Journal of the Neurological Sciences 316 (2012) 184–188

Contents lists available at SciVerse ScienceDirect

Journal of the Neurological Sciences
journal homepage: www.elsevier.com/locate/jns

Short communication

Measuring palinopsia: Characteristics of a persevering visual sensation
from cerebral pathology
S. Van der Stigchel a,⁎, T.C.W. Nijboer a, D.P. Bergsma b, J.J.S. Barton c, C.L.E. Paffen a
a
b
c

Experimental Psychology, Helmholtz Institute, Utrecht University, Utrecht, The Netherlands
UMC Radboud, Nijmegen, The Netherlands
Departments of Medicine (Neurology), and Ophthalmology and Visual Sciences, University of British Columbia, Vancouver, Canada

a r t i c l e

i n f o

Article history:
Received 4 November 2011
Received in revised form 24 December 2011
Accepted 6 January 2012
Available online 27 January 2012
Keywords:
Palinopsia
Afterimages
Visual perseveration

a b s t r a c t
Palinopsia is an abnormal perseverative visual phenomenon, whose relation to normal afterimages is unknown. We measured palinoptic positive visual afterimages in a patient with a cerebral lesion. Positive afterimages were confined to the left inferior quadrant, which allowed a comparison between afterimages in the
intact and the affected part of his visual field. Results showed that negative afterimages in the affected quadrant were no different from those in the unaffected quadrant. The positive afterimage in his affected field,
however, differed both qualitatively and quantitatively from normal afterimages, being weaker but much
more persistent, and displaced from the location of the inducing stimulus. These findings reveal distinctions
between pathological afterimages of cerebral origin and physiological afterimages of retinal origin.
© 2012 Elsevier B.V. Open access under the Elsevier OA license.

1. Introduction
Palinopsia (or ‘visual perseveration’) is a symptom in which images of a visual stimulus persist or recur after its physical disappearance. In the immediate type the stimulus is immediately followed by
a persistent positive or negative afterimage which gradually fades
after a period of time longer than that for normal afterimages [1]. In
the delayed type an afterimage of a previously seen object reappears
after an interval of minutes to hours, sometimes repeatedly for days
or even weeks [2]. There are many possible causes for palinopsia,
such as drug intoxication, metabolic and psychiatric conditions [3].
In rare situations, palinopsia is caused by a cerebral lesion, usually
in the right-hemisphere [1,4,5]. The anatomic correlate of palinopsia
is not clear: both occipito-parietal [1,6–8] and occipito-temporal lesions have been implicated [5,9–11], Palinopsia with cerebral lesions
is often [2,5] but not always [1,12] associated with visual field defects.
In spite of numerous case reports [1,5–11], the characteristics of
the afterimage and the conditions that evoke it are unknown. There
have been only two studies in which the characteristics of the afterimage have been investigated [2,13]. In both these studies, the cerebral lesions underlying the palinopsia were not clearly revealed by
imaging. Kinsbourne and Warrington [2] concluded that the characteristics of the palinoptic afterimage were similar to the normal
after-images experienced by healthy subjects, differing quantitatively
rather than qualitatively.
⁎ Corresponding author at: Experimental Psychology, Helmholtz Institute, Heidelberglaan 2, 3584 CS Utrecht, The Netherlands. Tel.: + 31 30 253 3356; fax: + 31 30
253 4511.
E-mail address: S.VanderStigchel@uu.nl (S. Van der Stigchel).
0022-510X © 2012 Elsevier B.V. Open access under the Elsevier OA license.
doi:10.1016/j.jns.2012.01.006

In the present report, we describe the properties of palinopsia in a
patient with a known cerebral lesion. Unlike the previously studied patients, in whom palinopsia occurred throughout the visual field [2,13],
our patient's immediate-type palinoptic afterimages were confined to
the left inferior quadrant. This allowed a unique comparison between
afterimages in the intact and the affected part of his visual field.
In a first set of experiments, we investigated the properties of his
negative afterimages in the affected quadrant, in terms of intensity,
the strength of the adapting stimulus needed to produce them, and
their duration, and compared the results to those of an unaffected
quadrant. In a second set of experiments, we quantified his positive
palinoptic images along similar lines, measuring the intensity, the
strength of the adapting stimulus, and the duration of the afterimage.
2. Case report
At the moment of testing, CQ was a 64 years old right-handed man
with no history of psychiatric disorders nor substance abuse. He had
had an intra-cerebral hemorrhage five months before testing. At the
time he was examined in our laboratory for the first time, he showed
normal language, memory and executive function (Table 1). There
were no signs of neglect or deficits in visual perception.
Magnetic resonance images of his brain showed extensive focal
cerebral gliosis. In the left hemisphere there was mainly damage to
white matter in the lateral occipito-temporal region, possibly in the
region of area V5 and lateral occipital cortex (LOC). On the right
there was cortical damage involving V1 at the occipital pole with
white matter extension to the ventricle and underlying early extrastriate V2 and V3 regions (Fig. 1). In the right hemisphere, there is
also a thalamo-capsular lesion, likely unrelated to the palinopsia.

S. Van der Stigchel et al. / Journal of the Neurological Sciences 316 (2012) 184–188

185

Fig. 1. Axial images of the lesion of CQ (T1 weighted MRI). The left side of each slice represents the right side of the brain, and the images proceed from ventral slices on the left to
dorsal slices on the right.

When shown an Amsler grid, he reported that elements in the left
inferior quadrant were distorted and blurry with either eye viewing.
Goldmann perimetry showed an incongruous relative defect in his left
inferior quadrant, more prominent in the left eye (Fig. 2). He reported
an afterimage in this quadrant, which, like the afterimages he had in
daily life, he described as being a copy of the original image, instead of
a negative afterimage (i.e. with contrast and colors reversed in polarity).
CQ explained that the afterimages in his left inferior quadrant
were always present and that he frequently had to close his eyes to
remove them when their persistence bothered him. A particularly
vivid example was that, when driving through a tunnel, the lights in
the tunnel evoked a trail of persisting afterimages (creating a form
of phenomenological overlap between polyopia and palinopsia). The
afterimages made reading difficult because the words previously fixated would persist. However, CQ was always confident that he could
distinguish between real images and afterimages.
Informed consent was obtained prior to the study in accordance
with the guidelines of the Helsinki Declaration.
In all experiments described below for both negative and positive afterimages, CQ had to fixate a central fixation point. Eye movements
were monitored with an Eyelink1000 video eye-tracking system
(www.sr-research.com) and showed that CQ maintained excellent fixation. Pilot experiments confirmed that CQ only experienced abnormal
afterimages with stimuli in the left inferior quadrant. We therefore
chose to contrast the right superior (unaffected) and left inferior (affected) quadrants.

All experiments were performed in a dimly lit room, with lighting
constant across Experiments 1–7. In Experiments 1–5, the background of the monitor was black (b0.1 cd/m 2). Experiments 6 and 7
used more eccentric stimuli projected onto a blank wall.
In all experiments, trials were blocked by location. In Experiments
1–5, stimuli were presented at 135° counterclockwise from vertical
(left inferior) or 45° clockwise from vertical (right superior), at an eccentricity of 6.3° of visual angle. In Experiments 6 and 7, stimulus location was adjusted according to CQ's experience and instructions.
3. Negative afterimages (Experiments 1 to 3)
A negative afterimage is the phenomenon where exposure to a visual stimulus leads to an afterimage of opposite polarity (e.g. perceiving an illusory black spot after exposure to a white spot). Such
afterimages are normal, and are believed to arise at the level of the
retina [e.g. [14]]. To exclude the possibility that CQ's abnormal afterimages were mediated by abnormalities in his negative afterimage,
we started by assessing his negative afterimages.
Experiment 1. This experiment measured the intensity of his negative
afterimage using a nulling procedure [15]. CQ viewed for 12 s an adaptor that was a white spot with an intensity following a Gaussian distribution, with a standard deviation of 1.6° of visual angle, and a peak
luminance of 12.4 cd/m 2. After an inter-stimulus interval of 1 s, a
probe stimulus that was another white spot of the same size was

Fig. 2. Goldmann perimetry of visual fields of the patient. Perimetry showed an incongruous relative defect in his left inferior quadrant.

186

S. Van der Stigchel et al. / Journal of the Neurological Sciences 316 (2012) 184–188

shown for 2.5 s, accompanied at its onset by an alerting auditory beep.
CQ's task was to indicate whether he perceived this white probe spot.
The luminance of the probe spot was varied using a 1-up, 1-down
adaptive staircase method [see [16]] that converged after eight reversals at the Point of Subjective Equality (the nulling point). The luminance of the probe spot was changed by a factor of 1.4 until the
third reversal, and by a factor of 1.1 until the eight reversal, the
point at which the staircase terminated. The initial peak luminance
of the probe was 12.4 cd/m 2. The stronger the negative afterimage
generated by the adapting spot, the brighter the probe spot would
have to be to ‘null’ the afterimage. Hence the point of subjective
equality is a measure of the strength of the afterimage.
Experiment 2. This experiment measured the duration of exposure
needed to create an afterimage. For this we used the same nulling
concept as above. CQ again had to indicate whether the probe spot
was perceived, but now the procedure varied the duration of the
adaptor using a 1-up/1-down adaptive staircase, rather than the luminance of the probe spot. The duration of the adaptor was changed
by a factor of 1.4 until the third reversal, and by a factor of 1.1 until
the eight reversal, the point at which the staircase terminated. The
initial duration of the adaptor was 1 s. In this experiment, the peak luminance of the adaptor was kept constant at 30.0 cd/m 2, and that of
the probe spot kept constant at 23.9 cd/m 2 (these values were
based on pilot experiments on author CP). The point at which the
staircase converged indicates the duration of the adaptor at which
the afterimage nulled the white probe spot 50% of the time. This experiment reveals the time the adaptor needs to be present to produce
an afterimage.
Experiment 3. This simply measured the duration of his negative afterimage. CQ was instructed to press a button when the black afterimage (occurring after adaptation to the white spot) was gone. Thus no
probe spot was involved. The adaptor spot had a peak luminance of
58.3 cd/m 2. We measured five durations for each of the two locations.
4. Results
Experiment 1 showed that the intensity needed to null the afterimage was 12.5 cd/m 2 (s.d. 0.9 cd/m 2) in the unaffected right superior quadrant and 10.2 cd/m 2 (s.d. 2.0 cd/m 2) in the affected left
inferior quadrant. A paired sample t-test revealed that the strength
of CQ's negative afterimage did not differ between the affected and
unaffected quadrants (t(3) = 1.6, p = 0.2).
Experiment 2 showed that the adaptor needed to be present for
3.7 s (s.d. 0.9 s) to null a white probe spot in the right superior quadrant, and for 3.0 s (s.d. 0.3 s) in the left inferior quadrant. Again, a
paired t-test showed that there was no difference between these
two quadrants (t(3) = 2.3, p = 0.1).
Experiment 3 revealed that the mean duration of CQ's negative afterimages was 11.3 s (s.d. 2.9 s) in the right superior quadrant and
15.9 s (s.d. 4.0 s) in the left inferior quadrant. The difference between
the results of the two quadrants was not significant (t(4) = −1.9,
p = 0.1).
5. Positive afterimages (Experiments 4–7)
Experiments 1–3 show that CQ's negative afterimages at the affected region of the visual field were no different in strength
(Experiments 1–2) and duration (Experiment 3) from those at his unaffected region. In the experiments below, we assessed CQ's positive
afterimages parametrically. To be able to distinguish positive from
negative afterimages, we chose to use a green adapting stimulus, so
it would be easy for him to label his perceived afterimage. A nulling
procedure as we used to assess his negative afterimage could not be
used, because CQ stated that his positive afterimages were displaced

with respect to the stimulated location. Therefore, we devised other
experimental methods to investigate CQ's positive afterimage.
Experiment 4. Whereas Experiment 1 measured the intensity of the
negative afterimage, in this experiment we assessed the intensity of
the positive afterimage. CQ matched the appearance of the afterimage
with that of a probe stimulus presented at fixation. The probe stimulus was presented in either the left inferior or right superior quadrants, at the same locations used in Experiment 1–3. The probe
presented at fixation was a green Gaussian spot (standard deviation
of 1.6 deg, luminance = 4.1 cd/m 2, u′ = 0.1, v′ = 0.6 in CIE 1976
color space) of the same size as the adaptor. As in Experiment 1, an
adaptor spot was presented for 12 s (with same luminosity and hue
as the initial value of the probe spot). One second later, the probe
stimulus appeared at fixation, accompanied by a beep. CQ's task was
to indicate whether the probe was brighter than the afterimage produced by the adaptor. The intensity of the probe presented at fixation
was varied using a 1 up/1 down staircase. The luminance of the probe
was changed by a factor of 1.5 until the third reversal, and by a factor
of 1.1 until the eight reversal, the point at which the staircase terminated. The point at which the staircase converged indicates the intensity of the probe stimulus at fixation that matched the intensity of
CQ's afterimage.
Experiment 5. This measured how intense the adaptor had to be to
produce a positive afterimage. The size of the spots was also identical
to those used in Experiments 1–4. In this experiment we varied between trials the luminance of a green adaptor, again using a 1 up/1
down adaptive staircase. The luminance of the adaptor was changed
by a factor of 1.4 until the third reversal, and by a factor of 1.1 until
the eight reversal, the point at which the staircase terminated. The
initial luminance of the adaptor was 4.1 cd/m 2 (u′ = 0.1, v′ = 0.6 in
CIE 1976 color space). The adaptor's duration was kept constant at
12 s. After an interval of 1 s following the disappearance of the adaptor, an auditory beep sounded, signaling to CQ that he had to indicate
whether or not he saw a green afterimage. The point at which the
staircase converged indicates the adaptor luminance at which CQ
reported a positive aftereffect in 50% of the trials.
Experiment 6. CQ reported that the palinopsia in daily life was stronger the more eccentric the stimuli were in his left inferior quadrant.
Therefore, we performed a variant of Experiment 3, assessing the duration of the afterimage, using an image projected onto a wall at a distance of 220 cm from CQ, rather than displayed on a computer
monitor. First, we asked CQ to indicate the location where he believed
his afterimages to be maximal, which was at an eccentricity of 14° of
visual angle, along the diagonal in the left inferior quadrant, or 135°
counterclockwise from vertical. Next, we assessed the duration of
his positive afterimage at this location. For two trials, we presented
an adaptor that was a green Gaussian spot, with increased size (standard deviation of 2.5° of visual angle), peak luminance 8.41 cd/m 2,
(u' = 0.169, v' = 5.43) for 12s. The luminance of the background
was 4.0 cd/m 2. CQ's task was to report via a key press when the positive afterimage had disappeared.
Experiment 7. As CQ reported that his abnormal afterimages were
not located at the same position as the original image, we mapped
the spatial offset of his afterimage. For this, we presented an adaptor
for 12 s at an eccentricity of either 14.4° or 18.9° of visual angle. In the
three trials for each location, CQ had to indicate where the afterimage
was perceived, by verbally guiding a pencil held by the experimenter,
who marked on the wall where the after image was reported.
6. Results
Experiment 4. CQ did not perceive a positive afterimage in the
right superior quadrant in either this or the following experiments.

S. Van der Stigchel et al. / Journal of the Neurological Sciences 316 (2012) 184–188

Cognitive domain

Neuropsychological test

Performance Percentile/decile

Language

Boston naming test
Verbal fluency
A
N

175/180

93rd percentile

21

90–95th
percentile
95th percentile
72nd percentile

Working memory
Memory

Executive
functions
Visual perception

Neglect

Semantic fluency
Token test
Digit span
RAVT immediate
RAVT delay
RAVT recognition
Rey delay

21
38
21/21
11
3/6/12/9/12
9
30/30
14.5

BADS-NL rule shift cards

4

Judgment of line
orientation
Facial recognition test

27/30

72nd percentile

45/54

33–59th
percentile

Corvist
Symbol acuity
Shape discrimination
Size discrimination
Shape detection
Hue detection
Scattered dot counting
Fragmented numbers
Face perception
Crowding
Rey copy
Star cancelation

33/36
4/4
2/2
4/4
4/4
4/4
4/4
8/8
4/4
33.25
54/54

75th percentile
9th decile
8th decile
No abnormalities
35–50th
percentile

50th percentile

For the left inferior location, a probe stimulus at fixation of 0.41 cd/
m2 (s.d. 0.1 cd/m2) matched that of the positive eccentric afterimage
perceived by CQ.
Experiment 5. The luminance of green needed to generate a positive aftereffect in the left inferior quadrant was 1.0 cd/m 2 (s.d. 0.9 cd/
m 2). Thus, a light of very small intensity is enough to produce a positive afterimage.
Experiment 6. The duration of CQ's positive afterimage at the left
inferior location was 121 s (s.d. 3.5 s).
Experiment 7. When the adapting stimulus was presented at an
eccentricity of 14.4° of visual angle along the diagonal meridian,
thus corresponding to 10.2° left and 10.2° down from fixation, the
positive afterimage occurred 13.7° (s.d. 1.5) left and 20.5° (s.d. 0.1)
down from fixation (Fig. 3). An adapting stimulus at an eccentricity
of 18.9° of visual angle. 13.3° left and 13.3° below fixation, produced
an afterimage at 21.5° (s.d. 0.7) left and 20.3° (s.d. 1.7) below fixation.
These results confirm a displacement of the afterimage to a more eccentric location in the affected quadrant.
7. Discussion

about 10% of the brightness of the stimulus inducing it. This contrasts
with the negative afterimages in his affected and unaffected quadrants, which were almost as strong as the stimuli inducing these. Importantly, his positive palinoptic afterimages lasted for over two
minutes, 7 to 10 times as long as his negative afterimages, and were
systematically displaced from the inducing location, in distinct contrast to the highly retinotopic nature of normal negative afterimages.
Palinopsia has previously been explained by damaged connections
that prevent inhibition of signals to neural systems subserving vision
[2]. This inhibition is necessary to remove a stimulus from visual
awareness once it is no longer present in the visual environment.
Kinsbourne and Warrington [2] stated that release from this inhibition results in overactivity of the visual system, which causes the observed positive visual phenomenon. The spatial displacement of the
afterimage from the inducing stimulus location in our case suggests
that on occasion this loss of inhibitory connections may result in
overactivity of neighboring neurons in the visual cortex. The existence of these inhibitory signals has also been proposed in explanations of saccadic suppression, in which the brain cuts off the
processing of blurry images during an eye movement [17].
There are two possible mechanisms for the lack of inhibitory signals
in the visual cortex. First, the neural areas responsible for sending inhibitory signals to the visual cortex may be damaged, resulting in the sustained visual experience of the stimulus. Second, the relative field defect
at the location of the afterimage may prevent or reduce transmission of
the visual information required to generate the inhibitory signals. Regardless, the location of the brain lesions in CQ appears to implicate
lower visual areas as responsible for these inhibitory mechanisms.
The current study aimed to quantify the visual afterimage in a patient with palinopsia caused by brain damage. So far, most studies
have merely described the phenomenon without providing quantitative details about the afterimage. Our results concur in part with previous conclusions that palinoptic afterimages show quantitative
differences with normal afterimages [2,13]: CQ's palinoptic afterimages were inducible by weaker stimuli, fainter, but lasted longer
than his own normal negative afterimages. However, besides being
positive rather than negative afterimages, they also showed

Location of positive afterimage
25
20
15
10
5

y (deg)

Table 1
Performance of CQ on standard neuropsychological tests (5 months post-stroke).

187

0
-5

We studied the properties of palinopsia due to a known cerebral
lesion, in which positive afterimages were confined to a region in
the lower left quadrant. The results of Experiments 1–3 show that
the negative afterimages in CQ's affected quadrant were no different
from those in his unaffected quadrant. The experience of an
immediate-type palinopsia is therefore not necessarily accompanied
by distortion of normal adaptive phenomena, in particular when the
palinoptic images are positive rather than negative afterimages. The
aim of Experiment 4–7 was to provide objective measures of CQ's
positive palinoptic afterimages, concerning their intensity, duration
and location. The results show that the palinoptic afterimage could
be evoked by an adaptor of relatively low intensity (1.0 cd/m 2). CQ
was able to match the intensity of his positive afterimage with a
real physical stimulus, showing that his palinoptic afterimage was

-10
-15
-20
-25
-25

-20

-15

-10

-5

0

5

10

15

20

25

x (deg)
Fig. 3. Perceived locations of CQ's positive afterimage in visual angle. The + indicates the
location of the fixation cross; the dashed line the limits of the projected background.
Adapting probes were presented at the location of the red and green circle. Crosses indicate the location of the perceived afterimages (red for the red adaptor location, blue for
the blue adaptor location). The small crosses indicate the perceived locations at individual
trials, the large crosses and error bars the average and standard deviations respectively.

188

S. Van der Stigchel et al. / Journal of the Neurological Sciences 316 (2012) 184–188

qualitative differences from normal phenomena, particularly in the
systematic spatial displacement of the afterimage in relation to the
inducing stimulus. Hence these results indicate distinctions between
pathologic afterimages of cerebral origin and physiologic afterimages
of retinal origin.
Conflict of interest
There is no conflict of interest.
Acknowledgments
This research was funded by two grants from NWO (Netherlands organization for Scientific Research): grant 451-09-019 to SvdS and grant
451-10-013 to TCWN. JB was supported by a Canada Research Chair.
References
[1] Bender MB, Feldman M, Sobin AJ. Palinopsia Brain 1968;91:321–38.
[2] Kinsbourne M, Warrington E. A study of visual perseveration. J Neurol Neurosurg
Psychiatry 1963;26:468–75.
[3] Abert B, Ilsen PF. Palinopsia. Optometry 2010;81:394–404.

[4] Jacobs L, Feldman M, Bender MB. The persistence of visual and auditory percepts
as symptoms of irritative lesions of the cerebrum of man. Z Neurol 1972;203.
[5] Meadows JC, Munro SS. Palinopsia. J Neurol Neurosurg Psychiatry 1977;40:5–8.
[6] Critchley M. Types of visual persevaration: ‘palinopsia’ and ‘illusory visual spread’.
Brain 1951;74:267–99.
[7] Cleland P, Saunders M, Rosser R. An usual case of visual perseveration. J Neurol
Neurosurg Psychiatry 1981;44:262–3.
[8] Hayashi R, Shimizu S, Watanabe R, Katsumata Y, Mimura M. Palinopsia and perilesional hyperperfusion following subcortical hemorrhage. Acta Neurol Scand
2002;105:228–31.
[9] Michel EM, Troost BT. Palinopsia: cerebral localization with computed tomography.
Neurology 1980;30:887–9.
[10] Lazaro RP. Palinopsia: rare but ominous symptom of cerebral dysfunction. Neurosurgery 1983;13:310–3.
[11] Werring DJ, Marsden CD. Visual hallucinations and palinopsia due to an occipital
lobe tuberculoma. J Neurol Neurosurg Psychiatry 1999;66:684.
[12] Ritsema ME, Murphy MA. Palinopsia from posterior visual pathway lesions without
visual field defects. J Neuroophthalmol 2007;2:115–7.
[13] Blythe IM, Bromley JM, Ruddock KH, Kennard C, Traub M. A study of systematic
visual perseveration involving central mechanisms. Brain 1986;109:661–75.
[14] Loomis JM. The photopigment bleaching hypothesis of complementary after-images:
a psychophysical test. Vis Res 1972;12:1587–94.
[15] Kelly DH, Martinez-Uriegas E. Measurements of chromatic and achromatic afterimages. J Opt Soc Am 1993;10:29–37.
[16] Levitt H. Transformed up–down methods in psychoacoustics. J Acoust Soc Am
1971;49(Suppl. 2) 467+.
[17] Dodge R. Visual perception during eye movement. Psychol Rev 1900;7:454–65.