracy, whereas damage to regions of the temporal lobe had an effect mainly on the response bias.

DISCUSSION

The aim of this study was to assess the effect of the spatial position of visual objects on various aspects of immediate and delayed memory in USN. Right hemisphere–damaged patients with chronic signs of left-sided neglect were compared with healthy controls. The study was conducted in two parts: first, free recall of objects was tested immediately following presentation of pictures of objects; second, delayed recognition was tested and analyzed in relation to the performance in the first part. The comparative analysis of patients' memory in the immediate and delayed phases shed new light on the phenomenon of lateralized memory loss in USN.

Immediate Recall

Uneven capacity for immediate recall of environmental information was anticipated in the patient group, in view of the chronicity of USN signs revealed in all patients in formal neglect tests (Table 1). Indeed, the immediate recall task showed that when the visual scene contains several unrelated objects, patients' capacity to recall immediately the identity of objects on the left (especially lower left) is significantly inferior compared with their recall capacity for right-sided objects. Inability reporting correctly on left-sided information is considered an essential component of the USN syndrome as defined by Heilman, Valenstein, and Watson (1985): “a failure to report, respond, or orient to novel or meaningful stimuli presented on the side opposite a brain lesion, which cannot be attributed to either elemental sensory or motor deficits.” This component of the syndrome could emerge however from a variety of reasons: impaired perceptual processing on the left (i.e., difficulty creating clear, recognizable percepts from left-sided visual stimuli); impaired maintenance of correctly formed percepts in visual WM; or a state of physiological disconnection, where the access of correctly formed percepts to verbal declarative systems on the left hemisphere is unstable. The possibility for multiple underlying mechanisms is in accord with current USN theorizing, where the syndrome is conceived as a manifestation of disrupted activity in a distributed neural network mediating spatial attention (there are however several network models, assigning somewhat different roles to different cortical and subcortical structures: Corbetta, Kincade, Lewis, Snyder, & Sapir, 2005; Posner & Petersen, 1990; Mesulam, 1981). It is not unlikely that damage in one part of the right hemisphere will induce the left-sided recall inferiority in one way whereas damage to another component of the network will result in left-sided inferiority via another mechanism. The performance of the USN patients in the immediate free recall task revealed a spatial bias not only in the horizontal but also in the vertical dimension, with significant superiority in recall of items shown in the RUQ compared with items shown in the LLQ. This finding is consistent with a growing body of evidence pointing to a vertical component in spatial neglect (Cazzoli, Nyfeller, Hess, & Müri, 2011; Halligan & Marshall, 1989). The anatomical substrate underlying the vertical bias in USN awaits clarification.

Note that in the current study we compared memory performance of USN patients and age-matched healthy controls, without including a group of RBD patients without USN. Given that USN has no gold standard for its diagnosis (moreover, double dissociation has been shown for its different diagnostic paradigms), we applied demonstration of left-sided neglect in at least two of the four diagnostic tests used, as an inclusion criterion (see Methods). This means that patients could perform beyond the cutoff for normality in one or two of the tests and yet be included in the research group. It also means that RBD patients who failed to answer our operational definition of USN (the above inclusion criterion) were still likely to have some degree of USN, differently manifested or of lesser severity. Thus, instead of comparing two patient groups separated on the basis of a questionable diagnostic definition, we decided to compare different aspects of delayed recognition for neglected versus nonneglected items in the same participants. This analysis is based on the assumption that neglect is not an all-or-none phenomenon but a statistical issue, that is, an increment (beyond normal) of the likelihood to miss left-sided objects and events. Indeed, almost all USN patients in the chronic stage detect a given portion of left-sided stimuli quickly and accurately; thus, we believe it is legitimate to compare delayed recognition for manifested versus not-manifested occasions of lateralized inattention, on a trial-by-trial basis.

Delayed Recognition

The delayed recognition task revealed several important findings. First, a spatial bias (right-side advantage) was found not only in the patient group (who showed a similar bias in the immediate recall task) but also in the performance of the healthy controls (who did not show a spatial bias in the immediate recall task). The concordant bias shown by the USN participants in the two memory tasks could be anticipated: If the objects presented on the left received less or no attention in the learning phase and as a consequence failed to form coherent percepts and/or stable memory traces, late retrieval failure for these items was bound to occur.

The right-side bias shown by the healthy control participants is more difficult to explain, as these participants performed equally well on the left and right sides in the immediate recall task. This unexpected finding is in line with a recent report by Emrich, Burianová, and Ferber (2011), who found in normal participants a right-side advantage in recognition performance, in conditions of high visual WM load (simultaneous presentation of three objects, different in shape and color). The authors referred to this phenomenon as to a “transient perceptual neglect” and showed by fMRI that recognition performance was associated with deactivation of the right TPJ and activation of regions associated with visual WM, around the intraparietal sulcus (IPS; Emrich et al., 2011). Another study (Sheremata, Bettencourt, & Somers, 2010) demonstrated, using a change detection visual WM task, robust bilateral field representations in right hemisphere IPS and, in contrast, strongly biased contralateral field representations in left hemisphere IPS. This asymmetry only emerged when participants had to hold more than one object in STM (in the current study, the number of simultaneously presented items, four, also corresponds to a high WM load). If however asymmetric hemispheral involvement in conditions of increased WM load was responsible for the right-side advantage (as right-sided objects recruit both hemispheres whereas left-sided objects recruit mainly the right hemisphere in these conditions), one would expect the bias to be revealed in immediate recall as well as in delayed recognition. Indeed, ceiling effect in the former task (around 86.5% correct recall in all quadrants) could mask small side differences, yet the fact that lateral preference was significant only in delayed recognition might have another explanation.

Given the lack of significant side differences in the immediate recall task, the lateral preference demonstrated by healthy participants in delayed recognition may have its origin in a later stage in memory formation. We propose that the relative advantage of right-sided items in delayed recognition is related to the dominant role of the contralateral left hemisphere in verbal memory formation (i.e., in preservation of the name attribute of the presented object beyond the time immediately following the learning phase; D'Esposito et al., 1998; Smith & Jonides, 1998). This conjecture is supported by a second finding in the current study—the demonstration that successful verbal report on a shown object immediately following its presentation not only raises the likelihood of correct recognition at a later stage but also eliminates to a large extent (more notably in the patients group) the effect of the object's spatial position on delayed recognition.

A plausible explanation for this attenuation of the spatial bias, which was more salient in the USN group, is that restriction of the analysis of delayed recognition performance to objects successfully recalled immediately after presentation removes from the analysis trials where failed delayed recognition was a consequence of left-sided inattention with impaired object perception and encoding in the learning phase. In fact, such restriction seems to equalize the starting point of USN participants and healthy controls with respect to the initial stages of memory formation by removing from the analysis items affected by perceptual neglect. However, even in this restricted analysis, USN participants still lagged behind controls in overall recognition rate, probably because of nonspatially lateralized factors operating in this symptom complex (Husain & Rorden, 2003). The fact that right-side advantage remained significant in healthy participants and became statistically not significant in the USN group, when the analysis was restricted to items successfully recalled immediately after presentation, is likely to reflect greater variance and reduced statistical power of side differences in the latter group (large variance, both inter- and intrapersonal, associated with performance instability, is an inherent feature of USN; see Mesulam, 2002; Bartolomeo, Siéroff, Chokron, & Decaix, 2001; Anderson, Mennemeier, & Chatterjee, 2000; here the standard deviation of recognition performance across participants with USN was 0.183, compared with 0.088 across control participants).

Thus, we propose a common explanatory mechanism for (1) the overall right-side advantage revealed in delayed recognition memory in the two groups and (2) the attenuation of this side difference, mainly in the USN group, when only items reported immediately after presentation are taken into account. These two findings may reflect a similar pattern of interhemispheric dynamics, where encoding of the verbal (name) attribute of presented objects in left hemisphere synaptic space strengthens the memory trace and facilitates its delayed recognition. When perceptual processing occurs equally well in both sides of space (in the control group), verbal encoding in the left hemisphere is likely to give a small advantage to objects captured from the contralateral right hemispace. This is because the left hemisphere's vector of focused attention is directed to the right hemispace by default (Halligan & Marshall, 1994), facilitating the integrated processing of different features of right-sided objects and the concomitant encoding of these features (the physical appearance and the spatial location attributes of an object, along with its verbal attribute—the name). This facilitatory action is pointed also by the lower rate of positioning errors for right- compared with left-sided objects in this group, which had a marginal statistical significance (p = .064). When impaired and unstable perceptual processing on the left (in the USN group) leads to impaired memory formation, a left-sided object in a trial escaping perceptual neglect and ending in correct naming is in a much smaller relative disadvantage, because its verbal attribute (its name) is encoded in the nondamaged left hemisphere and the overall strength of the memory trace increases (Smith & Jonides, 1998, 1999).

Unlike the current study, which focused on the effect of objects' spatial position on their delayed recognition, almost all the previous studies of representational neglect employed recall tasks, where patients had to describe verbally a familiar square or a learned array, or to draw a figure from memory. The theoretical accounts that emerged from these studies include inability to attend to the left part of the representational space, similar to the inability to attend to the left part of perceptual space; inability to spatially reconstruct left-sided information; inability to address and retrieve left-sided data from correctly represented spatial information (Bisiach, 1993; Meador et al., 1987; Baddeley & Lieberman, 1980; Bisiach & Luzzatti, 1978). Later studies (Logie, Della Sala, Beschin, & Denis, 2005; Della Sala, Logie, Beschin, & Denis, 2004) reported on preserved capacities of visuospatial transformation and mental rotation in patients with representational neglect, possibly pointing to impaired maintenance of left-sided visuospatial information in WM as the essential underlying dysfunction.

The marked attenuation of the spatial bias in delayed recognition performance, when only objects successfully recalled immediately after presentation were entered into the analysis, raises the question why objects on the left part of a well-known location fail to be retrieved (as in Bisiach & Luzzatti's, 1978, famous “piazza effect”). If encoding of the verbal attribute (the name) of left-sided objects in the learning phase is what enabled these objects' escape from “representational neglect” in the delayed recognition task, one should assume that the verbal code is likely to facilitate recognition memory (as tested in the second part of the current study) but is much less effective in facilitating free recall (as tested in the original “piazza” experiment). An essential difference between recognition and free recall with respect to activation of the verbal code (the object's name) can explain the selective facilitatory effect in the former. In a recognition task, the actual sight of an object presented during the testing phase is likely to evoke a silent naming process. The evoked name (verbal code) can be matched then with the name stored during the learning phase, thus reducing the dependence of the retrieval process on other attributes of the memory trace, that is, the spatial location of the object, its shape, color, and other contextual and semantically related factors, which seem to be the primary cues for retrieval in tasks like description of a known place. This mechanism is likely to operate in the “testing effect,” demonstrated in healthy participants, where administration of a free recall task after exposure to a list of items affects delayed recognition performance in a positive manner (Chan & McDermott, 2007; Carrier & Pashler, 1992). The effect seemingly comprises of strengthening the memory trace through the verbal iteration of objects' names, that is, by creating a “dual” (verbal/visual) coding” condition.

Another crucial difference between our recognition test and recall tasks used in previous representational neglect research is the fact that we presented just one object (in central position) for testing at a time. In contrast, trying to retrieve from memory details on the left side of a known place is likely to be hindered in USN by strong attraction coming from objects on the right side of the reconstructed memory. The additional two factors making our recognition test distinct from typical recall tasks used to elicit representational neglect (e.g., describing the surroundings of Milan's square) are the small eccentricity of items presented in the learning phase and the small radial distance (50 cm) of the display where the items were presented (i.e., in peripersonal rather than extrapersonal space). The impact of the imagined radial and lateral distance on the ability to retrieve left-sided data from memory needs to be further clarified. However, there are reported cases pointing to a distinction between representation and memory for data in peripersonal space versus data in more distant segments of extrapersonal space (De Nigris et al., 2013; Palermo, Piccardi, Nori, Giusberti, & Guariglia, 2010). Guariglia and Piccardi (2010) proposed that a distinction should be made between two types of representational neglect—one revealed in recollection of environmental information (town squares, paths, rooms of a house, etc.) involving navigation and the other in spatial information that is not gained through navigation (objects on a desktop, the interior of a car, etc). There is evidence for a double dissociation between these two types of representational neglect (e.g., Ortigue et al., 2003; Grossi, Modafferi, Pelosi, & Trojano, 1989).

Memory of Spatial Information

It is of interest that USN patients could state the correct spatial location (the quadrant on the screen where the object was initially presented), in no less than 60% of the objects they recognized in the second task (much above chance level, 25%). Among the correctly localized objects, 34% were on the left, thus showing that objects may escape representational neglect even when a “left” spatial tag is maintained in relation to their memory trace. It should be noted however that correct recollection of the spatial location was significantly higher for objects originally presented on the right side of the screen and erroneous misplacement from left to right was frequent. Thus, if the spatial tag maintained in memory acted in a facilitatory manner during delayed recognition, it did so more often when on the right. In the case of left-sided objects, it is possible that in a certain proportion of cases erroneous misplacement to the right improved recognition performance in a way similar to the improvement demonstrated for such displacement errors on auditory perception in USN (Deouell & Soroker, 2000; Soroker, Calamaro, & Myslobodsky, 1995). The patients' relatively accurate retrieval of the spatial location of recognized items was quite unexpected, considering previous evidence of impaired spatial WM in USN (Malhotra et al., 2005; Pisella et al., 2004). Although spatial WM was not tested in the current study, it could be expected that such impairment will affect delayed recall of the spatial location, but the patients recalled the locations of recognized items quite well (60% accuracy, compared with 70% in healthy controls). Note that participants were not instructed to remember the objects' spatial positions and were not aware that a recognition task would follow the verbal recall task. Thus, the learning of objects' positions was incidental (without intention) in nature. A study comparing free recall of items and recall of the items' spatial location, conducted in stroke patients with right and left brain damage (Vakil, Soroker, & Biran, 1992), found that the recall of the spatial location by RBD patients was better under incidental compared with intentional learning conditions. The relative preservation of location information in the current study compared with the documented impairment in spatial WM may relate to the fact that typical spatial WM tasks involve intentional rather than incidental learning of the spatial position of objects. In both healthy controls and USN patients, object recognition performance and the accuracy of spatial position recollection did not correlate, possibly reflecting incongruent involvement of ventral and dorsal pathways mediating recognition of objects' identities and objects' spatial locations, respectively (Takahashi, Ohki, & Kim, 2013).

Explicit Recollection of Neglected Information

In the delayed recognition task, USN patients recognized 19.4% of the left-sided items they failed to recall immediately after presentation. This rate was far beyond their rate of false alarm responses for distracter stimuli (4.5%); therefore, it can be considered a manifestation of explicit recollection of neglected information. This finding corroborates data reported originally by Bisiach et al. (1999) on the basis of a smaller sample of items, proving that inability of USN patients to verbally report on a presented object immediately after presentation does not necessarily mark lack of perception and encoding. Note that upon naming the recalled objects immediately after presentation, the patients often made a remark like “I think there was something there as well (pointing to the left), but I can't remember what,” reflecting lack of immediate recall for detected (not necessarily recognized) contralateral items in such trials. USN patients (but not healthy control participants) revealed here a spatial bias, with right-side advantage in delayed recognition of objects that have not been verbally recalled immediately after presentation. This finding reflects the superiority of encoding/retrieval processes for ipsilesional compared with contralesional information in USN, a bias shown to be attenuated considerably in recognition of objects the patients were able to name immediately after presentation, as discussed above.

Anatomical Considerations

The brain lesions of the patients who participated in the study varied largely in size and location, but in most cases (9 of 12 patients for whom normalized lesion data are available) the hippocampus and the parahippocampal gyrus were not involved, making it unlikely that direct damage to these structures was a major factor in the memory impairments shown by the USN participants as a group. Yet, the beta ratio (denoting the criterion along the liberal/conservative range, used by the patients' in their old/new judgments) maintained a significant negative correlation with the extent of damage only in the hippocampus, parahippocampal gyrus, amygdala, and middle temporal gyri—regions classically associated to the formation and storage of LTM traces. The negative correlation here implies attenuation of the conservative criterion (very small rate of false alarms) that characterized the patients' attitude in general.

In contrast, d′ values correlated negatively with the extent of damage to a classical neglect region (Table 2)—the angular gyrus of the inferior parietal cortex as well as frontal regions (the precentral, middle frontal, and inferior frontal gyri). This finding raises the question of the role of parietal damage in formation of the memory impairments comprising representational neglect and the role of the posterior parietal cortex (PPC) in recognition memory in general. The PPC role in memory is under debate in recent years, as activations of this area during the retrieval phase of recognition memory tasks have been reported repeatedly, both in ERP and fMRI studies (Wagner, Shannon, Kahn, & Buckner, 2005; Rugg, 1995, 2004), whereas virtual lesions in this region by rTMS failed to induce clear memory impairments (Rossi et al., 2006). A lesion study in stroke patients (Haramati, Soroker, Dudai, & Levy, 2008) revealed impaired recognition of objects' pictures and sounds (but not words) following right hemisphere damage including the PPC. Yet, lesion–behavior correlation analyses (region-based correlations and voxel-based lesion-symptom mapping) suggested a nonparietal origin of the impairments in recognition memory, specifically implying damage to frontal and lateral temporal areas. In the current study, a correlation between recognition performance and lesion extent in frontal areas (middle and inferior frontal gyri, precentral gyrus) was again found. To our knowledge, no published study has systemically investigated the anatomical structures underlying representational neglect. A case report by Rode et al. (2010) raised the possibility that callosal damage may play a role by creating a dissociative state between LTM and spatial processing of information. Yet, another study reported two cases of pure visual neglect without a representational component, following posterior cerebral artery infarctions affecting the splenium of the corpus callosum (Tomaiulo et al., 2010). Two other case reports reported a right thalamic lesion causing representational neglect (Ortigue et al., 2001; Beschin, Basso, & Della Sala, 2000). Thus, it seems that damage to a variety of structures can underlie the manifestations of representational neglect by disruption of neural circuits essential to the representation of spatial relationships, the encoding and maintenance in LTM of these relationships, and their ordered retrieval upon demand.

In conclusion, in this study we sought to investigate the effect of the spatial location of objects on various aspects of immediate and delayed visual memory in USN. As expected, we found a laterality effect in immediate free recall, as well as a vertical bias, with the worse recall performance being for items located in the LLQ of the screen. In the delayed recognition task, we found a right-side advantage in the USN group, as well as in healthy participants. Limiting the analysis to items that have been verbally recalled immediately after presentation eliminated to a large extent this spatial bias, pointing to difference between classical (“piazza”-like) recall tasks used to demonstrate contralesional memory loss in representational neglect research, and the current recognition task. This finding points to a positive role of objects' verbal attributes (objects names) as a factor strengthening the memory traces and facilitating delayed recognition. Recollection of the original spatial location of items was much above chance level, suggesting the retention of visuospatial information about the items and the ability of a certain proportion of objects, correctly tagged as left sided, to reach conscious awareness. Explicit recollection of neglected information was also shown, both for objects' identities and locations. Lesion analysis suggests that discriminability (d′) and bias (beta) in memory retrieval are affected by damage to different regions, a finding demanding further investigation. Overall, the findings point to a substantial amount of contralesional information being perceived and encoded in USN.

Acknowledgments

This study was supported by a Physician-Scientist grant from Hadassah Hebrew University Medical Center.

Reprint requests should be sent to Elior Moreh, Department of Physical Medicine and Rehabilitation, Hadassah Hebrew University Hospital, Mount Scopus, P.O.B. 24035, Jerusalem 92140, Israel, or via e-mail: elior@hadassah.org.il.

REFERENCES

REFERENCES
Anderson
,
B.
,
Mennemeier
,
M.
, &
Chatterjee
,
A.
(
2000
).
Variability not ability: Another basis for performance decrements in neglect.
Neuropsychologia
,
38
,
785
796
.
Baddeley
,
A. D.
, &
Lieberman
,
K.
(
1980
).
Spatial working memory.
In R. S. Nickerson (Ed.)
,
Attention and performance VIII
(pp.
521
539
).
Hillsdale, NJ
:
Erlbaum
.
Bartolomeo
,
P.
,
Siéroff
,
E.
,
Chokron
,
S.
, &
Decaix
,
C.
(
2001
).
Variability of response times as a marker of diverted attention.
Neuropsychologia
,
39
,
358
363
.
Berti
,
A.
, &
Rizzolatti
,
G.
(
1992
).
Visual processing without awareness: Evidence from unilateral neglect.
Journal of Cognitive Neuroscience
,
4
,
345
351
.
Beschin
,
N.
,
Basso
,
A.
, &
Della Sala
,
S.
(
2000
).
Perceiving left and imagining right: Dissociation in neglect.
Cortex
,
36
,
401
414
.
Beschin
,
N.
,
Cocchini
,
G.
,
Della Sala
,
S.
, &
Logie
,
R. H.
(
1997
).
What the eyes perceive, the brain ignores: A case of pure unilateral representational neglect.
Cortex
,
33
,
3
26
.
Bisiach
,
E.
(
1993
).
Mental representation in unilateral neglect and related disorders.
Quarterly Journal of Experimental Psychology
,
46
,
435
461
.
Bisiach
,
E.
, &
Luzzatti
,
C.
(
1978
).
Unilateral neglect of representational space.
Cortex
,
14
,
129
133
.
Bisiach
,
E.
,
Ricci
,
R.
,
Silani
,
G.
,
Cossa
,
F. M.
, &
Crespi
,
M. G.
(
1999
).
Hyperamnesia in unilateral neglect.
Cortex
,
35
,
701
711
.
Bourlon
,
C.
,
Oliviero
,
B.
,
Wattiez
,
N.
,
Pouget
,
P.
, &
Bartolomeo
,
P.
(
2011
).
Visual mental imagery: What the head's eye tells the mind's eye.
Brain Research
,
1367
,
287
297
.
Carrier
,
M.
, &
Pashler
,
H.
(
1992
).
The influence of retrieval on retention.
Memory & Cognition
,
20
,
633
642
.
Cazzoli
,
D.
,
Nyfeller
,
T.
,
Hess
,
C. W.
, &
Müri
,
R. M.
(
2011
).
Vertical bias in neglect: A question of time?
Neuropsychologia
,
49
,
2369
2374
.
Chan
,
J. C. K.
, &
McDermott
,
K. B.
(
2007
).
The testing effect in recognition memory: A dual process account.
Journal of Experimental Psychology: Learning, Memory, and Cognition
,
33
,
431
437
.
Corbetta
,
M.
,
Kincade
,
M. J.
,
Lewis
,
C.
,
Snyder
,
A. Z.
, &
Sapir
,
A.
(
2005
).
Neural basis and recovery of spatial attentional deficits in spatial neglect.
Nature Neuroscience
,
8
,
1424
1425
.
Damasio
,
H.
, &
Damasio
,
A. R.
(
1989
).
Lesion analysis in neuropsychology.
New York
:
Oxford University Press
.
De Nigris
,
A.
,
Piccardi
,
L.
,
Bianchini
,
F.
,
Palermo
,
L.
,
Incoccia
,
C.
, &
Guariglia
,
C.
(
2013
).
Role of visuo-spatial working memory in path integration disorders in neglect.
Cortex
,
49
,
920
930
.
Della Sala
,
S.
,
Logie
,
R. H.
,
Beschin
,
N.
, &
Denis
,
M.
(
2004
).
Preserved visuo-spatial transformations in representational neglect.
Neuropsychologia
,
42
,
1358
1364
.
Denis
,
M.
,
Beschin
,
N.
,
Logie
,
R. H.
, &
Della Salla
,
S.
(
2002
).
Visual perception and verbal descriptions as sources for generating mental representations: Evidence from representational neglect.
Cognitive Neuropsychology
,
19
,
97
112
.
Deouell
,
L. Y.
,
Bentin
,
S.
, &
Soroker
,
N.
(
2000
).
Electrophysiological evidence for an early (preattentive) information processing deficit in patients with right hemisphere damage and unilateral neglect.
Brain
,
123
,
353
365
.
Deouell
,
L. Y.
,
Sacher
,
Y.
, &
Soroker
,
N.
(
2005
).
Assessment of spatial attention after brain damage with a dynamic reaction time test.
Journal of the International Neuropsychological Society
,
11
,
697
707
.
Deouell
,
L. Y.
, &
Soroker
,
N.
(
2000
).
What is extinguished in auditory extinction?
NeuroReport
,
11
,
3059
3062
.
D'Esposito
,
M.
,
Aguirre
,
G. K.
,
Zarahn
,
E.
,
Ballard
,
D.
,
Shin
,
R. K.
, &
Lease
,
J.
(
1998
).
Functional MRI studies of spatial and nonspatial working memory.
Brain Research, Cognitive Brain Research
,
7
,
1
13
.
Emrich
,
S. M.
,
Burianová
,
H.
, &
Ferber
,
S.
(
2011
).
Transient perceptual neglect: Visual working memory load affects conscious object processing.
Journal of Cognitive Neuroscience
,
10
,
2968
2982
.
Forti
,
S.
, &
Humphreys
,
G. W.
(
2007
).
The representation of unseen objects in visual neglect: Effects of view and object identity.
Cognitive Neuropsychology
,
24
,
661
680
.
Grossi
,
D.
,
Modafferi
,
A.
,
Pelosi
,
L.
, &
Trojano
,
L.
(
1989
).
On the different roles of the cerebral hemispheres in mental imagery: The “o'Clock Test” in two clinical cases.
Brain Cognition
,
10
,
18
27
.
Guariglia
,
C.
,
Padovani
,
A.
,
Pantano
,
P.
, &
Pizzamiglio
,
L.
(
1993
).
Unilateral neglect restricted to visual imagery.
Nature
,
364
,
235
237
.
Guariglia
,
C.
, &
Piccardi
,
L.
(
2010
).
Environmental orientation and navigation in different types of unilateral neglect.
Experimental Brain Research
,
206
,
163
169
.
Halligan
,
P. W.
, &
Marshall
,
J. C.
(
1989
).
Is neglect (only) lateral? A quadrant analysis of line cancellation.
Journal of Clinical and Experimental Neuropsychology
,
11
,
793
798
.
Halligan
,
P. W.
, &
Marshall
,
J. C.
(
1994
).
Toward a principled explanation of unilateral neglect.
Cognitive Neuropsychology
,
11
,
167
206
.
Haramati
,
S.
,
Soroker
,
N.
,
Dudai
,
Y.
, &
Levy
,
D. A.
(
2008
).
The posterior parietal cortex in recognition memory: A neuropsychological study.
Neuropsychologia
,
46
,
1756
1766
.
Heilman
,
K. M.
,
Valenstein
,
E.
, &
Watson
,
R. T.
(
1985
).
The neglect syndrome.
In J. A. M. Fredricks (Ed.)
,
Handbook of clinical neurology, 1. (45): Clinical neuropsychology
(pp.
153
183
).
Amsterdam
:
Elsevier
.
Husain
,
M.
,
Mannan
,
S.
,
Hodgson
,
T.
,
Wojciulik
,
E.
,
Driver
,
J.
, &
Kennard
,
C.
(
2001
).
Impaired spatial working memory across saccades contributes to abnormal search in parietal neglect.
Brain
,
124
,
941
952
.
Husain
,
M.
, &
Rorden
,
C.
(
2003
).
Non-spatially lateralized mechanisms in hemispatial neglect.
Nature Reviews Neuroscience
,
4
,
26
36
.
Logie
,
R. H.
,
Della Sala
,
S.
,
Beschin
,
N.
, &
Denis
,
M.
(
2005
).
Dissociating mental transformations and visuo-spatial storage in working memory: Evidence from representational neglect.
Memory
,
13
,
430
434
.
Mah
,
L.
,
Arnold
,
M. C.
, &
Grafman
,
J.
(
2004
).
Impairment of social perception associated with lesions of the prefrontal cortex.
American Journal of Psychiatry
,
161
,
1247
1255
.
Malhotra
,
P.
,
Jäger
,
H. R.
,
Parton
,
A.
,
Greenwood
,
R.
,
Playford
,
E. D.
,
Brown
,
M. M.
,
et al
(
2005
).
Spatial working memory capacity in unilateral neglect.
Brain
,
128
,
424
435
.
Mannan
,
S. K.
,
Mort
,
D. J.
,
Hodgson
,
T. L.
,
Driver
,
J.
,
Kennard
,
C.
, &
Husain
,
M.
(
2005
).
Revisiting previously searched locations in visual neglect: Role of right parietal and frontal lesions in misjudging old locations as new.
Journal of Cognitive Neuroscience
,
17
,
340
354
.
McGlinchey-Berroth
,
R.
,
Milberg
,
W. P.
,
Verfaellie
,
M.
,
Alexander
,
M.
, &
Kilduff
,
P. T.
(
1993
).
Semantic processing in the neglected visual field: Evidence from a lexical decision task.
Cognitive Neuropsychology
,
10
,
79
108
.
Meador
,
K. J.
,
Loring
,
D. W.
,
Bowers
,
D.
, &
Heilman
,
K. M.
(
1987
).
Remote memory and neglect syndrome.
Neurology
,
37
,
522
526
.
Mesulam
,
M. M.
(
1981
).
A cortical network for directed attention and unilateral neglect.
Annals of Neurology
,
10
,
309
325
.
Mesulam
,
M. M.
(
2000
).
Principles of behavioral and cognitive neurology
(2nd ed.).
Oxford, UK
:
Oxford University Press
.
Mesulam
,
M. M.
(
2002
).
Functional anatomy of attention and neglect: From neurons to networks.
In
The cognitive and neural bases of spatial neglect
(pp.
33
46
).
New York
:
Oxford University Press
.
Ortigue
,
S.
,
Viaud-Delmon
,
I.
,
Annoni
,
J. M.
,
Landis
,
T.
,
Michel
,
C. M.
,
Blanke
,
O.
,
et al
(
2001
).
Pure representational neglect after right thalamic lesion.
Annals of Neurology
,
50
,
401
404
.
Ortigue
,
S.
,
Viaud-Delmon
,
I.
,
Michel
,
C. M.
,
Blanke
,
O.
,
Annoni
,
J. M.
,
Pegna
,
A.
,
et al
(
2003
).
Pure imagery hemi-neglect of far space.
Neurology
,
60
,
2000
2002
.
Palermo
,
L.
,
Piccardi
,
L.
,
Nori
,
R.
,
Giusberti
,
F.
, &
Guariglia
,
C.
(
2010
).
Does hemineglect affect visual mental imagery? Imagery deficits in representational and perceptual neglect.
Cognitive Neuropsychology
,
27
,
115
133
.
Pisella
,
L.
,
Berberovic
,
N.
, &
Mattingley
,
J. B.
(
2004
).
Impaired working memory for location but not for colour or shape in visual neglect: A comparison of parietal and non-parietal lesions.
Cortex
,
40
,
379
390
.
Posner
,
M. I.
, &
Petersen
,
S. E.
(
1990
).
The attention system of the human brain.
Annual Review of Neuroscience
,
13
,
25
42
.
Rode
,
G.
,
Cotton
,
F.
,
Revol
,
P.
,
Jacquin-Courtois
,
S.
,
Rossetti
,
Y.
, &
Bartolomeo
,
P.
(
2010
).
Representation and disconnection in imaginal neglect.
Neuropsychologia
,
48
,
2903
2911
.
Rode
,
G.
,
Rossetti
,
Y.
,
Perenin
,
M. T.
, &
Boisson
,
D.
(
2004
).
Geographic information has to be spatialized to be neglected: A representational neglect case.
Cortex
,
40
,
391
397
.
Rossi
,
S.
,
Pasqualetti
,
P.
,
Zito
,
G.
,
Vecchio
,
F.
,
Cappa
,
S.
,
Miniussi
,
C.
,
et al
(
2006
).
Prefrontal and parietal cortex in human episodic memory: An interference study by repetitive transcranial magnetic stimulation.
European Journal of Neuroscience
,
23
,
793
800
.
Rugg
,
M. D.
(
1995
).
ERP studies of memory.
In M. D. Rugg & M. G. H. Coles (Eds.)
,
Electrophysiology of mind: Event-related brain potentials and cognition
(pp.
113
127
).
Oxford
:
Oxford University Press
.
Rugg
,
M. D.
(
2004
).
Retrieval processing in human memory: Electrophysiological and fMRI evidence.
In M. S. Gazzaniga (Ed.)
,
The cognitive neurosciences III
(pp.
727
738
).
Cambridge, MA
:
MIT Press
.
Sagiv
,
N.
,
Vuilleumier
,
P.
, &
Swick
,
D.
(
2000
).
The neural fate of extinguished faces: Electrophysiological correlates of conscious and unconscious perception in unilateral spatial neglect. Cognitive Neuroscience Society Meeting, San Francisco.
Journal of Cognitive Neuroscience
,
12(Suppl.)
,
95
.
Sheremata
,
S. L.
,
Bettencourt
,
K. C.
, &
Somers
,
D. C.
(
2010
).
Hemispheric asymmetry in visuotopic posterior parietal cortex emerges with visual short-term memory load.
Journal of Neuroscience
,
30
,
12581
12588
.
Smith
,
E. E.
, &
Jonides
,
J.
(
1998
).
Neuroimaging analyses of human working memory.
Proceedings of the National Academy of Sciences,
, U.S.A.,
95
,
12061
12068
.
Smith
,
E. E.
, &
Jonides
,
J.
(
1999
).
Storage and executive processes in the frontal lobes.
Science
,
283
,
1657
1661
.
Soroker
,
N.
,
Calamaro
,
N.
, &
Myslobodsky
,
M. S.
(
1995
).
Ventriloquist effect reinstates responsiveness to auditory stimuli in the “ignored” space in patients with hemispatial neglect.
Journal of Clinical and Experimental Neuropsychology
,
17
,
243
255
.
Takahashi
,
E.
,
Ohki
,
K.
, &
Kim
,
D. S.
(
2013
).
Dissociation and convergence of the dorsal and ventral visual working memory streams in the human prefrontal cortex.
Neuroimage
,
65
,
488
498
.
Tomaiulo
,
F.
,
Voci
,
L.
,
Bresci
,
M.
,
Cozza
,
S.
,
Posteraro
,
F.
,
Oliva
,
M.
,
et al
(
2010
).
Selective visual neglect in right brain damaged patients with splenial interhemispheric disconnection.
Experimental Brain Research
,
206
,
209
217
.
Treccani
,
B.
,
Cubelli
,
R.
,
Sellaro
,
R.
,
Umilta
,
C.
, &
Della Sala
,
S.
(
2012
).
Dissociation between awareness and spatial coding: Evidence from unilateral neglect.
Journal of Cognitive Neuroscience
,
24
,
854
867
.
Tzourio-Mazoyer
,
N.
,
Landeau
,
B.
,
Papathanassiou
,
D.
,
Crivello
,
F.
,
Etard
,
O.
,
Delcroix
,
N.
,
et al
(
2002
).
Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain.
Neuroimage
,
15
,
273
289
.
Vakil
,
E.
,
Soroker
,
N.
, &
Biran
,
N.
(
1992
).
Differential effect of right and left hemispheric lesions on two memory tasks: Free recall of items and recall of spatial location.
Neuropsychologia
,
30
,
1041
1051
.
Volpe
,
B. T.
,
LeDoux
,
J. E.
, &
Gazzaniga
,
M. S.
(
1979
).
Information processing in an “extinguished” visual field.
Nature
,
282
,
722
724
.
Vuilleumier
,
P.
,
Schwartz
,
S.
,
Clarke
,
K.
,
Husain
,
M.
, &
Driver
,
J.
(
2002
).
Testing memory for unseen visual stimuli in patients with extinction and spatial neglect.
Journal of Cognitive Neuroscience
,
14
,
873
886
.
Wagner
,
A. D.
,
Shannon
,
B. J.
,
Kahn
,
I.
, &
Buckner
,
R. L.
(
2005
).
Parietal lobe contributions to episodic memory retrieval.
Trends in Cognitive Science
,
9
,
445
453
.
Wilson
,
B.
,
Cockburn
,
J.
, &
Halligan
,
P. W.
(
1987
).
Behavioural Inattention Test.
Bury St. Edmunds, UK
:
Thames Valley Test Company
.
Wojciulik
,
E.
,
Husain
,
M.
,
Clarke
,
K.
, &
Driver
,
J.
(
2001
).
Spatial working memory deficit in unilateral neglect.
Neuropsychologia
,
39
,
390
396
.