NOTE
IPSILATERAL NEGLECT: REVERSAL OF BIAS OR
EXAGGERATED CROSS-OVER PHENOMENON?
John C. Adair1, Anjan Chatterjee2, Ronald L. Schwartz3 and Kenneth M. Heilman3
(1Department of Neurology, University of New Mexico Health Sciences Center, and the
Neurology Service of the Albuquerque VA Medical Center; 2Department of Neurology,
University of Alabama in Birmingham; 3Department of Neurology, University of Florida,
and the Neurology Service of the Gainesville VA Medical Center)

ABSTRACT
When right brain injury produces contralesional neglect (CN), patients typically
misbisect lines to the right. However, others demonstrate so-called “ipsilateral neglect” (IN)
with misbisection to the left of midpoint. Paradoxically, most patients with CN also
demonstrate a ‘cross-over’ phenomenon whereby they misbisect short lines to the left. It is
not known whether patients with IN actually have a contralesional bias opposite the
ipsilesional bias observed with CN, or if their performance reflects an exaggerated crossover. These alternatives can be distinguished by power function analysis which evaluates
the relationship between magnitude of perception and stimulus magnitude. Using line
bisection tasks to derive a power function, an IN patient showed a reduced exponent
(β=0.841), falling outside 95% confidence intervals (CI) for controls but within the CI for
CN patients. The IN patient showed a greatly increased constant (K=7.82), extending outside
the CI for both controls and CN patients. The results suggest that the anomalous leftward
misbisection with IN is associated with an exaggerated cross-over point and not simply
reversal of spatial bias.
Key words: ipsilateral neglect, power function analysis
INTRODUCTION
Kwon and Heilman (1991) reported a patient who demonstrated leftward deviation on
line bisection following a right frontal stroke. He initially showed contralesional neglect,
including rightward bias on line bisection and target cancellation tasks. However, after
several weeks of recovery, the patient demonstrated what they termed “ipsilateral neglect”,
bisecting lines to the left of midpoint, and failure to inhibit saccades to contralesional stimuli
(“visual grasp”). Butter, Rapcsak, Watson et al.(l988) observed similar ocular behavior
following right frontal stroke, explaining their findings on the basis of insufficient control of
collicular impulses driving visual attention to the left. Uncertain that such an account
pertained to line bisection performance, Kwon and Heilman (1991) speculated that ipsilateral
neglect (IN) might alternatively result from disruption of frontally-mediated “avoidance”
behavior, with consequent release of “approach” behavior directed contralesionally.
The observation that line bisection error is influenced by the length of the stimulus line
may be relevant to IN. As lines get longer, patients with contralesional neglect (CN) appear
to misapprehend larger proportions of the stimulus (Bisiach, Bulgarelli, Sterzi, et al., 1983;
Butter, Mark and Heilman, 1988). Marshall and Halligan (1989) noted that patients who
demonstrated CN when bisecting standard size lines may also reverse the side of their
bisection error for smaller lines, a phenomenon they termed “cross-over”. Perhaps what
Kwon and Heilman witnessed was actually an example of exaggerated cross-over.
Chatterjee, Mennemeier and Heilman (1994a) recently framed the relationship between
Cortex, (1998) 34, 147-153

148

John C. Adair and Others

an individual’s estimation of line length and actual line length in terms of a power function.
Following psychophysical experiments conducted by Stevens (1958), they applied line
bisection tasks as a means of inferring a subject’s perception of stimulus magnitude.
Assuming that the subject marked the line at the perceived midpoint, they calculated the
subject’s estimation of line length as twice the distance from the right end of the line to the
mark. Measuring bisection error across a series of lines with different length, they found
that the data closely fit a power function, C = K fb, in which C represents the
psychological magnitude estimate and f represents the objective physical length of the line.
Power functions with an exponent (b) of approximately 1 and a constant (K) of
approximately 1 account for a large amount of the variance observed in line bisections of
normal subjects while patients with CN usually demonstrate a reduced exponent and an
increased constant (Chatterjee, Dajani and Gage, 1994b). The diminished exponent
descriptively captures a pattern of performance whereby, for a given increase with line
length, patients with neglect place bisection marks progressively further to the right of true
midpoint. When a constant greater than 1 is obtained, as is the case in most instances of
CN, patients will cross over at shorter line lengths and misbisect to the left (For detailed
discussions, see Chatterjee et al., l994b; Chatterjee, 1995).
While power function analysis has never been applied to patients demonstrating IN,
several possible outcomes might be predicted. First, a patient with IN might produce the
same direction and degree of error regardless of stimulus magnitude. Such a patient would
produce misbisected lines on the same contralesional side of midline for all line lengths,
never demonstrating the cross-over phenomenon. Next, it is also conceivable that an
individual with IN would demonstrate a mirror reversed spatial bias that follows a power
function. In this case, patients (with right hemisphere lesions) would bisect longer lines to
the left of midline but bisect shorter lines to the right of midline, retaining a cross-over
point of similar magnitude to those with contralateral neglect. Finally, instead of showing a
qualitative difference in their line bisection performance (i.e. reversal), perhaps patients
with IN demonstrate a quantitative difference compared to patients with CN such that their
cross-over point is dramatically exaggerated. Such a scenario predicts that patients with IN
from right hemisphere lesions would misbisect of very long lines to the right of center and
standard sized lines to the left of true midpoint. Figure 1 depicts graphically the relationship
between stimulus amplitude and bisection error in normal subjects (line 1), in patients with
CN (line 2), and for each of these 3 alternative situations in patients with IN (lines 3-5).
Determining which pattern of bisection applies to IN might provide an initial clue to
understanding what appears to be a puzzling observation.
We recently evaluated a patient 3 years after right hemisphere stroke. In screening tests,
he misbisected lines to the left of actual midline, a performance consistent with IN. By
analyzing bisection error on a series of lines of various lengths, we tested which of the
above relationships best characterize his performance.
CASE REPORT
A 51 year-old man presented to the Gainesville VA Medical Center in July 1994.
Pertinent history included a right hemisphere infarct sustained in 1992. Unfortunately,
records from this hospitalization were not available. The current physical examination
demonstrated spastic hemiparesis involving the left face, arm, and leg. Cranial nerve
assessment also suggested a left temporal visual field deficit to single stimuli but extraocular
movements, including saccade and pursuit, were normal. There was no deviation of the
eyes or head. The patient was completely alert and fully aware of his neurologic deficits.
Language function was normal, including reading of regular words and non-word letter
strings. The patient canceled all targets on the Albert test (Albert, 1973) and copied line
drawings accurately. Bilateral simultaneous stimulation failed to produce extinction in any
modality. In contrast, bisection of 20 cm lines demonstrated systematic deviation to the left
(mean –4.22 mm) which exceeded the bias observed in normal individuals (95% confidence
interval –0.91 - –3.27).
Cranial CT scan revealed a large area of encephalomalacia in the distribution of the

Ipsilateral neglect

149

3
1
4

Log Perceived Line Length

5
2

A

B

Log Actual Line Length
Fig. 1 – Graphic depiction of the mathematical relationship (power function) between actual
stimulus magnitude and perception of stimulus magnitude. (1) Normal performance-exponent and
constant approximately 1. (2) Left (contralesional) neglect-exponent <1 and constant >1 with crossover point at line length ‘A’. (3) Ipsilateral neglect with constant error-exponent approximately 1 and
constant >1 with no cross-over point. (4) Ipsilateral neglect with reversed bias-exponent >1 and
constant <1 with same cross-over point as (2). (5) Ipsilateral neglect with “exaggerated” cross-overexponent <1 and constant >> 1 with cross-over at line length ‘B’.

right middle cerebral artery (Figure 2). Analysis of the scan through plotting onto standard
templates (Damasio and Damasio, 1989) demonstrated that the lesion involved the
dorsolateral frontal cortex with lesser involvement of the antero-inferior parietal lobe.
MATERIALS AND METHODS
Control Subjects
The patient’s performance on line bisection was compared to a control group. Control
subjects (N = 5) were right-handed by self-report and denied history of neurological or
psychiatric disorder. Members of the control group were age- and gender-matched to the
patient.

150

John C. Adair and Others

Fig. 2 – Cranial CT scan demonstrates large region of decreased attenuation involving the right
frontal and parietal regions.

Line Bisections
The patient and control subjects bisected lines that were 2 mm wide and 14, 17, 20, 23,
and 30 cm long. Stimuli were centered on a sheet of legal-sized paper (216 mm × 366
mm). Lines were placed horizontally on a table surface such that they were parallel to the
body’s coronal plane and their midpoint was aligned with the body’s midsagittal plane.
Trials consisted of blocks of 5 different line lengths presented in random order and subjects
completed a total of eight trials at each length (40 total lines). Subjects were instructed to
mark each line at the midpoint as judged by visual inspection.
Data Analysis
The distance from the right end of each line to the subject’s mark was measured with
calipers to the nearest 0.5 mm. The dependent variable was defined as twice this distance.
Therefore, if a subject accurately bisects the line, this variable equals the line length. For

TABLE I

Power Function Values: Patient versus Controls and Previously Reported Contralateral Neglect Patients
Subject

Exponent

Constant

r2

Patient
Control 1
Control 2
Control 3
Control 4
Control 5
CI control
CI neglect*

0.841
0.983
1.014
0.995
0.971
1.024
0.977-1.017
0.693-0.897

7.820
1.176
1.121
1.080
1.461
1.207
1.078-1.340
1.076-1.802

0.962
0.991
0.988
0.981
0.987
0.986

CI: 95% confidence intervals.
* Data obtained from Chatterjee et al., 1994a, and Chatterjee et al., 1994b.

Ipsilateral neglect

151

subjects with IN, the variable exceeds the line length and for subjects with CN, the variable
is less than the line length. The independent variable was the actual length of stimulus
lines. Both measures were logarithmically transformed to allow regression analysis from
which the power function was derived as follows. The equation for the regression line (log
[dependent variable] = β log [independent variable] + Constant) is mathematically
equivalent to a power function (dependent variable = Constant [independent variable]b; see
Chatterjee et al., 1994b, for derivation). Hence, the slope of the regression line is identical
to the exponent of the power function while the constant of the power function can be
calculated from the y-axis intercept.
From the control subjects’ performance, 95% confidence intervals were calculated for
both the exponent and constant. In addition, 95% confidence intervals were derived for
both variables from data reported previously on line bisection performance in patients with
right hemisphere injury, most of whom demonstrated contralateral neglect (Chatterjee et al.,
1994b).
RESULTS
Line bisection performance for both the patient and control subjects could be accurately
described by a power function (Table I). The control subjects had a mean exponent of 0.997
(SEM 0.01) and a mean constant of 1.209 (SEM 0.07), values nearly identical to previously
published observations [Chatterjee et al., 1994b]. Similarly, the regression equation
generated from logarithmically transformed data accounted for a substantial proportion of
the variance in control performance (mean 98.ó%).
In contrast, the patient’s exponent was reduced at 0.841. The value of the patient’s
exponent fell outside the 95% confidence interval for control subject performance.
Furthermore, the value of his exponent was similar to those reported in other patients with
left (contralesional) hemispatial neglect.
The patient’s constant was increased at 7.82. Again, this figure extended far beyond the
95% confidence interval for control subjects. In addition, the value of the patient’s constant
was greatly inflated even when compared to the performance of other patients with left
hemispatial neglect secondary to right hemisphere injury.
As depicted in Figure 1, our patient’s performance corresponds to the situation in which
the y-axis intercept of the regression line is greater than and the slope of the regression line
is similar to “typical” patients with left (contralesional) hemispatial neglect (line 5).

DISCUSSION
Patients with left hemispatial neglect misbisect lines to the right of true midpoint.
Different investigators have interpreted this phenomenon several ways. One theory
emphasizes attention-perceptual factors, suggesting that patients with neglect distort or
undervalue the leftward line segment (Harvey, Milner and Roberts, 1995; Milner, Harvey,
Roberts et al., l989). Some authors suggest that deviated bisections could result from
deficiency of moving in or toward the left space (Bisiach, Geminiani, Berti et al., 1990;
Coslett, Bowers, Fitzpatrick et al., 1990; Tegner and Levander, 1991) while others speculate
that the deviated bisection relates to increased attraction to right-sided stimuli (De Renzi,
Gentilini, Faglioni et al., 1989; Kinsbourne, 1987; Mark, Kooistra and Heilman, 1988).
None of these accounts easily accommodates the observation of patients with bisection
errors ipsilateral to the lesion.
Kwon and Heilman (1991) noted that right frontal lobe lesions can induce what they
have termed ipsilateral neglect. When attempting line bisections, rather than bisecting lines
to the right as would patients with neglect from right parietal injury, their patient bisected
the line to the left of true midpoint. The patient we now report demonstrated a similar
phenomenon. There are at least two mechanisms that could account for these findings. First,
according to Denny-Brown and Chambers (1958), the frontal lobes are critical in mediating

152

John C. Adair and Others

avoidance behaviors while the parietal lobes mediate approach behaviors. Therefore, a
lesion of the right frontal lobe may have “released” the ipsilateral parietal lobes approach
behavior and biased attention toward left-sided stimuli. Second, patients with right frontal
lesions demonstrate a “visual grasp” whereby they make obligatory saccades to
contralesional stimuli. To explain this ocular approach behavior, it has been posited that the
frontal brain tonically inhibits the superior colliculus, allowing saccadic eye movements
under voluntary control (Wurtz and Albano, 1976). With frontal injury, abnormal
involuntary saccades might occur because of collicular disinhibition. Evidence of such a
mechanism for approach behavior has been documented in experimental paradigms with
animals (Sprague, 1966) as well as from clinical observation in humans (Pierrot-Deseilligny,
Rosa, Masmoudi et al., 1991).
While parietal and collicular approach hypotheses can explain some aspects of the IN
phenomenon, neither concept can entirely account for why our patient showed less error
with longer lines. If approach behavior were the only factor influencing his bisection
performance, we would have predicted that as stimulus magnitude increased as lines became
longer, the impact of the approach response would have been stronger and the percent error
would have increased. The observation that our patient demonstrated less neglect with
longer lines suggests that there may be additional conflicting influences on behavior.
The dorsolateral frontal lobe, together with the parietal lobe, cingulate gyrus, and the
reticular system have been proposed to be components of a system that mediates attention
to contralateral stimuli (Mesulam, 1990; Watson, Valenstein and Heilman, 1981). Studies
of patients with CN from injury in this modular system indicates that they have two core
deficits, an ipsilesional attentional bias and a reduction of global attentional capacity
(Chatterjee, Mennemeier and Heilman, 1992; Robertson, 1993). Perhaps longer lines
demand more attentional resources than shorter lines. If so, then reduction of attentional
capacity rnight therefore disproportionately impact bisection of stimuli of greater magnitude
(i.e. Longer Lines). Release of approach behavior toward contralesional space may
supersede ipsilesional attention bias only when stimulus parameters place minimal demands
on attentional capacity (i.e. shorter lines), although the release may be at a representational
level (Chatterjee, 1995).
Our patient shares certain clinical features with Kwon and Heilman’s description of
“ipsilateral neglect”. We think it is likely that this patient would have demonstrated CN
early in the course of recovery. However, he demonstrated leftward line bisection errors in
the chronic phase of his injury, a finding which begs for an explanation given his right
hemisphere injury.
Even though our patient resembles previously published cases, analysis of his line
bisection performance suggests that the term “ipsilateral neglect” may be a misnomer. Using
methods described by Chatterjee and associates (Chatterjee et al., 1994a, 1994b), we used
performance on line bisections for different line lengths to establish quantitative measures,
the constant and the exponent, which characterize the patient’s spatial bias. Results suggest
that the patient exhibits an exaggerated cross-over point since their power function had a
reduced constant and greatly increased exponent. The psychophysical power law applies to
subjective magnitude estimates of a variety of perceptual continua (Stevens, 1958). Previous
studies reported that exponents greater than 1, as observed in perception of painful stimuli,
indicate that for each increment of the physical stimulus, the psychological perception
increases at an accelerated rate. With exponents less than 1, as observed in CN, there is a
compressed appreciation of objective increases in stimulus magnitude. In other words, the
data indicate that our patient was quantitatively different (i.e. increased constant) but
qualitatively similar (i.e. similar exponent) to patients with CN. These observations reinforce
the complexity of interactions of spatial attention and spatial representation along the
horizontal axis, and suggest that “release” phenomenon toward contralesional space may, in
some circumstances, subvert a more basic bias away from this space.
Acknowledgment. This work was supported by the Research Service at the VA Medical
Center, Albuquerque, New Mexico, and the VA Medical Center, Gainesville, Florida.

Ipsilateral neglect

153

REFERENCES
ALBERT, M.D. A simple test of visual neglect. Neurology, 23: 658-664, 1973.
BISIACH, E., BULGARELLI, C., STERZI, R., and VALLAR, G., Line bisection and cognitive plasticity of
unilateral neglect of space. Brain and Cognition 2: 32-38, 1983.
BISIACH, E., GEMINIANI, G., BERTI, A., and RUSCONI, M.L. Perceptual and premotor factors of unilateral
neglect. Neurology, 40: 1278-1281, 1990.
BUTTER, C.M., MARK, V.W., and HEILMAN, K.M. An experimental analysis of factors underlying neglect
in line bisection. Journal of Neurology, Neurosurgery, and Psychiatry, 51: 1581-1583, 1988.
BUTTER, C.M., RAPCSAK, S.Z., WATSON, R.T., and HEILMAN, K.M. Changes in sensory inattention,
directional akinesia, and release of the frontal fixation reflex following a unilateral frontal lesion: a
case report. Neuropsychologia, 26: 533-545, 1988.
CHATTERJEE, A. Cross-over, completion and confabulation in unilateral spatial neglect. Brain, 118: 455465, 1995.
CHATTERJEE, A., DAJANI, B.M. and GAGE, R.J. Psychophysical constraints on behavior in unilateral
spatial neglect. Neuropsychiatry, Neuropsychology and Behavioral Neurology, 7: 267-274, 1994a.
CHATTERJEE, A., MENNEMEIER, M., and HEILMAN, K.M. Search patterns in neglect. Neuropsychologia,
30: 657-672, 1992.
CHATTERJEE, A., MENNEMEIER, M., and HEILMAN, K.M. The psychophysical power law and unilateral
spatial neglect. Brain and Cognition, 25: 92-107, 1994b.
COSLETT, H.B., BOWERS, D., FITZPATRICK, E., HAWS, B., and HEILMAN, K.M. Directional hypokinesia
and hemispatial inattention in neglect. Brain, 113: 475-486, 1990.
DAMASIO, H., and DAMASIO, A.R. Lesion Analysis in Neuropsychology. New York: Oxford University
Press, 1989.
DENNY-BROWN, D., and CHAMBERS, R.A. The parietal lobe and behavior. Research Publications
Association for Research in Nervous and Mental Disease, 36: 35-117, 1958.
DE RENZI, E., GENTILINI, M., FAGLIONI, P., and BARBIERI, C. Attentional shift towards the rightmost
stimuli in patients with left visual neglect. Cortex, 25: 231-237, 1989.
HARVEY, M., MILNER, A.D., and ROBERTS, R.C. An investigation of hemispatial neglect using the
landmark task. Brain and Cognition, 27: 59-78, 1995.
KINSBOURNE, M. Mechanisms of unilateral neglect. In M. Jeannerod (Ed.), Neurophysiological and
Neuropsychological Aspects of Spatial Neglect. Amsterdam: Elsevier, 1987.
KWON, S.E., and HEILMAN, K.M. Ipsilateral neglect in a patient following a unilateral frontal lesion.
Neurology, 41: 2001-2004, 1991.
MARK, V.W., KOOISTRA, C.A., and HEILMAN, K.M. Hemispatial neglect affected by non-neglected
stimuli. Neurology, 38: 1207-1211, 1988.
MARSHALL, J.C., and HALLIGAN, P.W. When right goes left: an investigation of line bisection in a case
of visual neglect. Cortex, 25: 503-515, 1989.
MESULAM, M.-M. Large-scale neurocognitive networks and distributed processing for attention,
language, and memory. Annals of Neurology, 28: 597-613, 1990.
MILNER, A.D., HARVEY, M., ROBERTS, R.C., and FORSTER, S.V. Line bisection errors in visual neglect:
misguided action or size distortion? Neuropsychologia, 31: 39-49, 1993.
PIERROT-DESEILLIGNY, C., ROSA, A., MASMOUDI, K, AND RIVAUD, S. Saccade deficits after a unilateral
lesion affecting the superior colliculus. Journal of Neurology, Neurosurgery, and Psychiatry, 54:
1106-1109, 1991.
ROBERTSON, I.H. The relationship between lateralised and non-lateralised attention in unilateral neglect.
In I.H. Robertson and J.C. Marshall (Eds.), Unilateral Neglect: Clinical and Experimental Studies.
Hillsdale, N.J.: Lawrence Erlbaum Associates, Inc., 1993, pp. 257-275.
SPRAGUE, J.M. Interaction of cortex and superior colliculus in mediation of visually guided behavior in
the cat. Science, 153: 1544-1547, 1966.
STEVENS, S.S. Measurement and man. Science, 127: 383-389, 1958
TEGNER, R., and LEVANDER, M. Through a looking glass: a new technique to demonstrate directional
hypokinesia in unilateral neglect. Brain, 114: 1943-1951, 1991.
WATSON, R.T., VALENSTEIN, E., and HEILMAN, K.M. Thalamic neglect: the possible role of the medial
thalamus and nucleus reticularis thalami in behavior. Archives of Neurology, 38: 501-507, 1981.
WURTZ, R.H., and ALBANO, J.E. Visuo-motor function of the primate superior colliculus. Annual Review
of Neuroscience, 39: 766-772, 1976.
John C. Adair, Department of Neurology, ACC 2nd Floor, University of New Mexico Medical Center 2211 Lomas Blvd. NE
Albuquerque, New Mexico 82131 USA. E-mail:Adair.John@Albuquerque. VA. Gov