- Email: [email protected]
Residual oculomotor and exploratory deficits in patients with recovered hemineglect
Neuropsychologia 42 (2004) 1203–1211
Residual oculomotor and exploratory deficits in patients with recovered hemineglect Tobias Pflugshaupt, Stefanie Almoslöchner Bopp, Dörthe Heinemann, Urs P. Mosimann, Roman von Wartburg, Thomas Nyffeler, Christian W. Hess, René M. Müri∗ Departments of Neurology and Clinical Research, Perception and Eye Movement Laboratory, University of Berne, Inselspital, Freiburgstrasse 10, 3010 Berne, Switzerland Received 19 May 2003; received in revised form 3 February 2004; accepted 4 February 2004
Abstract Several studies on hemineglect have reported that patients recover remarkably well when assessed with neuropsychological screening tests, however, they show deficits on novel or complex tasks. We investigated whether such deficits can be revealed with eye movement analysis, applying two basic oculomotor tasks as well as two exploratory tasks. Eye movements were recorded in eight hemineglect patients at least eleven months after right-hemisphere brain damage had occurred. Sixteen healthy volunteers participated in the control group. Regarding the basic oculomotor tasks, only the overlap task revealed residual deficits in patients, suggesting that a directional deficit in disengaging attention persisted during recovery. Further residual deficits were evident in the exploratory tasks. When everyday scenes were explored, patients showed a bias in early orienting towards the ipsilateral hemispace. In a search task, they demonstrated the same orienting bias as well as a non-directional deficit concerning search times. Moreover, patients preferentially fixated in the contralateral hemispace, but did not benefit from this asymmetry in terms of search times, i.e. they did not detect contralateral targets faster than ipsilateral ones. This suggests a dissociation between oculomotor processes and attentional ones. In conclusion, we have identified behavioural aspects that seem to recover slower than others. A disengagement deficit and biases in early orienting have been the most pronounced residual oculomotor deficits. © 2004 Elsevier Ltd. All rights reserved. Keywords: Human; Attention; Eye movements; Recovery of function
1. Introduction Acute hemineglect is characterised by a failure to report, respond, or orient to stimuli presented in the space contralateral to the lesioned hemisphere (Halligan & Marshall, 1994). This tendency towards the ipsilateral hemispace—i.e. the ipsilateral bias—can affect perceptual, motor, or representational processes (Kirk, 2001). Clinical experiences show that most hemineglect symptoms diminish in approximately 75% of patients within 6 months (Stone, Patel, Greenwood, & Halligan, 1992). However, this percentage is related to overt symptoms, such as test performance in copying or line bisection tasks. Several studies have revealed that other aspects of behaviour benefit less from recovery (Campbell & Oxbury, 1976; Goodale, Milner, Jakobson, & Carey, 1990; Harvey & Milner, 1999; Jehkonen et al., 2000; Karnath, 1988). These ∗ Corresponding author. Tel.: +41-316323081/368; fax: +41-316329679. E-mail address: [email protected] (R.M. Müri).
0028-3932/$ – see front matter © 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropsychologia.2004.02.002
findings suggest that residual hemineglect deficits tend to occur in novel or complex rather than simple or familiar settings. Karnath (1988) postulated that the speed of recovery from hemineglect was deficit-specific. According to his theory, acute hemineglect is characterised by three main components: (a) an early orienting of attention towards the ipsilateral side; (b) an impairment in disengaging attention from the ipsilateral side and reorienting towards contralateral stimuli; and (c) a non-directional reduction in information processing capacity. Karnath proposed that the second component recovered faster than the other two. Recent evidence (Mattingley, Bradshaw, Bradshaw, & Nettleton, 1994; Olk, Harvey, & Gilchrist, 2002) supports his notion with regard to the bias in early orienting. This deficit seems to be particularly persistent during recovery. Eye movement analysis provides an accurate assessment of hemineglect (Sprenger, Kömpf, & Heide, 2002) and its measurement is facilitated due to the limited number of relevant variables (i.e. fixations and saccades). In recent years,
1204
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211
many studies have used eye movement data to investigate particular aspects of acute hemineglect, such as saccadic eye movement programming (Walker & Findlay, 1996), visual space exploration (Karnath & Niemeier, 2002; Karnath, Niemeier, & Dichgans, 1998), or visual search behaviour (Husain et al., 2001; Sprenger et al., 2002). The aim of the present eye movement study was to reveal residual oculomotor and exploratory deficits in hemineglect patients who show substantial recovery with regard to clinical reports and in neuropsychological screening tests. First, two basic oculomotor tasks (i.e. gap and overlap task as described by Saslow (1967)) were administered to investigate speed and accuracy of visually guided saccades. Acute hemineglect patients have demonstrated delayed and hypometric saccades into the contralateral hemifield during these tasks (Heide & Kömpf, 1998; Walker & Findlay, 1996). We were interested in whether these deficits persisted during recovery. Second, exploratory behaviour was tested with a free exploration task and a search task. Since the centre of exploration is shifted towards the ipsilateral side in acute hemineglect patients (Karnath et al., 1998), we expected that such a shift would be less pronounced in residual hemineglect. The results of the exploratory tasks were further analysed to investigate early orienting. In accordance with a recent case report that focussed on residual hemineglect deficits in search behaviour (Olk et al., 2002), we hypothesised that our patient group would show a bias towards the ipsilateral side regarding early orienting. The present eye movement study on residual deficits in hemineglect differs from previous ones on acute hemineglect in two essential respects. First, we chose everyday photographs in order to represent real life situations. Previous studies used random arrays of letters (Husain et al., 2001; Karnath & Niemeier, 2002; Karnath et al., 1998) or geometrical shapes (Sprenger et al., 2002). Second, each of our patients was confronted with a large number of images (i.e. 32) per exploratory task. In previous studies (Husain et al., 2001; Karnath & Niemeier, 2002; Karnath et al., 1998), results were based on <10 images per exploratory task. By increasing the number of images per task, the results produced may depend less on image content and thus reflect systematic exploratory strategies more accurately.
ure and shape copying subtest of the behavioural inattention test (Wilson, Cockburn, & Halligan, 1987) as well as a line bisection task, which was similar to the one Schenkenberg, Bradford, & Ajax (1980) used. During hospitalisation (T1), Patients had to demonstrate pronounced asymmetries in one or several of these screening tests to enter the study. Since the manual of the behavioural inattention test does not provide side-specific cut-off scores, we adopted the scoring procedure of the Bells test (Gauthier, Dehaut, & Joanette, 1989) for the star and letter cancellation. A relevant asymmetry was thus characterised by a left-minus-right difference of more than two omissions (Rousseaux et al., 2001). For the figure and shape copying subtest, any left sided omission was considered as a relevant asymmetry (Azouvi et al., 2002). In line with the scoring procedure proposed by a clinical study (Ferber & Karnath, 2001), we applied a 14% cut-off criterion (i.e. mean percentage rightward deviation of the bisection mark in relation to the correct length) for the line bisection task. Furthermore, confrontational visual field testing at the end of hospitalisation verified that none of the patients suffered from visual field deficits. At the time of the experiment (T2), the four neuropsychological screening tests and a clinical interview were applied to assure that patients could be identified as a group of recovered hemineglect patients in terms of clinically relevant asymmetries. Eight patients met these conditions and participated in the study. Demographic and lesion data of the patient group is summarised in Table 1, clinical reports and neuropsychological test results are given in Table 2, and Fig. 1 depicts two illustrative examples of the recovery observed in the screening tests. Sixteen healthy volunteers participated in the control group. Patients and controls did not differ with regard to age (patients: median = 47, range = 32; control subjects: median = 47, range = 34), gender, and handedness (Mann–Whitney and χ2 -tests: n.s.). None of them suffered from depression (Beck Depression Inventory/BDI, Hautzinger, Bailer, Wollall, & Keller, 1994), colour blindness (Ishihara colour plates, Ishihara, 1999), visuo-motor
2. Methods 2.1. Patients and control subjects The hemineglect patients who participated in the study were former inpatients at our division of neuropsychological rehabilitation. All patients showed a history of acute hemineglect with typical clinical signs (e.g. bumping into things placed in the left half of space) after right-hemisphere damage. The hemineglect diagnosis was based on a wide range of neurological and neuropsychological examinations, including the star cancellation, letter cancellation, and the fig-
Fig. 1. Figure copying examples of patient 4 (top) and patient 8 (bottom). Original stimulus (left), copying performance during hospitalisation (T1; middle) and at the time of the experiment (T2; right).
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211
1205
Table 1 Demographic and lesion data of the patients (f: frontal, t: temporal, p: parietal, th: thalamic)
P1 P2 P3 P4 P5 P6 P7 P8
Sex
Age
Aetiology
Localisation of lesion
Time since onset of brain damage in months
Male Male Female Female Female Male Male Male
39 29 54 54 43 43 61 51
Ischaemic stroke Ischaemic stroke Ischaemic stroke Ischaemic stroke Ischaemic stroke Haematoma Haematoma Haematoma
f, t f th f, p f, t p, t f, t f, t
17 35 15 60 19 11 16 70
Mean
47
30
S.D.
10
23
Table 2 Clinical reports and neuropsychological test results of the patient group Clinical reportsa
P1 P2 P3 P4 P5 P6 P7 P8
Cancellation tasks
Line bisection
Figure copying
T1
T2
T1
T2
T1
T2
T1
T2
X X X X X X X X
– – – – – – – –
X X X X X X X X
– – – – – – – –
– X – X – X – X
– – – – – – – –
– – – X X – – X
– – – – – – – –
Summary of clinically relevant asymmetries (X) during hospitalisation (T1) and at the time of the experiment (T2). a Refers to clinical observations (e.g. bumping into things placed in the left half of space) during hospitalisation (T1), respectively to the clinical interview at the time of the experiment (T2).
disturbances (e.g. ocular motor paralysis, strabism) or limited visual acuity (normal or corrected to normal). Both patients as well as control subjects gave written informed consent prior to participation. The study was approved by the local ethical committee. 2.2. Apparatus Eye movements were recorded using an infrared-based video tracking system (EyeLink, SensoMotoric Instruments TM, GmbH, Berlin, Germany). This system collects eye movement data at a sampling rate of 250 Hz with a spatial resolution of 0.01◦ . It provides an accuracy of gaze-position relative to stimulus coordinates of 0.5◦ –1.0◦ , depending largely on participants’ fixation accuracy during calibration. A chin rest was used to both ensure constant distance and minimise head movements. Moreover, head motion compensation was computed by the tracking system. Minimal fixation duration was set to 120 ms and saccades had to exceed 0.75◦ in order to be analysed. Participants were seated 70 cm in front of a 19 in. screen, subtending a visual angle of approximately 29◦ × 22◦ . Eye movements of either the right or the left eye were registered, with both eyes equally often recorded across all participants. The system was cal-
ibrated prior to each block of stimuli by means of a 3 × 3 point grid. 2.3. Eye movement recording 2.3.1. Oculomotor tasks Two types of basic oculomotor tasks were applied and both of them required participants to perform visually guided reflexive saccades. The time course of stimulus presentation in these two tasks is shown in Fig. 2. During the gap task, the central fixation (CF) point was presented pseudo-randomly with durations ranging from
Fig. 2. Gap and overlap task. Time course of stimulus presentation.
1206
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211
2000 to 3000 ms. The offset of the CF preceded the onset of the lateral target (LT) by 200 ms (i.e. temporal gap). Target amplitude (range 3.7◦ –9.1◦ ) and direction were randomised in the horizontal plane. The LT was presented for 1000 ms. Following the offset of the LT, the next CF was presented immediately. Participants were instructed to fixate the point that was presented, regardless of whether it appeared centrally (CF) or laterally (LT). During the overlap task, stimulus presentation was identical to the gap task except for the CF that remained visible during the presentation of the LT. The participants therefore had to actively disengage attention from the CF before looking at the LT. This contrasts the gap task, in which disengagement is facilitated by the temporal gap. The overlap task required participants to look at the LT as soon as it appeared and then to refixate the CF as soon as the LT disappeared. Overall, 28 targets (i.e. 14 ipsilateral and 14 contralateral targets) were shown per participant and task. The off-line analysis contained a comparison of ipsi—versus contralateral values with regard to the following parameters: latency of the first saccade, gain (i.e. saccade amplitude/target amplitude) of the first and second saccade, and the gap effect, which corresponds to the difference between overlap and gap latency. It is a well described finding that a temporal gap between fixation point offset and target onset typically results in shorter saccadic latencies than if the fixation point remains visible (Pratt, Bekkering, Abrams, & Adam, 1999). Two trial exclusion criteria were applied for both oculomotor tasks. First, anticipatory saccades were excluded since
our focus was on reflexive saccades. Saccades with negative latencies (i.e. saccade initiation occurred prior to the onset of the LT) or with latencies below 80 ms were therefore not analysed. Second, saccades with a starting point that deviated more than one degree from the CF were excluded as well, since these saccades would strongly interfere with the validity of gain calculations. 2.3.2. Exploratory tasks Exploratory behaviour was investigated under two experimental conditions. First, a free exploration task that consisted of everyday scenes was applied. Each scene was presented for 7 s. Second, patients and control subjects had to solve a search task, during which they were instructed to find and click a predefined object (i.e. leaf, screw, clock, roll, snake, butterfly, pear, or scissors) placed in an everyday scene as quickly as possible. Search times longer than 7 s were treated as omissions. Object size was 1.8◦ × 1.5◦ on average and the overall distribution of objects was balanced across quadrants. Fig. 3 depicts two example stimuli for each exploratory task. Each participant was confronted with 32 images per exploratory task in total. Blocks of 8 images were formed to recalibrate the system between blocks. In the search task, the nature of the object (i.e. leaf, screw, etc.) was varied on a block-to-block interval. All images were 29◦ × 22◦ in size and information content was balanced to prevent an image-induced ipsi—or contralateral bias. Prior to every image, a central fixation point was presented to assure that
Fig. 3. Free exploration of everyday scenes (left) and search task (right). Two examples; superimposed white arrows show targets (clock, snake) in the search task examples.
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211
1207
Table 3 Basic oculomotor tasks Left
Right
AQ
patients
controls
patients
controls
patients
controls
P
Gap
Latency Gain1 Gain2
158 (64) 0.85 (0.50) 1.00 (0.10)
157 (84) 0.91 (0.30) 1.02 (0.15)
183 (126) 0.85 (0.54) 0.96 (0.16)
162 (70) 0.91 (0.29) 0.99 (0.10)
0.89 (0.51) 1.01 (0.74) 0.94 (0.17)
0.96 (0.50) 0.99 (0.48) 0.96 (0.19)
0.153 0.653 0.350
Overlap
Latency Gain1 Gain2
307 (265) 0.87 (0.70) 0.97 (0.12) 158 (273)
240 (150) 0.95 (0.40) 1.00 (0.15) 82 (159)
222 (155) 0.93 (0.59) 0.98 (0.22) 36 (137)
219 (187) 0.93 (0.34) 0.99 (0.14) 60 (142)
1.29 1.07 1.01 0.30
1.10 1.00 0.99 0.62
0.007∗∗ 0.153 0.417 0.023∗
Gap effect
(0.82) (1.30) (0.27) (0.92)
(0.41) (0.30) (0.23) (1.35)
Latency and gap effect in ms; gain of the first (gain1) and second (gain2) saccade as quotients. Values are medians; ranges in parentheses; AQ: asymmetry quotient; P-values (Mann–Whitney tests) refer to group comparisons of AQs. ∗ Significant at the 0.05 level. ∗∗ Significant at the 0.01 level.
patients and control subjects started exploring the images in a central location. The off-line analysis included plots of the horizontal distribution of fixation frequency. These plots illustrate the centre of exploration in a descriptive manner. Furthermore, parameters such as the number of fixations, the fixation duration, and the saccadic amplitude were compared between groups. The main focus was on the comparison of ipsi—with contralateral values regarding first saccade parameters (i.e. direction, latency and amplitude) and overall fixation patterns (i.e. number of fixations and total fixation duration). Since exploratory behaviour was likely to change during stimulus presentation, total fixation duration was further analysed applying a second-by-second design. Together with first saccade parameters, this analysis allowed an investigation of early orienting in exploratory behaviour. In the search task, two additional parameters were examined: the percentage of hits and the average search time defined as the duration from stimulus onset to the mouseclick. 2.4. Experimental procedure The order of experimental conditions was held constant across participants. Basic oculomotor tasks were followed by exploratory tasks. During eye movement recordings, participants were seated in a dimly lit room. All tasks were preceded by verbal instructions and practise trials. 2.5. Data analysis For most parameters of oculomotor and exploratory behaviour, asymmetry quotients (AQs) based on individual medians were calculated (i.e. ipsilateral/contralateral value1 ). However, asymmetries with regard to latencies and search times refer to inverted ratios (i.e. contralateral/ipsilateral 1 Although the expressions ‘ipsilateral’ and ‘contralateral’ relate to lesion sites and are therefore inadequate with regard to healthy control subjects, we nonetheless use these expressions for both patients and control subjects (i.e. ‘ipsilateral’ = right, ‘contralateral’ = left), for the purpose of a consistent terminology across groups.
value). This inversion was necessary to illustrate asymmetries in a consistent manner. As a consequence, asymmetries towards the contralateral side are represented by AQs < 1, AQs > 1 indicate asymmetries towards the ipsilateral side and an AQ = 1 describes exact symmetry. Since AQs were not normally distributed (Kolmogorov–Smirnov test), group comparisons were based on non-parametric tests (Mann–Whitney or χ2 -tests). Similarly, medians and ranges will be presented as central tendency and dispersion measures. A P-value of <0.05 was considered statistically significant and all tests were two-tailed.
3. Results 3.1. Basic oculomotor tasks In general, data loss due to anticipatory saccades or central fixation problems was relatively low in both tasks with group-specific medians below 5%. In the gap task, no statistically significant group differences were found. However, Table 3 shows that the overlap task revealed significant differences. Patients’ latencies were increased when the LT appeared on the contralateral as opposed to the ipsilateral side. Corresponding AQs differed significantly between groups. As a consequence of this deficit, the gap effect of patients was considerably less pronounced on the ipsilateral than on the contralateral side. Again, corresponding AQs differed significantly between groups. Both results indicate a task-specific dissociation. Residual deficits in patients’ basic oculomotor behaviour were apparent in the overlap task only. 3.2. Exploratory tasks 3.2.1. Centre of exploration Plotting the horizontal distribution of fixation frequency allowed the assessment of the centre of exploration in a descriptive manner. In contrast to acute hemineglect patients,
1208
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211 contralateral
ipsilateral
1 2
25
3 20
4 5
15
6 7 .4
10
.6
.8
1.0
1.2
1.4
1.6
AQ 5
Fig. 5. Free exploration of everyday scenes. Time course of AQs with regard to total fixation duration from first to seventh second (y-axis); patients (black) vs. controls (grey); based on medians.
0 25
20
15
10
5
0
Fig. 4. Horizontal distribution of fixation frequency. Free exploration of everyday scenes (top), search task (bottom); eight columns of 3.6◦ each (x-axis), percentage of fixations (y-axis); patients (black) vs. controls (grey).
Fig. 4 illustrates that the group of recovered patients did not show an ipsilaterally shifted centre of exploration. In fact, a slight shift towards the contralateral side was evident in both tasks.
3.2.2. Free exploration of everyday scenes When exploring everyday scenes, patients made fewer fixations in number, longer fixations in duration, as well as smaller saccades in amplitude than controls. None of these differences, however, reached statistical significance. Comparing ipsi- and contralateral values with regard to overall fixation patterns did not reveal significant group differences either. Nevertheless, significant differences were observed when early orienting was analysed, with the first saccade being particularly sensitive to residual deficits. The patient group made more and larger first saccades towards the ipsilateral as opposed to the contralateral hemispace while the control subjects made relatively more first saccades towards the contralateral side. Corresponding AQs differed significantly between groups (Table 4). Based on a second-by-second design, Fig. 5 shows the time course of exploring everyday scenes with regard to AQs in total fixation duration. Statistical analyses revealed significant group differences within the first two seconds. During the first second, patients’ total fixation time was distributed rather symmetrically whereas controls spent more fixation time on the contralateral side (Mann–Whitney: Z = −2.14, P = 0.032). Corresponding AQs, however, changed substantially during the following second. Controls moved
Table 4 Parameters of the first saccade Leftward
Rightward
AQ
patients
controls
patients
controls
patients
controls
P
Free exploration of everyday scenes
Direction Latency Amplitude
45 (41) 333 (143) 3.1 (4.2)
56 (63) 257 (163) 4.4 (3.8)
55 (41) 289 (131) 4.3 (3.2)
44 (63) 286 (179) 4.1 (4.2)
1.21 (2.48) 1.08 (0.55) 1.33 (0.70)
0.78 (2.86) 0.96 (0.66) 0.89 (1.28)
0.038∗ 0.136 0.023∗
Search task
Direction Latency Amplitude
46 (53) 284 (231) 3.4 (4.3)
59 (31) 261 (131) 4.7 (4.7)
54 (53) 291 (140) 4.1 (4.5)
41 (31) 241 (153) 4.8 (5.1)
1.17 (3.94) 1.11 (0.75) 1.40 (1.21)
0.69 (0.95) 1.08 (0.61) 1.11 (1.21)
0.019∗ 0.350 0.023∗
Direction in %, latency in ms and amplitude in ◦ . Values are medians, ranges in parentheses; P-values (Mann–Whitney tests) refer to group comparisons of AQs. ∗ Significant at the 0.05 level.
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211
1209
Table 5 Fixation patterns and search performance Left
Right
patients
controls
patients
AQ controls
patients
controls
P
Free exploration of everyday scenes
Number of fixations Total fixation duration
11 (4) 3.10 (1.64)
11 (6) 3.08 (1.07)
9 (6) 2.66 (1.81)
10 (7) 2.94 (1.19)
0.82 (0.81) 0.81 (0.96)
0.93 (0.65) 0.95 (0.65)
0.264 0.383
Search task
Number of fixations Total fixation duration Search time Percentage of hits
4 (7) 2.47 (2.07) 3.01 (3.31) 62.5 (50)
3 (3) 1.33 (1.24) 1.48 (1.94) 75 (50)
4 (5) 2.21 (2.34) 2.98 (2.97) 59.4 (56)
3 (2) 1.41 (1.05) 1.77 (1.87) 75 (50)
1.00 0.91 1.07 1.00
1.00 1.04 0.83 1.00
0.787 0.038∗ 0.038∗ 0.928
(1.00) (0.25) (0.72) (0.43)
(1.20) (0.70) (1.14) (0.27)
Number of fixations, total fixation duration in s, search time in s, and percentage of hits in %. Values are medians, ranges in parentheses; P-values (Mann–Whitney tests) refer to group comparisons of AQs. ∗ Significant at the 0.05 level.
to the ipsilateral (i.e. right) side while patients preferentially explored the contralateral side (Mann–Whitney: Z = −2.45, P = 0.013). Thus, when exploring everyday scenes freely, patients differed from control subjects during the initial phase of exploration only. 3.2.3. Search task In the search task, patients made more fixations in number, shorter fixations in duration, as well as smaller saccades in amplitude than controls. Statistical analyses showed that the difference concerning the number of fixations was significant (Mann–Whitney: Z = −2.29, P = 0.023). However, this significance must be put into perspective, since patients’ search times were significantly longer for both ipsilaterally (Mann–Whitney: Z = −2.388, P = 0.016) and contralaterally presented targets (Mann–Whitney: Z = −3.123, P = 0.001), resulting in a higher number of fixations. In addition, an interesting dissociation occurred in patients’ search behaviour. Despite relatively longer total fixation durations in the contralateral hemispace, patients did not benefit from this asymmetry in terms of search times, i.e. they did not detect contralateral targets faster than ipsilateral ones. Since controls’ total fixation durations were rather symmetrical and search times contralaterally biased, both group comparisons regarding AQs reached statistical significance (Table 5). With regard to the percentage of hits, patients found slightly less targets than controls (Mann–Whitney tests: n.s.), irrespective of whether the target was localised ipsi- or contralaterally. Corresponding AQs, however, did not differ significantly between groups. Due to short search times, a second-by-second analysis of fixation behaviour was not appropriate for the search task. Nevertheless, early orienting could be examined by focussing on first saccade parameters. Results were highly similar to the ones obtained in the free exploration of everyday scenes. Significant group differences were found with regard to direction and amplitude of the first saccade. The patient group made more and larger saccades towards the ipsilateral as opposed to the contralateral hemispace while the control subjects made relatively more first saccades towards the contralateral side (Table 4).
4. Conclusion The results of the present study can be summarised as follows: At least eleven months after the onset of brain damage, eight hemineglect patients showed no clinically relevant asymmetries with regard to a clinical interview, in neuropsychological screening tests, and concerning the centre of exploration in two exploratory tasks. This is in accordance with the assumption that most neglect symptoms diminish within six months (Stone et al., 1992). However, residual deficits occurred in a basic oculomotor task, in early orienting, and with regard to search performance. We thus identified behavioural aspects that seem to recover slower than others. With regard to basic oculomotor behaviour, the overlap task revealed ipsilateral biases in patients whereas the gap task did not. While acute hemineglect patients have shown ipsilateral biases in gap and overlap tasks (Behrmann, Ghiselli-Crippa, & Dimatteo, 2001/2002; Heide & Kömpf, 1998; Karnath, Schenkel, & Fischer, 1991; Walker & Findlay, 1996), our results suggest that the overlap task is particularly sensitive to residual hemineglect deficits. With reference to Posner, Walker, Friedrich, & Rafal (1984, 1987), we assume that this indicates impaired disengagement of attention, when the following saccade is directed towards the contralateral side. The underlying assumption is that the overlap task requires participants to actively disengage attention whereas during the gap task, disengagement is facilitated by the temporal gap (Pratt et al., 1999). This directional disengagement deficit may be a first behavioural aspect that seems to persist during recovery from hemineglect. Patients’ exploratory behaviour was characterised by residual deficits during the initial phase of exploration. When everyday scenes were explored, patients initially oriented towards the ipsilateral side as opposed to control subjects who preferentially oriented towards the contralateral (i.e. left) side. Biases in early orienting of acute hemineglect patients have been reported in many studies. (e.g. Azouvi et al., 2002; Gainotti, D’Emre, & Bartolomeo, 1991; Mattingley et al., 1994; Olk et al., 2002). A recent study based on 206 patients (Azouvi et al., 2002) has provided
1210
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211
evidence that a right-sided starting point in a cancellation task was the most sensitive paper and pencil measure for acute hemineglect. The present study suggests that this bias in early orienting may be a second behavioural aspect that persists during recovery. This is consistent with previous studies on residual hemineglect (Mattingley et al., 1994; Olk et al., 2002) and with Karnath’s model of recovery (1988). Similar to previous findings (Gainotti et al., 1991; Mattingley et al., 1994; Tant et al., 2002), the present study detected early orienting asymmetries in healthy subjects as well. In contrast to the ipsilateral bias found in neglect patients, healthy subjects initially oriented towards the contralateral (i.e. left) hemifield, irrespective of task. It has been suggested that this asymmetry results from dominant right hemisphere activation for spatial processing (Tant et al., 2002). An alternative account is that acquired scanning habits determine healthy subjects’ early orienting. Such habits may be strongly related to language aspects and reading (Nicholls, Bradshaw, & Mattingley, 1999; Sakhuja, Gupta, Singh, & Vaid, 1996). The search task revealed three interesting findings. First and similar to the free exploration of everyday scenes, patients’ early orienting was ipsilaterally biased. Second, search times were significantly longer in patients relative to controls, irrespective of whether the object was presented contra- or ipsilaterally. This was also shown by a previous study (Heide & Kömpf, 1998) and it may correspond to the non-directional reduction in information processing capacity that is seen as a major component in acute and residual hemineglect (Karnath, 1988). Finally, patients preferentially fixated in the contralateral hemifield, but they did not benefit from this asymmetry in terms of search times, i.e. they did not find contralateral targets faster than ipsilateral ones. The latter finding can be interpreted as a dissociation between underlying processes. While oculomotor behaviour in terms of overall fixation patterns seems to have recovered considerably, attentional processes as indicated by search times may recover slower. A similar finding has been described in acute hemineglect (Làdavas, Zeloni, Zaccara, & Gangemi, 1997). When instructed to maintain central fixation while responding manually to peripheral targets, patients made oculomotor responses to the contralateral side without a corresponding shift of attention as documented by the absence of manual responses. This finding was discussed in terms of a partial separation of attentional processes from oculomotor ones. The present study provides evidence that this separation is likely to persist during recovery. Therefore, one may ask whether recovery from acute hemineglect should be regarded as a compensatory mechanism rather than true functional recovery. The dissociation between oculomotor and attentional processes during the search task favours compensatory mechanisms in our patient group. Nevertheless, patients’ normal performances in the gap task or with regard to the centre of exploration in the exploratory tasks indicates functional recovery, at least to some extent.
Acknowledgements This work was supported by grants from the Thomas Stanely Johnson Foundation.
References Azouvi, P., Samuel, C., Louis-Dreyfus, A., Bernati, T., Bartolomeo, P., & Beis, J.-M. et al., (2002). Sensitivity of clinical and behavioural tests of spatial neglect after right hemisphere stroke. Journal of Neurology, Neurosurgery, and Psychiatry, 73, 160–166. Behrmann, M., Ghisseli-Crippa, T., & Dimatteo, I. (2001/2002). Impaired initiation but not execution of contralesional saccades in hemispatial neglect. Behavioural Neurology, 13, 39–60. Campbell, D., & Oxbury, J. (1976). Recovery from unilateral visuospatial neglect. Cortex, 12, 303–312. 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. Gainotti, G., D’Emre, P., & Bartolomeo, P. (1991). Early orientation of attention toward the half-space ipsilateral to the lesion in patients with unilateral brain damage. Journal of Neurology, Neurosurgery, and Psychiatry, 54, 1082–1089. 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. Goodale, M. A., Milner, A. D., Jakobson, L. S., & Carey, D. P. (1990). Kinematic analysis of limb movements in neuropsychological research: Subtle deficits and recovery of function. Canadian Journal of Psychology, 44, 180–195. Halligan, P. W., & Marshall, J. C. (1994). Towards a principled explanation of unilateral neglect. Cognitive Neuropsychology, 11, 167–206. Harvey, M., & Milner, A. D. (1999). Residual perceptual distortion in ‘recovered’ hemispatial neglect. Neuropsychologia, 37, 745–750. Hautzinger, M., Bailer, M., Wollall, H., Keller, F. (1994). BeckDepressions-Inventar (BDI): Testhandbuch. Bern: Hans Huber. Heide, W., & Kömpf, D. (1998). Combined deficits of saccades and visuospatial orientation after cortical lesions. Experimental Brain Research, 123, 164–171. 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. Ishihara, S. (1999). Ishihara’s test for color deficiency. Tokyo: Kanehara & Co. Jehkonen, M., Ahonen, J.-P., Dastidar, P., Koivisto, A.-M., Laippala, P., & Vilkki, J. et al., (2000). Visual neglect as a predictor of functional outcome one year after stroke. Acta Neurologica Scandinavica, 101, 195–201. Karnath, H.-O. (1988). Deficits of attention in acute and recovered visual neglect. Neuropsychologia, 26, 27–43. Karnath, H.-O., & Niemeier, M. (2002). Task-dependent differences in the exploratory behaviour of patients with spatial neglect. Neuropsychologia, 40, 1577–1585. Karnath, H.-O., Niemeier, M., & Dichgans, J. (1998). Space exploration in neglect. Brain, 121, 2357–2367. Karnath, H.-O., Schenkel, P., & Fischer, B. (1991). Trunk orientation as the determining factor of the contralateral deficit in the neglect syndrome and as the physical anchor of the internal representation of body orientation in space. Brain, 114, 1997–2014. Kirk, A. (2001). Spatial neglect. Current Neurology and Neuroscience Reports, 1, 541–546. Làdavas, E., Zeloni, G., Zaccara, G., & Gangemi, P. (1997). Eye movements and orienting of attention in patients with visual neglect. Journal of Cognitive Neuroscience, 9, 67–74.
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211 Mattingley, J. B., Bradshaw, J. L., Bradshaw, J. A., & Nettleton, N. C. (1994). Residual rightward attentional bias after apparent recovery from right hemisphere damage: Implications for a multicomponent model of neglect. Journal of Neurology, Neurosurgery, and Psychiatry, 57, 597–604. Nicholls, M. E., Bradshaw, J. L., & Mattingley, J. B. (1999). Free-viewing perceptual asymmetries for the judgement of brightness, numerosity and size. Neuropsychologia, 37, 307–314. Olk, B., Harvey, M., & Gilchrist, I. D. (2002). First saccades reveal biases in recovered neglect. Neurocase, 8, 306–313. Posner, M.I., Rafal, R.D. (1987). Cognitive theories of attention and the rehabilitation of attentional deficits. In M. J. Meier, A. L. Benton, L. Diller (Eds.), Neuropsychological rehabilitation (pp. 182–201). New York: Guilford Press. Posner, M. I., Walker, J., Friedrich, F. J., & Rafal, R. D. (1984). Effects of parietal injury on covert orienting of visual attention. Journal of Neuroscience, 4, 1863–1874. Pratt, J., Bekkering, H., Abrams, R. A., & Adam, J. (1999). The gap effect for spatially oriented responses. Acta Psychologica, 102, 1–12. Rousseaux, M., Beis, J.-M., Pradat-Diehl, P., Martin, Y., Bartolomeo, P., & Bernati, T. et al., (2001). Présentation d’une batterie de dépistage de la négligence spatiale. Normes et effet de l’ˆage, du niveau d’éducation, du sexe, de la main et de la latéralité. Revue Neurologique, 157, 1385– 1400.
1211
Sakhuja, T., Gupta, G. C., Singh, M., & Vaid, J. (1996). Reading habits affect asymmetries in facial affect judgements: A replication. Brain and Cognition, 32, 162–165. Saslow, M. G. (1967). Latency for saccadic eye movements. Journal of Optical Society of America, 57, 1030–1033. Schenkenberg, T., Bradford, D. C., & Ajax, E. T. (1980). Line bisection and unilateral visual neglect patients with neurological impairment. Neurology, 30, 509–517. Sprenger, A., Kömpf, D., Heide, W., (2002). Visual search in patients with left visual hemineglect. In J. Hyönä, D. P. Munoz, W. Heide, R. Radach (Eds.), Progress in Brain Research (Vol. 140, pp. 397–418) Amsterdam: Elsevier. Stone, S. P., Patel, P., Greenwood, R. J., & Halligan, P. W. (1992). Measuring visual neglect in acute stroke and predicting its recovery: the visual neglect index. Journal of Neurology, Neurosurgery, and Psychiatry, 55, 431–436. Tant, M. L. M., Kuks, J. B. M., Kooijman, A. C., Cornelissen, F. W., & Brouwer, W. H. (2002). Grey scales uncover similar attentional effects in homonymous hemianopia and visual hemineglect. Neuropsychologia, 40, 1474–1481. Walker, R., & Findlay, J. M. (1996). Saccadic eye movement programming in unilateral neglect. Neuropsychologia, 34, 493–508. Wilson, B., Cockburn, J., Halligan, P.W. (1987). The Behavioural Inattention Test. Bury St. Edmunds, UK: Thames Valley Test Company.
Residual oculomotor and exploratory deficits in patients with recovered hemineglect Tobias Pflugshaupt, Stefanie Almoslöchner Bopp, Dörthe Heinemann, Urs P. Mosimann, Roman von Wartburg, Thomas Nyffeler, Christian W. Hess, René M. Müri∗ Departments of Neurology and Clinical Research, Perception and Eye Movement Laboratory, University of Berne, Inselspital, Freiburgstrasse 10, 3010 Berne, Switzerland Received 19 May 2003; received in revised form 3 February 2004; accepted 4 February 2004
Abstract Several studies on hemineglect have reported that patients recover remarkably well when assessed with neuropsychological screening tests, however, they show deficits on novel or complex tasks. We investigated whether such deficits can be revealed with eye movement analysis, applying two basic oculomotor tasks as well as two exploratory tasks. Eye movements were recorded in eight hemineglect patients at least eleven months after right-hemisphere brain damage had occurred. Sixteen healthy volunteers participated in the control group. Regarding the basic oculomotor tasks, only the overlap task revealed residual deficits in patients, suggesting that a directional deficit in disengaging attention persisted during recovery. Further residual deficits were evident in the exploratory tasks. When everyday scenes were explored, patients showed a bias in early orienting towards the ipsilateral hemispace. In a search task, they demonstrated the same orienting bias as well as a non-directional deficit concerning search times. Moreover, patients preferentially fixated in the contralateral hemispace, but did not benefit from this asymmetry in terms of search times, i.e. they did not detect contralateral targets faster than ipsilateral ones. This suggests a dissociation between oculomotor processes and attentional ones. In conclusion, we have identified behavioural aspects that seem to recover slower than others. A disengagement deficit and biases in early orienting have been the most pronounced residual oculomotor deficits. © 2004 Elsevier Ltd. All rights reserved. Keywords: Human; Attention; Eye movements; Recovery of function
1. Introduction Acute hemineglect is characterised by a failure to report, respond, or orient to stimuli presented in the space contralateral to the lesioned hemisphere (Halligan & Marshall, 1994). This tendency towards the ipsilateral hemispace—i.e. the ipsilateral bias—can affect perceptual, motor, or representational processes (Kirk, 2001). Clinical experiences show that most hemineglect symptoms diminish in approximately 75% of patients within 6 months (Stone, Patel, Greenwood, & Halligan, 1992). However, this percentage is related to overt symptoms, such as test performance in copying or line bisection tasks. Several studies have revealed that other aspects of behaviour benefit less from recovery (Campbell & Oxbury, 1976; Goodale, Milner, Jakobson, & Carey, 1990; Harvey & Milner, 1999; Jehkonen et al., 2000; Karnath, 1988). These ∗ Corresponding author. Tel.: +41-316323081/368; fax: +41-316329679. E-mail address: [email protected] (R.M. Müri).
0028-3932/$ – see front matter © 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropsychologia.2004.02.002
findings suggest that residual hemineglect deficits tend to occur in novel or complex rather than simple or familiar settings. Karnath (1988) postulated that the speed of recovery from hemineglect was deficit-specific. According to his theory, acute hemineglect is characterised by three main components: (a) an early orienting of attention towards the ipsilateral side; (b) an impairment in disengaging attention from the ipsilateral side and reorienting towards contralateral stimuli; and (c) a non-directional reduction in information processing capacity. Karnath proposed that the second component recovered faster than the other two. Recent evidence (Mattingley, Bradshaw, Bradshaw, & Nettleton, 1994; Olk, Harvey, & Gilchrist, 2002) supports his notion with regard to the bias in early orienting. This deficit seems to be particularly persistent during recovery. Eye movement analysis provides an accurate assessment of hemineglect (Sprenger, Kömpf, & Heide, 2002) and its measurement is facilitated due to the limited number of relevant variables (i.e. fixations and saccades). In recent years,
1204
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211
many studies have used eye movement data to investigate particular aspects of acute hemineglect, such as saccadic eye movement programming (Walker & Findlay, 1996), visual space exploration (Karnath & Niemeier, 2002; Karnath, Niemeier, & Dichgans, 1998), or visual search behaviour (Husain et al., 2001; Sprenger et al., 2002). The aim of the present eye movement study was to reveal residual oculomotor and exploratory deficits in hemineglect patients who show substantial recovery with regard to clinical reports and in neuropsychological screening tests. First, two basic oculomotor tasks (i.e. gap and overlap task as described by Saslow (1967)) were administered to investigate speed and accuracy of visually guided saccades. Acute hemineglect patients have demonstrated delayed and hypometric saccades into the contralateral hemifield during these tasks (Heide & Kömpf, 1998; Walker & Findlay, 1996). We were interested in whether these deficits persisted during recovery. Second, exploratory behaviour was tested with a free exploration task and a search task. Since the centre of exploration is shifted towards the ipsilateral side in acute hemineglect patients (Karnath et al., 1998), we expected that such a shift would be less pronounced in residual hemineglect. The results of the exploratory tasks were further analysed to investigate early orienting. In accordance with a recent case report that focussed on residual hemineglect deficits in search behaviour (Olk et al., 2002), we hypothesised that our patient group would show a bias towards the ipsilateral side regarding early orienting. The present eye movement study on residual deficits in hemineglect differs from previous ones on acute hemineglect in two essential respects. First, we chose everyday photographs in order to represent real life situations. Previous studies used random arrays of letters (Husain et al., 2001; Karnath & Niemeier, 2002; Karnath et al., 1998) or geometrical shapes (Sprenger et al., 2002). Second, each of our patients was confronted with a large number of images (i.e. 32) per exploratory task. In previous studies (Husain et al., 2001; Karnath & Niemeier, 2002; Karnath et al., 1998), results were based on <10 images per exploratory task. By increasing the number of images per task, the results produced may depend less on image content and thus reflect systematic exploratory strategies more accurately.
ure and shape copying subtest of the behavioural inattention test (Wilson, Cockburn, & Halligan, 1987) as well as a line bisection task, which was similar to the one Schenkenberg, Bradford, & Ajax (1980) used. During hospitalisation (T1), Patients had to demonstrate pronounced asymmetries in one or several of these screening tests to enter the study. Since the manual of the behavioural inattention test does not provide side-specific cut-off scores, we adopted the scoring procedure of the Bells test (Gauthier, Dehaut, & Joanette, 1989) for the star and letter cancellation. A relevant asymmetry was thus characterised by a left-minus-right difference of more than two omissions (Rousseaux et al., 2001). For the figure and shape copying subtest, any left sided omission was considered as a relevant asymmetry (Azouvi et al., 2002). In line with the scoring procedure proposed by a clinical study (Ferber & Karnath, 2001), we applied a 14% cut-off criterion (i.e. mean percentage rightward deviation of the bisection mark in relation to the correct length) for the line bisection task. Furthermore, confrontational visual field testing at the end of hospitalisation verified that none of the patients suffered from visual field deficits. At the time of the experiment (T2), the four neuropsychological screening tests and a clinical interview were applied to assure that patients could be identified as a group of recovered hemineglect patients in terms of clinically relevant asymmetries. Eight patients met these conditions and participated in the study. Demographic and lesion data of the patient group is summarised in Table 1, clinical reports and neuropsychological test results are given in Table 2, and Fig. 1 depicts two illustrative examples of the recovery observed in the screening tests. Sixteen healthy volunteers participated in the control group. Patients and controls did not differ with regard to age (patients: median = 47, range = 32; control subjects: median = 47, range = 34), gender, and handedness (Mann–Whitney and χ2 -tests: n.s.). None of them suffered from depression (Beck Depression Inventory/BDI, Hautzinger, Bailer, Wollall, & Keller, 1994), colour blindness (Ishihara colour plates, Ishihara, 1999), visuo-motor
2. Methods 2.1. Patients and control subjects The hemineglect patients who participated in the study were former inpatients at our division of neuropsychological rehabilitation. All patients showed a history of acute hemineglect with typical clinical signs (e.g. bumping into things placed in the left half of space) after right-hemisphere damage. The hemineglect diagnosis was based on a wide range of neurological and neuropsychological examinations, including the star cancellation, letter cancellation, and the fig-
Fig. 1. Figure copying examples of patient 4 (top) and patient 8 (bottom). Original stimulus (left), copying performance during hospitalisation (T1; middle) and at the time of the experiment (T2; right).
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211
1205
Table 1 Demographic and lesion data of the patients (f: frontal, t: temporal, p: parietal, th: thalamic)
P1 P2 P3 P4 P5 P6 P7 P8
Sex
Age
Aetiology
Localisation of lesion
Time since onset of brain damage in months
Male Male Female Female Female Male Male Male
39 29 54 54 43 43 61 51
Ischaemic stroke Ischaemic stroke Ischaemic stroke Ischaemic stroke Ischaemic stroke Haematoma Haematoma Haematoma
f, t f th f, p f, t p, t f, t f, t
17 35 15 60 19 11 16 70
Mean
47
30
S.D.
10
23
Table 2 Clinical reports and neuropsychological test results of the patient group Clinical reportsa
P1 P2 P3 P4 P5 P6 P7 P8
Cancellation tasks
Line bisection
Figure copying
T1
T2
T1
T2
T1
T2
T1
T2
X X X X X X X X
– – – – – – – –
X X X X X X X X
– – – – – – – –
– X – X – X – X
– – – – – – – –
– – – X X – – X
– – – – – – – –
Summary of clinically relevant asymmetries (X) during hospitalisation (T1) and at the time of the experiment (T2). a Refers to clinical observations (e.g. bumping into things placed in the left half of space) during hospitalisation (T1), respectively to the clinical interview at the time of the experiment (T2).
disturbances (e.g. ocular motor paralysis, strabism) or limited visual acuity (normal or corrected to normal). Both patients as well as control subjects gave written informed consent prior to participation. The study was approved by the local ethical committee. 2.2. Apparatus Eye movements were recorded using an infrared-based video tracking system (EyeLink, SensoMotoric Instruments TM, GmbH, Berlin, Germany). This system collects eye movement data at a sampling rate of 250 Hz with a spatial resolution of 0.01◦ . It provides an accuracy of gaze-position relative to stimulus coordinates of 0.5◦ –1.0◦ , depending largely on participants’ fixation accuracy during calibration. A chin rest was used to both ensure constant distance and minimise head movements. Moreover, head motion compensation was computed by the tracking system. Minimal fixation duration was set to 120 ms and saccades had to exceed 0.75◦ in order to be analysed. Participants were seated 70 cm in front of a 19 in. screen, subtending a visual angle of approximately 29◦ × 22◦ . Eye movements of either the right or the left eye were registered, with both eyes equally often recorded across all participants. The system was cal-
ibrated prior to each block of stimuli by means of a 3 × 3 point grid. 2.3. Eye movement recording 2.3.1. Oculomotor tasks Two types of basic oculomotor tasks were applied and both of them required participants to perform visually guided reflexive saccades. The time course of stimulus presentation in these two tasks is shown in Fig. 2. During the gap task, the central fixation (CF) point was presented pseudo-randomly with durations ranging from
Fig. 2. Gap and overlap task. Time course of stimulus presentation.
1206
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211
2000 to 3000 ms. The offset of the CF preceded the onset of the lateral target (LT) by 200 ms (i.e. temporal gap). Target amplitude (range 3.7◦ –9.1◦ ) and direction were randomised in the horizontal plane. The LT was presented for 1000 ms. Following the offset of the LT, the next CF was presented immediately. Participants were instructed to fixate the point that was presented, regardless of whether it appeared centrally (CF) or laterally (LT). During the overlap task, stimulus presentation was identical to the gap task except for the CF that remained visible during the presentation of the LT. The participants therefore had to actively disengage attention from the CF before looking at the LT. This contrasts the gap task, in which disengagement is facilitated by the temporal gap. The overlap task required participants to look at the LT as soon as it appeared and then to refixate the CF as soon as the LT disappeared. Overall, 28 targets (i.e. 14 ipsilateral and 14 contralateral targets) were shown per participant and task. The off-line analysis contained a comparison of ipsi—versus contralateral values with regard to the following parameters: latency of the first saccade, gain (i.e. saccade amplitude/target amplitude) of the first and second saccade, and the gap effect, which corresponds to the difference between overlap and gap latency. It is a well described finding that a temporal gap between fixation point offset and target onset typically results in shorter saccadic latencies than if the fixation point remains visible (Pratt, Bekkering, Abrams, & Adam, 1999). Two trial exclusion criteria were applied for both oculomotor tasks. First, anticipatory saccades were excluded since
our focus was on reflexive saccades. Saccades with negative latencies (i.e. saccade initiation occurred prior to the onset of the LT) or with latencies below 80 ms were therefore not analysed. Second, saccades with a starting point that deviated more than one degree from the CF were excluded as well, since these saccades would strongly interfere with the validity of gain calculations. 2.3.2. Exploratory tasks Exploratory behaviour was investigated under two experimental conditions. First, a free exploration task that consisted of everyday scenes was applied. Each scene was presented for 7 s. Second, patients and control subjects had to solve a search task, during which they were instructed to find and click a predefined object (i.e. leaf, screw, clock, roll, snake, butterfly, pear, or scissors) placed in an everyday scene as quickly as possible. Search times longer than 7 s were treated as omissions. Object size was 1.8◦ × 1.5◦ on average and the overall distribution of objects was balanced across quadrants. Fig. 3 depicts two example stimuli for each exploratory task. Each participant was confronted with 32 images per exploratory task in total. Blocks of 8 images were formed to recalibrate the system between blocks. In the search task, the nature of the object (i.e. leaf, screw, etc.) was varied on a block-to-block interval. All images were 29◦ × 22◦ in size and information content was balanced to prevent an image-induced ipsi—or contralateral bias. Prior to every image, a central fixation point was presented to assure that
Fig. 3. Free exploration of everyday scenes (left) and search task (right). Two examples; superimposed white arrows show targets (clock, snake) in the search task examples.
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211
1207
Table 3 Basic oculomotor tasks Left
Right
AQ
patients
controls
patients
controls
patients
controls
P
Gap
Latency Gain1 Gain2
158 (64) 0.85 (0.50) 1.00 (0.10)
157 (84) 0.91 (0.30) 1.02 (0.15)
183 (126) 0.85 (0.54) 0.96 (0.16)
162 (70) 0.91 (0.29) 0.99 (0.10)
0.89 (0.51) 1.01 (0.74) 0.94 (0.17)
0.96 (0.50) 0.99 (0.48) 0.96 (0.19)
0.153 0.653 0.350
Overlap
Latency Gain1 Gain2
307 (265) 0.87 (0.70) 0.97 (0.12) 158 (273)
240 (150) 0.95 (0.40) 1.00 (0.15) 82 (159)
222 (155) 0.93 (0.59) 0.98 (0.22) 36 (137)
219 (187) 0.93 (0.34) 0.99 (0.14) 60 (142)
1.29 1.07 1.01 0.30
1.10 1.00 0.99 0.62
0.007∗∗ 0.153 0.417 0.023∗
Gap effect
(0.82) (1.30) (0.27) (0.92)
(0.41) (0.30) (0.23) (1.35)
Latency and gap effect in ms; gain of the first (gain1) and second (gain2) saccade as quotients. Values are medians; ranges in parentheses; AQ: asymmetry quotient; P-values (Mann–Whitney tests) refer to group comparisons of AQs. ∗ Significant at the 0.05 level. ∗∗ Significant at the 0.01 level.
patients and control subjects started exploring the images in a central location. The off-line analysis included plots of the horizontal distribution of fixation frequency. These plots illustrate the centre of exploration in a descriptive manner. Furthermore, parameters such as the number of fixations, the fixation duration, and the saccadic amplitude were compared between groups. The main focus was on the comparison of ipsi—with contralateral values regarding first saccade parameters (i.e. direction, latency and amplitude) and overall fixation patterns (i.e. number of fixations and total fixation duration). Since exploratory behaviour was likely to change during stimulus presentation, total fixation duration was further analysed applying a second-by-second design. Together with first saccade parameters, this analysis allowed an investigation of early orienting in exploratory behaviour. In the search task, two additional parameters were examined: the percentage of hits and the average search time defined as the duration from stimulus onset to the mouseclick. 2.4. Experimental procedure The order of experimental conditions was held constant across participants. Basic oculomotor tasks were followed by exploratory tasks. During eye movement recordings, participants were seated in a dimly lit room. All tasks were preceded by verbal instructions and practise trials. 2.5. Data analysis For most parameters of oculomotor and exploratory behaviour, asymmetry quotients (AQs) based on individual medians were calculated (i.e. ipsilateral/contralateral value1 ). However, asymmetries with regard to latencies and search times refer to inverted ratios (i.e. contralateral/ipsilateral 1 Although the expressions ‘ipsilateral’ and ‘contralateral’ relate to lesion sites and are therefore inadequate with regard to healthy control subjects, we nonetheless use these expressions for both patients and control subjects (i.e. ‘ipsilateral’ = right, ‘contralateral’ = left), for the purpose of a consistent terminology across groups.
value). This inversion was necessary to illustrate asymmetries in a consistent manner. As a consequence, asymmetries towards the contralateral side are represented by AQs < 1, AQs > 1 indicate asymmetries towards the ipsilateral side and an AQ = 1 describes exact symmetry. Since AQs were not normally distributed (Kolmogorov–Smirnov test), group comparisons were based on non-parametric tests (Mann–Whitney or χ2 -tests). Similarly, medians and ranges will be presented as central tendency and dispersion measures. A P-value of <0.05 was considered statistically significant and all tests were two-tailed.
3. Results 3.1. Basic oculomotor tasks In general, data loss due to anticipatory saccades or central fixation problems was relatively low in both tasks with group-specific medians below 5%. In the gap task, no statistically significant group differences were found. However, Table 3 shows that the overlap task revealed significant differences. Patients’ latencies were increased when the LT appeared on the contralateral as opposed to the ipsilateral side. Corresponding AQs differed significantly between groups. As a consequence of this deficit, the gap effect of patients was considerably less pronounced on the ipsilateral than on the contralateral side. Again, corresponding AQs differed significantly between groups. Both results indicate a task-specific dissociation. Residual deficits in patients’ basic oculomotor behaviour were apparent in the overlap task only. 3.2. Exploratory tasks 3.2.1. Centre of exploration Plotting the horizontal distribution of fixation frequency allowed the assessment of the centre of exploration in a descriptive manner. In contrast to acute hemineglect patients,
1208
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211 contralateral
ipsilateral
1 2
25
3 20
4 5
15
6 7 .4
10
.6
.8
1.0
1.2
1.4
1.6
AQ 5
Fig. 5. Free exploration of everyday scenes. Time course of AQs with regard to total fixation duration from first to seventh second (y-axis); patients (black) vs. controls (grey); based on medians.
0 25
20
15
10
5
0
Fig. 4. Horizontal distribution of fixation frequency. Free exploration of everyday scenes (top), search task (bottom); eight columns of 3.6◦ each (x-axis), percentage of fixations (y-axis); patients (black) vs. controls (grey).
Fig. 4 illustrates that the group of recovered patients did not show an ipsilaterally shifted centre of exploration. In fact, a slight shift towards the contralateral side was evident in both tasks.
3.2.2. Free exploration of everyday scenes When exploring everyday scenes, patients made fewer fixations in number, longer fixations in duration, as well as smaller saccades in amplitude than controls. None of these differences, however, reached statistical significance. Comparing ipsi- and contralateral values with regard to overall fixation patterns did not reveal significant group differences either. Nevertheless, significant differences were observed when early orienting was analysed, with the first saccade being particularly sensitive to residual deficits. The patient group made more and larger first saccades towards the ipsilateral as opposed to the contralateral hemispace while the control subjects made relatively more first saccades towards the contralateral side. Corresponding AQs differed significantly between groups (Table 4). Based on a second-by-second design, Fig. 5 shows the time course of exploring everyday scenes with regard to AQs in total fixation duration. Statistical analyses revealed significant group differences within the first two seconds. During the first second, patients’ total fixation time was distributed rather symmetrically whereas controls spent more fixation time on the contralateral side (Mann–Whitney: Z = −2.14, P = 0.032). Corresponding AQs, however, changed substantially during the following second. Controls moved
Table 4 Parameters of the first saccade Leftward
Rightward
AQ
patients
controls
patients
controls
patients
controls
P
Free exploration of everyday scenes
Direction Latency Amplitude
45 (41) 333 (143) 3.1 (4.2)
56 (63) 257 (163) 4.4 (3.8)
55 (41) 289 (131) 4.3 (3.2)
44 (63) 286 (179) 4.1 (4.2)
1.21 (2.48) 1.08 (0.55) 1.33 (0.70)
0.78 (2.86) 0.96 (0.66) 0.89 (1.28)
0.038∗ 0.136 0.023∗
Search task
Direction Latency Amplitude
46 (53) 284 (231) 3.4 (4.3)
59 (31) 261 (131) 4.7 (4.7)
54 (53) 291 (140) 4.1 (4.5)
41 (31) 241 (153) 4.8 (5.1)
1.17 (3.94) 1.11 (0.75) 1.40 (1.21)
0.69 (0.95) 1.08 (0.61) 1.11 (1.21)
0.019∗ 0.350 0.023∗
Direction in %, latency in ms and amplitude in ◦ . Values are medians, ranges in parentheses; P-values (Mann–Whitney tests) refer to group comparisons of AQs. ∗ Significant at the 0.05 level.
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211
1209
Table 5 Fixation patterns and search performance Left
Right
patients
controls
patients
AQ controls
patients
controls
P
Free exploration of everyday scenes
Number of fixations Total fixation duration
11 (4) 3.10 (1.64)
11 (6) 3.08 (1.07)
9 (6) 2.66 (1.81)
10 (7) 2.94 (1.19)
0.82 (0.81) 0.81 (0.96)
0.93 (0.65) 0.95 (0.65)
0.264 0.383
Search task
Number of fixations Total fixation duration Search time Percentage of hits
4 (7) 2.47 (2.07) 3.01 (3.31) 62.5 (50)
3 (3) 1.33 (1.24) 1.48 (1.94) 75 (50)
4 (5) 2.21 (2.34) 2.98 (2.97) 59.4 (56)
3 (2) 1.41 (1.05) 1.77 (1.87) 75 (50)
1.00 0.91 1.07 1.00
1.00 1.04 0.83 1.00
0.787 0.038∗ 0.038∗ 0.928
(1.00) (0.25) (0.72) (0.43)
(1.20) (0.70) (1.14) (0.27)
Number of fixations, total fixation duration in s, search time in s, and percentage of hits in %. Values are medians, ranges in parentheses; P-values (Mann–Whitney tests) refer to group comparisons of AQs. ∗ Significant at the 0.05 level.
to the ipsilateral (i.e. right) side while patients preferentially explored the contralateral side (Mann–Whitney: Z = −2.45, P = 0.013). Thus, when exploring everyday scenes freely, patients differed from control subjects during the initial phase of exploration only. 3.2.3. Search task In the search task, patients made more fixations in number, shorter fixations in duration, as well as smaller saccades in amplitude than controls. Statistical analyses showed that the difference concerning the number of fixations was significant (Mann–Whitney: Z = −2.29, P = 0.023). However, this significance must be put into perspective, since patients’ search times were significantly longer for both ipsilaterally (Mann–Whitney: Z = −2.388, P = 0.016) and contralaterally presented targets (Mann–Whitney: Z = −3.123, P = 0.001), resulting in a higher number of fixations. In addition, an interesting dissociation occurred in patients’ search behaviour. Despite relatively longer total fixation durations in the contralateral hemispace, patients did not benefit from this asymmetry in terms of search times, i.e. they did not detect contralateral targets faster than ipsilateral ones. Since controls’ total fixation durations were rather symmetrical and search times contralaterally biased, both group comparisons regarding AQs reached statistical significance (Table 5). With regard to the percentage of hits, patients found slightly less targets than controls (Mann–Whitney tests: n.s.), irrespective of whether the target was localised ipsi- or contralaterally. Corresponding AQs, however, did not differ significantly between groups. Due to short search times, a second-by-second analysis of fixation behaviour was not appropriate for the search task. Nevertheless, early orienting could be examined by focussing on first saccade parameters. Results were highly similar to the ones obtained in the free exploration of everyday scenes. Significant group differences were found with regard to direction and amplitude of the first saccade. The patient group made more and larger saccades towards the ipsilateral as opposed to the contralateral hemispace while the control subjects made relatively more first saccades towards the contralateral side (Table 4).
4. Conclusion The results of the present study can be summarised as follows: At least eleven months after the onset of brain damage, eight hemineglect patients showed no clinically relevant asymmetries with regard to a clinical interview, in neuropsychological screening tests, and concerning the centre of exploration in two exploratory tasks. This is in accordance with the assumption that most neglect symptoms diminish within six months (Stone et al., 1992). However, residual deficits occurred in a basic oculomotor task, in early orienting, and with regard to search performance. We thus identified behavioural aspects that seem to recover slower than others. With regard to basic oculomotor behaviour, the overlap task revealed ipsilateral biases in patients whereas the gap task did not. While acute hemineglect patients have shown ipsilateral biases in gap and overlap tasks (Behrmann, Ghiselli-Crippa, & Dimatteo, 2001/2002; Heide & Kömpf, 1998; Karnath, Schenkel, & Fischer, 1991; Walker & Findlay, 1996), our results suggest that the overlap task is particularly sensitive to residual hemineglect deficits. With reference to Posner, Walker, Friedrich, & Rafal (1984, 1987), we assume that this indicates impaired disengagement of attention, when the following saccade is directed towards the contralateral side. The underlying assumption is that the overlap task requires participants to actively disengage attention whereas during the gap task, disengagement is facilitated by the temporal gap (Pratt et al., 1999). This directional disengagement deficit may be a first behavioural aspect that seems to persist during recovery from hemineglect. Patients’ exploratory behaviour was characterised by residual deficits during the initial phase of exploration. When everyday scenes were explored, patients initially oriented towards the ipsilateral side as opposed to control subjects who preferentially oriented towards the contralateral (i.e. left) side. Biases in early orienting of acute hemineglect patients have been reported in many studies. (e.g. Azouvi et al., 2002; Gainotti, D’Emre, & Bartolomeo, 1991; Mattingley et al., 1994; Olk et al., 2002). A recent study based on 206 patients (Azouvi et al., 2002) has provided
1210
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211
evidence that a right-sided starting point in a cancellation task was the most sensitive paper and pencil measure for acute hemineglect. The present study suggests that this bias in early orienting may be a second behavioural aspect that persists during recovery. This is consistent with previous studies on residual hemineglect (Mattingley et al., 1994; Olk et al., 2002) and with Karnath’s model of recovery (1988). Similar to previous findings (Gainotti et al., 1991; Mattingley et al., 1994; Tant et al., 2002), the present study detected early orienting asymmetries in healthy subjects as well. In contrast to the ipsilateral bias found in neglect patients, healthy subjects initially oriented towards the contralateral (i.e. left) hemifield, irrespective of task. It has been suggested that this asymmetry results from dominant right hemisphere activation for spatial processing (Tant et al., 2002). An alternative account is that acquired scanning habits determine healthy subjects’ early orienting. Such habits may be strongly related to language aspects and reading (Nicholls, Bradshaw, & Mattingley, 1999; Sakhuja, Gupta, Singh, & Vaid, 1996). The search task revealed three interesting findings. First and similar to the free exploration of everyday scenes, patients’ early orienting was ipsilaterally biased. Second, search times were significantly longer in patients relative to controls, irrespective of whether the object was presented contra- or ipsilaterally. This was also shown by a previous study (Heide & Kömpf, 1998) and it may correspond to the non-directional reduction in information processing capacity that is seen as a major component in acute and residual hemineglect (Karnath, 1988). Finally, patients preferentially fixated in the contralateral hemifield, but they did not benefit from this asymmetry in terms of search times, i.e. they did not find contralateral targets faster than ipsilateral ones. The latter finding can be interpreted as a dissociation between underlying processes. While oculomotor behaviour in terms of overall fixation patterns seems to have recovered considerably, attentional processes as indicated by search times may recover slower. A similar finding has been described in acute hemineglect (Làdavas, Zeloni, Zaccara, & Gangemi, 1997). When instructed to maintain central fixation while responding manually to peripheral targets, patients made oculomotor responses to the contralateral side without a corresponding shift of attention as documented by the absence of manual responses. This finding was discussed in terms of a partial separation of attentional processes from oculomotor ones. The present study provides evidence that this separation is likely to persist during recovery. Therefore, one may ask whether recovery from acute hemineglect should be regarded as a compensatory mechanism rather than true functional recovery. The dissociation between oculomotor and attentional processes during the search task favours compensatory mechanisms in our patient group. Nevertheless, patients’ normal performances in the gap task or with regard to the centre of exploration in the exploratory tasks indicates functional recovery, at least to some extent.
Acknowledgements This work was supported by grants from the Thomas Stanely Johnson Foundation.
References Azouvi, P., Samuel, C., Louis-Dreyfus, A., Bernati, T., Bartolomeo, P., & Beis, J.-M. et al., (2002). Sensitivity of clinical and behavioural tests of spatial neglect after right hemisphere stroke. Journal of Neurology, Neurosurgery, and Psychiatry, 73, 160–166. Behrmann, M., Ghisseli-Crippa, T., & Dimatteo, I. (2001/2002). Impaired initiation but not execution of contralesional saccades in hemispatial neglect. Behavioural Neurology, 13, 39–60. Campbell, D., & Oxbury, J. (1976). Recovery from unilateral visuospatial neglect. Cortex, 12, 303–312. 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. Gainotti, G., D’Emre, P., & Bartolomeo, P. (1991). Early orientation of attention toward the half-space ipsilateral to the lesion in patients with unilateral brain damage. Journal of Neurology, Neurosurgery, and Psychiatry, 54, 1082–1089. 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. Goodale, M. A., Milner, A. D., Jakobson, L. S., & Carey, D. P. (1990). Kinematic analysis of limb movements in neuropsychological research: Subtle deficits and recovery of function. Canadian Journal of Psychology, 44, 180–195. Halligan, P. W., & Marshall, J. C. (1994). Towards a principled explanation of unilateral neglect. Cognitive Neuropsychology, 11, 167–206. Harvey, M., & Milner, A. D. (1999). Residual perceptual distortion in ‘recovered’ hemispatial neglect. Neuropsychologia, 37, 745–750. Hautzinger, M., Bailer, M., Wollall, H., Keller, F. (1994). BeckDepressions-Inventar (BDI): Testhandbuch. Bern: Hans Huber. Heide, W., & Kömpf, D. (1998). Combined deficits of saccades and visuospatial orientation after cortical lesions. Experimental Brain Research, 123, 164–171. 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. Ishihara, S. (1999). Ishihara’s test for color deficiency. Tokyo: Kanehara & Co. Jehkonen, M., Ahonen, J.-P., Dastidar, P., Koivisto, A.-M., Laippala, P., & Vilkki, J. et al., (2000). Visual neglect as a predictor of functional outcome one year after stroke. Acta Neurologica Scandinavica, 101, 195–201. Karnath, H.-O. (1988). Deficits of attention in acute and recovered visual neglect. Neuropsychologia, 26, 27–43. Karnath, H.-O., & Niemeier, M. (2002). Task-dependent differences in the exploratory behaviour of patients with spatial neglect. Neuropsychologia, 40, 1577–1585. Karnath, H.-O., Niemeier, M., & Dichgans, J. (1998). Space exploration in neglect. Brain, 121, 2357–2367. Karnath, H.-O., Schenkel, P., & Fischer, B. (1991). Trunk orientation as the determining factor of the contralateral deficit in the neglect syndrome and as the physical anchor of the internal representation of body orientation in space. Brain, 114, 1997–2014. Kirk, A. (2001). Spatial neglect. Current Neurology and Neuroscience Reports, 1, 541–546. Làdavas, E., Zeloni, G., Zaccara, G., & Gangemi, P. (1997). Eye movements and orienting of attention in patients with visual neglect. Journal of Cognitive Neuroscience, 9, 67–74.
T. Pflugshaupt et al. / Neuropsychologia 42 (2004) 1203–1211 Mattingley, J. B., Bradshaw, J. L., Bradshaw, J. A., & Nettleton, N. C. (1994). Residual rightward attentional bias after apparent recovery from right hemisphere damage: Implications for a multicomponent model of neglect. Journal of Neurology, Neurosurgery, and Psychiatry, 57, 597–604. Nicholls, M. E., Bradshaw, J. L., & Mattingley, J. B. (1999). Free-viewing perceptual asymmetries for the judgement of brightness, numerosity and size. Neuropsychologia, 37, 307–314. Olk, B., Harvey, M., & Gilchrist, I. D. (2002). First saccades reveal biases in recovered neglect. Neurocase, 8, 306–313. Posner, M.I., Rafal, R.D. (1987). Cognitive theories of attention and the rehabilitation of attentional deficits. In M. J. Meier, A. L. Benton, L. Diller (Eds.), Neuropsychological rehabilitation (pp. 182–201). New York: Guilford Press. Posner, M. I., Walker, J., Friedrich, F. J., & Rafal, R. D. (1984). Effects of parietal injury on covert orienting of visual attention. Journal of Neuroscience, 4, 1863–1874. Pratt, J., Bekkering, H., Abrams, R. A., & Adam, J. (1999). The gap effect for spatially oriented responses. Acta Psychologica, 102, 1–12. Rousseaux, M., Beis, J.-M., Pradat-Diehl, P., Martin, Y., Bartolomeo, P., & Bernati, T. et al., (2001). Présentation d’une batterie de dépistage de la négligence spatiale. Normes et effet de l’ˆage, du niveau d’éducation, du sexe, de la main et de la latéralité. Revue Neurologique, 157, 1385– 1400.
1211
Sakhuja, T., Gupta, G. C., Singh, M., & Vaid, J. (1996). Reading habits affect asymmetries in facial affect judgements: A replication. Brain and Cognition, 32, 162–165. Saslow, M. G. (1967). Latency for saccadic eye movements. Journal of Optical Society of America, 57, 1030–1033. Schenkenberg, T., Bradford, D. C., & Ajax, E. T. (1980). Line bisection and unilateral visual neglect patients with neurological impairment. Neurology, 30, 509–517. Sprenger, A., Kömpf, D., Heide, W., (2002). Visual search in patients with left visual hemineglect. In J. Hyönä, D. P. Munoz, W. Heide, R. Radach (Eds.), Progress in Brain Research (Vol. 140, pp. 397–418) Amsterdam: Elsevier. Stone, S. P., Patel, P., Greenwood, R. J., & Halligan, P. W. (1992). Measuring visual neglect in acute stroke and predicting its recovery: the visual neglect index. Journal of Neurology, Neurosurgery, and Psychiatry, 55, 431–436. Tant, M. L. M., Kuks, J. B. M., Kooijman, A. C., Cornelissen, F. W., & Brouwer, W. H. (2002). Grey scales uncover similar attentional effects in homonymous hemianopia and visual hemineglect. Neuropsychologia, 40, 1474–1481. Walker, R., & Findlay, J. M. (1996). Saccadic eye movement programming in unilateral neglect. Neuropsychologia, 34, 493–508. Wilson, B., Cockburn, J., Halligan, P.W. (1987). The Behavioural Inattention Test. Bury St. Edmunds, UK: Thames Valley Test Company.