DISENTANGLING ROBIN WALKER,*
NEGLECT
JOHN M. FINDLAY,*?
AND HEMIANOPIA
ANDREW W. YOUNG*
and JOHN WELCH:
*Department of Psychology, University of Durham, Durham DHI 3LE, U.K.; and :Regional Neuroscience Centre, Department of Clinical Psychology, Newcastle General Hospital, Westgate Road, Newcastle upon Tyne NE4 6BE, U.K. (Rrcriwd
22 Fdmmry
I99 I : ucwpted2 Maq’ I99 I )
Abstract-In this paper we report findings which question the diagnosis of a hemianopia in B.Q., a ho-year-old lady who shows unilateral spatial neglect following a lesion to the right parietal lobe. The presence of a hemianopia has been indicated following two independent assessments of B.Q.‘s visual fields. We examined B.Q.‘s performance on a visuo-spatial task in which single or double stimuli were displayed left and right of a central fixation point. B.Q. failed to report left stimuli when the fixation point was continuously displayed (OVERLAP CONDITION). This performance is consistent with the suggestion of a left hemianopia as indicated by perimetric field testing. In a further condition the fixation point was extinguished prior to stimuli onset (GAP CONDITION). B.Q. consistently responded to left stimuli in the gap condition and also showed improved performance to right stimuli. Eye movements were recorded on a separate testing session, in which B.Q. showed a normal saccadic response to left targets in the gap condition, but not during the overlap condition. These results suggest that B.Q.‘s failure to report left stimuli during field plotting is due to her neglecting left stimuli and not because she has a visual field defect. By using a testing procedure which reduces the severity of neglect B.Q. is able to respond to left stimuli. Reducing the severity of B.Q.‘s neglect also abolishes the apparent hemianopia. Standard field plotting techniques which use a consistently displayed fixation point may not be appropriate for the testing of neglect patients’ visual fields.
INTRODUCTION
[JNILATERAI.spatial neglect is a condition in which patients fail to respond to stimuli in the side of space opposite to the lesion site. Neglect is typically associated with damage to the right parietal lobe resulting in left side neglect. On neurological examination many neglect patients are diagnosed as having a homonymous hemianopia, with or without macular sparing. The visual field defect is not generally thought to be the cause of neglect as the patient is free to make compensatory head or eye movements to orient in the neglected direction. Such compensatory strategies are found in patients who have a homonymous hemianopia, without neglect 19, 121. Conversely, it is possible to find patients without visual held defects who are shown to have neglect symptoms [S, lo]. Neglect is therefore thought to have an attentional 16, 151, or representational [2], underlying cause and any hemianopia is regarded as a subsidiary finding. There is an immediate problem concerning a diagnosis of hemianopia in neglect patients. Standard visual field testing involves maintaining fixation while stimuli are presented at various locations in the visual field. Failure to report stimuli is considered to reflect a field defect such as hemianopia. However unilateral spatial neglect by definition involves a failure
tTo whom reprint
requests
should
be addressed. 1019
1020
R. WALK~K t’t al.
to respond to stimuli on one side of space. In effect standard tests of hemianopia may also constitute tests for unilateral neglect. Is it possible to tell whether failure to report stimuli on one side of space is due to a field defect, or to neglect? This paper describes a study of a patient B.Q. who has profound unilateral spatial neglect. She has also been considered to have a hemianopia using perimetry techniques. The issue of hemianopia and neglect are examined on a visuo-spatial orienting task and by an examination of eye movements during such a task. It is shown that what appears as a profound hemianopia under certain testing conditions can be modified under slightly different conditions to a much milder extinction like disorder.
CASE HISTORY B.Q.. a 66-year-old retired professional woman was admitted to hospital in August 1989, with a suspected stroke. Neurological examination revealed normal cranial nerves and fundi, hemianopia, deviation of the eyes and tongue IO the left. and left-sided Facial weakness. A CT scan revealed an extensive, low denstty lesion surrounding the right parictal region. involving the cortical grey matter and deep white matter. Thcrc was moderate mid-hnc displacement. The appearances were consistent with an infarct in the territory of the right middle ccrcbral artcry. Psychological examination in September 1989 showed no significant impairment of verbal recall or immediate memory span (digit span 7 forward and 4 backwards). Verbal comprehension. vocabulary naming and remote memory were all intact. B.Q. was found to have severe unilateral spatial neglect (line bisection 40 50% shift from midpoint). Visual copying tasks and visual short term memory were severely impaired. Lcttercancellation tasks and reading tasks also showed severe impairment. The block design subtest of the Weschlcr was impossible. Occupational therapists indicated problems with body image and spatial relations. and described her as having dressing dyspraxia. Visual field testing was carried out a second time in July 1990, using screen perimetry. The results agam indicated a dense left hemianopia, with prcscrvation at 5 (macular sparing). Crossing out tasks and line bisection tasks carried out at this time indicate that her neglect condition has remained static. Near visual acuity (Curry and Paxton “Curpax” test) was shown to be good (approximately 6;4). Contraht sensitivity was in the normal range (Vistcch VCTS 6000).
METHOD Stimuli were generated by a BBC Mater scrics microcomputer (Acorn Computers Ltd), by uaing the Mode I character set, and were dlsplayed on a VDU monitor (PhIllIps Monitor X0). Horizonlal eye movements wcrc recorded using an infrared binocular eye tracker (ACS model EM 130). Analog records were recorded on a Tinhcrg scrics 100 tape recorder; one channel recorded eye movements, while a second stored a marker signal which indicated trial start and finish.
Over a period of months B.Q. was tested on three diKerent versions of the task. On one further session eye movements were recorded. The procedure involved maintaining fixation on a cross, displayed centrally on a VDlJ screen. On each trial. either a single target was then dlsplayed in the left or right visual field, or two targets wcrc displayed simultaneously in both visual fields. B.Q. was asked to look at the lixation cross and then report an Indicator digit which appeared at the target location(s). Responses given provided a measure ofspatial neglect, and of the visual field capacity. B.Q. was seated centrally in front ofthe VDU screen, mounted at eye Icvcl, with B viewing distance of SO cm. Head movements were not constrained, except that a chin rest was used when eye movements were recorded. The length of test sessions vdried, since testing wan restricted by constraints such as, fatigue during testing session, other therapy being received, etc. The aims of recording eye movements were to show: (I ) If central fixation was achieved at trial onset. (2) If saccadc latencies were affected by target type and cxpcrimen~al condillon. (3) If any saccade5 occurred to left single targets, even though the indicator digit was not rcportcd. The main difference bctwccn the conditlon5 conccrncd the activity at the fixation point as the tnrgct wa\
HLMIANOPIA
OK NEGLLCT’!
1021
presented. In the “overlap” condition the fixation cross stayed on throughout the trial sequence and targets appeared 1000 msec after initial fixation onset. In the “0 gap” condition the fixation cross was displayed for X00 mscc. Fixation cross o&et was simultaneous with target onset. In the “+ 100 gap” condition the fixation cross was displayed for 1000 msec and was extinguished 100 msec before the target was presented, leaving a blank screen on to which targets were displayed. Targets were presented in blocks under one of the three experimental conditions. Each trial consisted of one of two possible types of display sequence. namely single target trials, or double target trials. There were equal numbers of single and double trials in each experimental block.
This condition provided an objective measure of the severity of neglect under gdp and overlap conditions. Target eccentricities wcrc 2.5,s and 7.5 (measured horizontally from a vertical centre line), on the horizontal and principal oblique axes. This produced nine left visual field and nine right visual held target locations. The single target appeared at one of these 1X possible locations in a pseudorandom order. On eye movement recording sessions, targets appeared on the horizontal axis only, at 4 and 8” left and right of the fixation cross. The horizontal axis alone was used to enable accurate interpretation of saccade records.
Double targets were used to obtain a measure of extinction. Targets appeared simultaneously at equal eccentricities in the left and right visual fields, on either the horizontal axis, or diametrically opposed on the principal oblique axes. In the eye movement recording trials they appeared at equal eccentricities on the horizontal axis only.
The timing sequence used for all testing sessions is shown in Fig. 1. Each trial began with a blank screen. A fixation cross (0.57’ ) then appeared in the centre of the screen. B.Q. was instructed to attend to this fixation point and respond verbally to it by saying “CROSS” when it was detected. This aimed to ensure that central fixation was achieved at the beginning ofeach trial. On trials where the cross was not reported any response made subsequently was not included in the analysis. B.Q. had little difficulty in following these instructions and less than 5% of trials were eliminated on this criterion. The target was a uniformly illuminated square of side 0.57 ‘, presented for 100 msec. A delay of 100 msec then occurred during which time it was expected that an eye movement would occur. An indicator digit was then displayed at the target position (two different indicators were presented on double target trials). The indicator digits wcrc selected from the digits I 4 using a pseudorandom sequence. They were displayed for 300 msec following which time they were replaced by a mask (the same square constituting the target). B.Q. was required to report any indicator digit detected and the experimenter keyed the responses into the computer. There was a delay of 500 msec between each trial.
RESULTS Figure 2 shows the percentage of indicator digits reported by B.Q., on sessions when eye movements were not recorded. The + 100 gap and overlap testing sessions were both carried out on the same day, some 10 months following the stroke. The 0 gap session was carried out 2 months earlier. An examination of each individual eye movement record indicated that the eyes remained steady on the central fixation cross until target onset. The procedure of using a verbal report of the fixation cross at trial onset, was successful in providing central fixation prior to target onset.
B.Q.‘s performance on the 0 gap and overlap conditions are highly comparison confirmed no significant difference between the results sessions. Right single and double target indicators were reported at locations. B.Q. reported the left single target indicators on only 8811%
similar. A Chi-square of these two testing all eccentricities and oftrials. The reported
R.
1027
WALKEK VI
NEGLECT
JOHN M. FINDLAY,*?
AND HEMIANOPIA
ANDREW W. YOUNG*
and JOHN WELCH:
*Department of Psychology, University of Durham, Durham DHI 3LE, U.K.; and :Regional Neuroscience Centre, Department of Clinical Psychology, Newcastle General Hospital, Westgate Road, Newcastle upon Tyne NE4 6BE, U.K. (Rrcriwd
22 Fdmmry
I99 I : ucwpted2 Maq’ I99 I )
Abstract-In this paper we report findings which question the diagnosis of a hemianopia in B.Q., a ho-year-old lady who shows unilateral spatial neglect following a lesion to the right parietal lobe. The presence of a hemianopia has been indicated following two independent assessments of B.Q.‘s visual fields. We examined B.Q.‘s performance on a visuo-spatial task in which single or double stimuli were displayed left and right of a central fixation point. B.Q. failed to report left stimuli when the fixation point was continuously displayed (OVERLAP CONDITION). This performance is consistent with the suggestion of a left hemianopia as indicated by perimetric field testing. In a further condition the fixation point was extinguished prior to stimuli onset (GAP CONDITION). B.Q. consistently responded to left stimuli in the gap condition and also showed improved performance to right stimuli. Eye movements were recorded on a separate testing session, in which B.Q. showed a normal saccadic response to left targets in the gap condition, but not during the overlap condition. These results suggest that B.Q.‘s failure to report left stimuli during field plotting is due to her neglecting left stimuli and not because she has a visual field defect. By using a testing procedure which reduces the severity of neglect B.Q. is able to respond to left stimuli. Reducing the severity of B.Q.‘s neglect also abolishes the apparent hemianopia. Standard field plotting techniques which use a consistently displayed fixation point may not be appropriate for the testing of neglect patients’ visual fields.
INTRODUCTION
[JNILATERAI.spatial neglect is a condition in which patients fail to respond to stimuli in the side of space opposite to the lesion site. Neglect is typically associated with damage to the right parietal lobe resulting in left side neglect. On neurological examination many neglect patients are diagnosed as having a homonymous hemianopia, with or without macular sparing. The visual field defect is not generally thought to be the cause of neglect as the patient is free to make compensatory head or eye movements to orient in the neglected direction. Such compensatory strategies are found in patients who have a homonymous hemianopia, without neglect 19, 121. Conversely, it is possible to find patients without visual held defects who are shown to have neglect symptoms [S, lo]. Neglect is therefore thought to have an attentional 16, 151, or representational [2], underlying cause and any hemianopia is regarded as a subsidiary finding. There is an immediate problem concerning a diagnosis of hemianopia in neglect patients. Standard visual field testing involves maintaining fixation while stimuli are presented at various locations in the visual field. Failure to report stimuli is considered to reflect a field defect such as hemianopia. However unilateral spatial neglect by definition involves a failure
tTo whom reprint
requests
should
be addressed. 1019
1020
R. WALK~K t’t al.
to respond to stimuli on one side of space. In effect standard tests of hemianopia may also constitute tests for unilateral neglect. Is it possible to tell whether failure to report stimuli on one side of space is due to a field defect, or to neglect? This paper describes a study of a patient B.Q. who has profound unilateral spatial neglect. She has also been considered to have a hemianopia using perimetry techniques. The issue of hemianopia and neglect are examined on a visuo-spatial orienting task and by an examination of eye movements during such a task. It is shown that what appears as a profound hemianopia under certain testing conditions can be modified under slightly different conditions to a much milder extinction like disorder.
CASE HISTORY B.Q.. a 66-year-old retired professional woman was admitted to hospital in August 1989, with a suspected stroke. Neurological examination revealed normal cranial nerves and fundi, hemianopia, deviation of the eyes and tongue IO the left. and left-sided Facial weakness. A CT scan revealed an extensive, low denstty lesion surrounding the right parictal region. involving the cortical grey matter and deep white matter. Thcrc was moderate mid-hnc displacement. The appearances were consistent with an infarct in the territory of the right middle ccrcbral artcry. Psychological examination in September 1989 showed no significant impairment of verbal recall or immediate memory span (digit span 7 forward and 4 backwards). Verbal comprehension. vocabulary naming and remote memory were all intact. B.Q. was found to have severe unilateral spatial neglect (line bisection 40 50% shift from midpoint). Visual copying tasks and visual short term memory were severely impaired. Lcttercancellation tasks and reading tasks also showed severe impairment. The block design subtest of the Weschlcr was impossible. Occupational therapists indicated problems with body image and spatial relations. and described her as having dressing dyspraxia. Visual field testing was carried out a second time in July 1990, using screen perimetry. The results agam indicated a dense left hemianopia, with prcscrvation at 5 (macular sparing). Crossing out tasks and line bisection tasks carried out at this time indicate that her neglect condition has remained static. Near visual acuity (Curry and Paxton “Curpax” test) was shown to be good (approximately 6;4). Contraht sensitivity was in the normal range (Vistcch VCTS 6000).
METHOD Stimuli were generated by a BBC Mater scrics microcomputer (Acorn Computers Ltd), by uaing the Mode I character set, and were dlsplayed on a VDU monitor (PhIllIps Monitor X0). Horizonlal eye movements wcrc recorded using an infrared binocular eye tracker (ACS model EM 130). Analog records were recorded on a Tinhcrg scrics 100 tape recorder; one channel recorded eye movements, while a second stored a marker signal which indicated trial start and finish.
Over a period of months B.Q. was tested on three diKerent versions of the task. On one further session eye movements were recorded. The procedure involved maintaining fixation on a cross, displayed centrally on a VDlJ screen. On each trial. either a single target was then dlsplayed in the left or right visual field, or two targets wcrc displayed simultaneously in both visual fields. B.Q. was asked to look at the lixation cross and then report an Indicator digit which appeared at the target location(s). Responses given provided a measure ofspatial neglect, and of the visual field capacity. B.Q. was seated centrally in front ofthe VDU screen, mounted at eye Icvcl, with B viewing distance of SO cm. Head movements were not constrained, except that a chin rest was used when eye movements were recorded. The length of test sessions vdried, since testing wan restricted by constraints such as, fatigue during testing session, other therapy being received, etc. The aims of recording eye movements were to show: (I ) If central fixation was achieved at trial onset. (2) If saccadc latencies were affected by target type and cxpcrimen~al condillon. (3) If any saccade5 occurred to left single targets, even though the indicator digit was not rcportcd. The main difference bctwccn the conditlon5 conccrncd the activity at the fixation point as the tnrgct wa\
HLMIANOPIA
OK NEGLLCT’!
1021
presented. In the “overlap” condition the fixation cross stayed on throughout the trial sequence and targets appeared 1000 msec after initial fixation onset. In the “0 gap” condition the fixation cross was displayed for X00 mscc. Fixation cross o&et was simultaneous with target onset. In the “+ 100 gap” condition the fixation cross was displayed for 1000 msec and was extinguished 100 msec before the target was presented, leaving a blank screen on to which targets were displayed. Targets were presented in blocks under one of the three experimental conditions. Each trial consisted of one of two possible types of display sequence. namely single target trials, or double target trials. There were equal numbers of single and double trials in each experimental block.
This condition provided an objective measure of the severity of neglect under gdp and overlap conditions. Target eccentricities wcrc 2.5,s and 7.5 (measured horizontally from a vertical centre line), on the horizontal and principal oblique axes. This produced nine left visual field and nine right visual held target locations. The single target appeared at one of these 1X possible locations in a pseudorandom order. On eye movement recording sessions, targets appeared on the horizontal axis only, at 4 and 8” left and right of the fixation cross. The horizontal axis alone was used to enable accurate interpretation of saccade records.
Double targets were used to obtain a measure of extinction. Targets appeared simultaneously at equal eccentricities in the left and right visual fields, on either the horizontal axis, or diametrically opposed on the principal oblique axes. In the eye movement recording trials they appeared at equal eccentricities on the horizontal axis only.
The timing sequence used for all testing sessions is shown in Fig. 1. Each trial began with a blank screen. A fixation cross (0.57’ ) then appeared in the centre of the screen. B.Q. was instructed to attend to this fixation point and respond verbally to it by saying “CROSS” when it was detected. This aimed to ensure that central fixation was achieved at the beginning ofeach trial. On trials where the cross was not reported any response made subsequently was not included in the analysis. B.Q. had little difficulty in following these instructions and less than 5% of trials were eliminated on this criterion. The target was a uniformly illuminated square of side 0.57 ‘, presented for 100 msec. A delay of 100 msec then occurred during which time it was expected that an eye movement would occur. An indicator digit was then displayed at the target position (two different indicators were presented on double target trials). The indicator digits wcrc selected from the digits I 4 using a pseudorandom sequence. They were displayed for 300 msec following which time they were replaced by a mask (the same square constituting the target). B.Q. was required to report any indicator digit detected and the experimenter keyed the responses into the computer. There was a delay of 500 msec between each trial.
RESULTS Figure 2 shows the percentage of indicator digits reported by B.Q., on sessions when eye movements were not recorded. The + 100 gap and overlap testing sessions were both carried out on the same day, some 10 months following the stroke. The 0 gap session was carried out 2 months earlier. An examination of each individual eye movement record indicated that the eyes remained steady on the central fixation cross until target onset. The procedure of using a verbal report of the fixation cross at trial onset, was successful in providing central fixation prior to target onset.
B.Q.‘s performance on the 0 gap and overlap conditions are highly comparison confirmed no significant difference between the results sessions. Right single and double target indicators were reported at locations. B.Q. reported the left single target indicators on only 8811%
similar. A Chi-square of these two testing all eccentricities and oftrials. The reported
R.
1027
WALKEK VI
