Investigators have known for decades that size perception is altered in patients with neglect ( Gainott i & Tiacci, 1971 ; Milner, Harvey, Roberts, & Forster, 1993 ; Halligan & Marshall, 1991 ; Bisiach, Bulgarelli, Sterzi, & Vallar, 1983 ). Th is has been explained as a magnifying eff ect of focused att ention on size perception and as unspecifi ed “perceptual factors” that work to alter size estimation across a hori- zontal axis of space.
Magnitude estimation is at least one, if not the primary, perceptual factor accounting for altered size perception in neglect. Although, bias in magnitude estimation is exaggerated in neglect, it is not specifi c to patients with neglect or dependent on spatial att ention. Both cerebral hemispheres calculate estimates of stimulus magnitude; however, factors such as arousal may produce diff erent estimates of stimulus intensity that result in diff erent behavioral abnormalities.
Finally, altered perception of stimulus intensity may be a trans-disorder process that can help explain comorbidities in neglect. We now turn to the evidence for these statements in studies of stroke patients with neglect.
M A G N I T U D E E S T I M A T I O N I S A LT E R E D A C R O S S S PA T I A L A N D N O N S PA T I A L P R O T H E T I C P E R C E P T U A L C O N T I N U A F O L L O W I N G U N I L A T E R A L S T R O K E
Prothetic refers to “how much” of something is perceived. A series of experiments in our laboratory examined how unilateral stroke alters estimates of stimulus magnitude. Results for several of these studies are summarized here for the fi rst time. We reasoned that if magnitude estimation operates independently of spatial att ention, representation, or movement, then it should be altered across prothetic perceptual continua regardless of whether they have inherent spatial qualities.
Magnitude estimates for 12 perceptual continua spanning six sensory modalities were examined in four groups of subjects ( Table 7.1 ). Normal con- trol subjects (NCS) included a total sample of 79 participants stratifi ed by age (young = 19–34; middle = 35–65; and older = 65–80 years); however, stroke patients were statistically compared only with NCS between the ages of 35 and 80 (n=39). Aft er screening stroke patients for sensory thresholds and the cogni- tive ability to perform magnitude estimation tests, the patient sample (n=34) included 13 participants with a unilateral right hemisphere lesion (RHL) due to stroke who did not have neglect (RHL – ) as defi ned by the Behavioral Inatt ention Test ( Wilson, Cockburn, & Halligan, 1987 ), nine RHL participants
www.Ebook777.com
who did have neglect (RHL + ), and 12 participants with a unilateral left hemi- sphere lesion (LHL) who did not have neglect. All stroke participants were tested at least one month aft er lesion onset. All four comparison groups were statistically equivalent with regard to gender, race, handedness, age, and educa- tion composition. General cognitive function was assessed with standardized Table 7.1 Magnitude Estimate Ranges for 12 Perceptual Continua Spanning Six Sensory Domains in Four Groups of Subjects
Sensory Domain
Perceptual Continua
Stimulus Description Ranges of Stimulus Magnitudes Visual Line length Line lengths on a 94- × -
555 cm sheet
10, 32, 60, 115, 170, 280, 390, 500 cm Area Squares on a 240- × -
240 cm sheet
100; 299; 605; 1,537;
2,894; 6,889; 12,589;
20,000 cm2 Numerosity Dots on a 265- × -330 cm
sheet
6, 10, 15, 22, 36, 50, 71, 90 dots
Refl ectance Chips of light refl ectance material 75 × 125 cm
84.2, 63.6, 46.8, 33, 22.19, 13.7, 7.7, 3.8%
light refl ectance Tactile Pressure (R & L) Pressure cuff applied to
forearm
20, 37, 54, 71, 88, 105, 122, 139 mmHg Von Frei (R & L) Monofi laments applied to
forefi nger
0.6, 2, 6, 10, 26, 60, 100, 180 grams of force Roughness
(R & L)
Textures of sand paper applied to fi ngertips
300; 700; 1,100; 1,280;
1,350; 1,400; 1,440;
1,476 grit Proprio-ceptive Finger span
(R & L)
Wooden spacers between thumb and forefi nger
4, 8, 14, 24, 34, 44, 54, 63 cm
Th ermal Temperature (R & L)
Heated disk (3.2 × 3.2 cm) presented to forearm
36, 38, 40, 42, 44, 46, 48, 50 ° F
Auditory Loudness (R & L)
Tones presented to one ear- 1000Hz
35, 42, 49, 56, 63, 70, 77, 84 dB
Gustatory Sweetness Sugar concentrations diluted in water
0.15, 0.30, 0.60, 0.70, 0.80, 0.90, 1.05, 1.20 M Saltiness Salt concentrations diluted
in water
0.19, 0.25, 0.32, 0.46, 0.60, 0.74, 0.87, 1.00 M
www.Ebook777.com
tests of IQ ( Axelrod, 2002 ), memory (Wechsler Memory Scale-Revised), lan- guage (Western Aphasia Batt ery), and executive functions. For each perceptual continuum, subjects rated the intensity of eight stimuli presented three times in random order, using numbers between 10 and 99 (low and high, respectively) without a modulus. Absolute and diff erence thresholds were tested for each con- tinuum to ensure that subjects could both perceive and distinguish among all stimulus intensities presented and data were excluded if subjects did not meet criteria for sensory thresholds. Power function exponents, constants, and ( r 2 ) values were derived for each subject in each continuum. Data were excluded if the r2 value was non-signifi cant; indicating they did not fi t a power function;
this occurred in only a few subjects. Brain lesions were analyzed using subtrac- tion techniques and volume analysis (Karnath , Fruhmann Berger, Kuker, &
Rorden, 2004 ; Golay, Schnider, & Ptak, 2008 ) with the MRIcro and MRIcron soft ware programs.
Th e following observations can be made concerning the results of the series of experiments involving normal control subjects and unilateral stroke subjects:
1. Power function parameters in normal control subjects (young, middle and older age groups) were not altered by age or gender. African American sub- jects had lower r 2 values (or greater variability in magnitude estimation) than non-African Americans. ( Figure 7.4 ). Neither IQ nor education related to the size of exponents or constants. In general, lower scores on cognitive tests were associated with lower r 2 values.
2. Power function parameters were signifi cantly diff erent between normal con- trol subjects and the patient groups. A signifi cant between-group eff ect was found for the power function exponent (F(3,69) = 7, p <.0001). Planned con- trasts showed that patients with right hemisphere lesions had lower exponents than normal control subjects and that the RHL+ group had lower exponents than RHL- group. A signifi cant between group eff ect was also found for the constant (F(3,69) = 6.4, p =.01). Patients with right hemisphere lesions had signifi cantly higher constants than did normal control subjects; however, the RHL+ group did not diff er from the RHL- group. Th ere was no diff er- ence between the LHL and NCS groups with regard to the exponent and constant. Finally, a signifi cant between-group eff ect was found for the r 2 val- ues (F(3,69) = 22.1, p <.0001). Whereas all patient data were determined to fi t power functions prior to statistical analyses; all stroke patient groups had lower r 2 values than did the NCS group. Further, the RHL+ group had lower r2 values than any other group. ( Figure 7.5 ).
3. Right hemisphere injury altered magnitude estimation across spatial and non- spatial perceptual continua ( Figure 7.6 ). Interestingly, whereas eff ect sizes for visuo-spatial stimuli were small and moderate those for kinesthetic and some gustatory stimuli were very large. Large eff ect sizes were also observed
www.Ebook777.com
for stimuli presented to the ipsilesional as well as the contralesional side of the body.
4. Group diff erences in magnitude estimation were not simply due to cognitive impairment. Although LHL patients were impaired on tests of verbal IQ , executive function, memory, and language, compared to normal control sub- jects, they had normal power function exponents and constants. Conversely, RHL – patients had close to normal cognitive scores, even though their power function parameters were altered relative to those of both NCS and LHL patients.
5. Lesion volume and subtraction analyses were completed using a total of 17 MRI and CT images obtained from patients with right hemisphere lesions.
Lesion volume was approximately 2.5 times greater for the patients with right hemisphere lesions and neglect that for those without neglect (t=1.99, df=15, P<.06); however, no diff erence in lesion volume was associated with a decreased
2.0 10
8 6 4 2 0 1.8
1.6 1.4
exponent constant
Young Middle Older Young Middle Older
9 8 7 6 5
r2
Young Middle Older 1.2
1.0 .8 .6 .4
Figure 7.4 No alteration of power function parameters in normal control subjects by age. Box plots are shown for exponent, constant, and r 2 values across all 12 continua for all age groups. Error bars represent the highest and lowest values.
The three horizontal lines forming the boxes represent the 25th, 50th, and 75th percentiles. Median values (dark lines) for exponent, constant, and r 2 values are similar across the age groups.
www.Ebook777.com
1.80 7.0 6.0 5.0 4.0 3.0 2.0
0.90
0.70
0.50
0.30 1.40 1.00
exponent constant
R-Squrared
0.60 0.20
Controls LHL RHL– RHL+ Controls LHL RHL– RHL+
Controls LHL RHL– RHL+
Figure 7.5 Power function parameters are significantly different between subject groups. Box plots of power function parameters are shown for each subject group.
Unilateral stroke patients had lower r 2 values than controls and RHL + patients had lower r 2 values than any other group. Both RHL + and RHL – patients had lower exponent and higher constant values than normal controls and LHL patients who were not different from each other. The RHL+ group had significantly lower exponents that the RHL- group. (The details of the box plots are the same as those described in Figure 7.4 .)
1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00
Exponent Parameter for all Continua area
loudnesssugur temp, right
temp, left numer
osit y pressure, left
roughness, r ight
roughness, leftvon f rei, left
von f rei, r
ight pressur
e, right
fingerspan, left salt reflectance
fingerspan, r length ight
Corrected Cohen’s d Observed Power
Figure 7.6 Magnitude estimation is altered across spatial and nonspatial perceptual continua following right hemisphere injury. Effect sizes (Cohen’s d corrected for sample size ) and the observed experimental power, calculated for the power function exponent, to discriminate between patients with right hemisphere lesions and normal control subjects is shown for each perceptual continuum. Moderate to large effect sizes were observed across a wide variety of kinesthetic, gustatory, somatosensory, and visual judgments.
www.Ebook777.com
exponent (or an increased constant) independent of neglect status. Lesions involving Brodmann’s Area (BA) 39 in the posterior parietal and BA 38/21 in the anterior temporal cortex and the white matt er pathways beneath BA 6, 4, and 3 were more common among patients with neglect than patients without neglect. Lesions involving the putamen were more common among patients without neglect than those with neglect. ( Figure 7.7 ). With regard to the power function exponent, lesions involving the anterior limb of the internal capsule and the superior thalamic radiations were more common among patients with a decreased power function exponent than those with either a normal or a high
Neglect
Exponent Constant
A B C D
A B A B
Figure 7.7 Results of a lesion subtraction analysis for 17 patients with right hemisphere lesions. The top row of pictures (Neglect) shows subtraction of RHL- from RHL+ scans. Brighter red areas indicate lesions common to RHL+ but not RHL-. Brighter blue indicates the opposite. Figure A shows brighter red areas—one involving white matter beneath BAs 6, 4 and 3 and another involving BA 38/21 in the anterior temporal lobe – and one area of brighter blue centered on the putamen.
Crosshairs in figure B show the center of these areas, respectively. Figure C shows area of red overlap involving BA 39 with corresponding crosshair in Figure D. The bottom row of pictures (Exponents) show scans of RHL patients with normal or high power function exponents subtracted from those with decreased exponents and (Constants) of patients with normal or low constants subtracted from those with high constants. Figure A under exponent shows a brighter red area centered on the anterior limb of the internal capsule. [Similar involvement of the thalamic radiations was observed in higher slices (not shown).] Figure A under constant shows a brighter red area centered on the anterior limb and genu of the internal capsule (see color insert).
www.Ebook777.com
exponent. For the constant, lesions involving the anterior limb and genu of the internal capsule were more common in patients who had an increased constant but not in those patients who had either a normal or a decreased constant.
In summary, perception of stimulus intensity was altered across multiple percep- tual systems and on both sides of the body. Whereas left and right hemisphere injury lead to greater variability in magnitude estimation; the power function constant was increased and the exponent decreased only in association with right hemisphere injury. Neglect was associated with a combination of defi cits including an increased power function constant and an exponent that was signifi cantly lower than in right hemisphere patients without neglect. Whereas neglect was associated with large lesions aff ecting cortical structures and white matt er pathways, altered perception of stimulus intensity was associated with damage to fi ber pathways that aff ect thalamocortical, corticothalamic, and corticofugal projections.
N E G L E C T I N V O LV E S A C O M B I N A T I O N O F D E F I C I T S I N P R O T H E T I C ( N O N S PA T I A L) A N D M E T A T H E T I C (e . g . , S PA T I A L) P E R C E P T U A L C O N T I N U A
Robertson ( Robertson, 1993 ) proposed that chronic unilateral neglect may require defi cits in two types of att entional systems: one for spatial orientation and another for vigilance. A similar type of proposal can be made with regard to psy- chophysical judgment. In fact, psychophysical distinctions between judgments of “where” (metathetic) versus “how much” (prothetic) reveal conceptual prob- lems for neuropsychological studies of neglect. Whereas most theories of neglect are designed to explain where neglect occurs in space (a metathetic continuum), most studies actually measure the amount of stimuli neglected (a prothetic con- tinuum) as a dependent variable in experiments. Th is issue was addressed by Jewell ( Jewell, 2003 ), working in our laboratory, by examining whether neglect represents a combination of defi cits in estimating space and stimulus intensity.
Magnitude estimates and category scales for the pitch and loudness of 1,000 Hz tones and for the area and spatial location of visual stimuli were examined in a prospective study of 32 subjects: 12 normal control subjects, 7 LHL, 6 RHL – , and 7 RHL + . Estimates of spatial location and pitch are considered metathetic because they fi t linear functions and correspond to category scaling. In contrast, judgments of area and loudness are considered prothetic because they fi t power functions and correspond to ratio scaling ( Stevens & Galanter, 1957 ). Results of the prospective study indicated that neglect was associated with magnitude esti- mation defi cits in both metathetic and prothetic continua (space, size, and loud- ness), whereas right hemisphere injury without neglect (RHL – ) was associated only with defi cits in prothetic continua (loudness and size) ( Figure 7.8 ). Eff ect sizes (Cohen’s d ) distinguishing normal control subjects from RHL subjects
www.Ebook777.com
on the basis of loudness, area, and lateral displacement were .96, 1.25, and 1.6, respectively. All subject groups were equally adept at categorizing stimuli.
Unfortunately, however, no subject group could distinguish pitch. Right hemi- sphere lesion volume and involvement of heteromodal association cortex were positively associated with both neglect and defi cits in magnitude estimation.
A combination of defi cits in prothetic and metathetic perceptual continua may distinguish those with clinical signs of neglect from those without these signs.
T H E L E F T C E R E B R A L H E M I S P H E R E C O N T R I B U T E S T O M A G N I T U D E E S T I M A T I O N
Aphasia prevents many patients with large left hemisphere lesions from per- forming magnitude estimation studies. Consequently, it is not possible to create
2.50 Lateral Displacement Left Lateral Displacement Right
Area Judgment Loudness Judgment
Slope, lateral left, ratio Slope, lateral right, ratio
Exponent, area, 10-99 Exponent, loudness, 10-99
2.00 1.50
0.40
0.30
0.20 0.50
NPC RHL– LHL RHL+
1.00
2.50 2.00 1.50
0.50
2.80 2.40 2.00 1.60 1.20 1.00
Study group
Metathetic
Prothetic
NPC RHL– LHL RHL+
Study group
NPC RHL– LHL RHL+
Study group
NPC RHL– LHL RHL+
Study group
Figure 7.8 Deficits in both prothetic and metathetic perceptual continua are a distinguishing feature of neglect. Results of a prospective study involving 32 subjects (12 normal control subjects, 7 LHL, 6 RHL – , and 7 RHL + ) indicated that neglect was associated magnitude estimation deficits in both metathetic and prothetic continua (space, size, and loudness); in contrast, RHL – was associated only with deficits in prothetic continua (loudness and size).
www.Ebook777.com
groups of RHL and LHL subjects who are equated on the basis of lesion location and volume. Th is issue was addressed in two case studies of magnitude estimation following left hemisphere injury ( Woods et al., 2006 ).
Case 1. Patient B.G. was a 45-year-old, right-handed woman, a former reg- istered nurse, who had expressive aphasia one year following a large, posterior middle cerebral artery infarction ( Figure 7.9 ). She did not have neglect accord- ing to the Behavioral Inatt ention Test, but her brain lesion was typical of those associated with neglect following right hemisphere injury. She completed 7 of 12 perceptual continua tests by writing her responses. Her r 2 values fell below the 95 percent confi dence limits of r 2 values obtained from normal control subjects (from the above mentioned study) in six of the seven continua that she completed.
Additionally, values for her power function exponents fell below those of controls on four of the seven continua and exceeded these values on two of the seven. Sizes of the power function constants fell below those for controls on three of the seven continua and also exceeded these sizes on three of the seven.
Case 2. Patient J.W. was a 39-year-old, right-handed man with 13 years of edu- cation who had neglect acutely, without aphasia, following a following a putam- inal hemorrhage of the left hemisphere. ( Figure 7.10 ). His r 2 values were lower than those of normal control subjects in 11 of 12 perceptual continua tests. Sizes of his power function exponents fell below those of controls in 7 of 12 continua but exceeded 95 percent confi dence limits on 2 of 12 continua. Power function
Figure 7.9 A large, posterior middle cerebral artery infarction in a 45-year-old, right-handed woman with expressive aphasia 1 year post-infarction.
www.Ebook777.com
constant sizes were lower than the 95 percent confi dence limits for 6 of 12 conti- nua and exceeded those of controls on one continuum.
Th ese case studies are the fi rst, to our knowledge, to show that left hemisphere injury can alter magnitude estimation in a manner similar, but not identical, to that observed following right hemisphere injury. In particular, values of the power function exponents were both increased and decreased following LHL, rather than decreasing, which is more typical following RHL. A similar patt ern of results was observed in a group study of right and left hemisphere stroke patients that examined the relationship between perception of strength and stimulus intensity ( Taylor-Cooke et al., 2006 ) (see next section).
M A G N I T U D E E S T I M A T E S A R E O P P O S I T E F O L L O W I N G L E F T A N D R I G H T H E M I S P H E R E I N J U R I E S
Falls by patients with neglect are commonly att ributed to decreased awareness of weakness and to impulsivity ( Webster et al., 1995 ; Rapport et al., 1993 ); however, the problem of altered strength perception aft er stroke has not been systemati- cally studied. To examine this problem, we tested 13 patients with unilateral right hemisphere lesions and 6 with left hemisphere lesions while undergoing stroke rehabilitation. Only two patients in the RHL group had neglect; no patient had anosognosia. From these studies, we derived an index score of strength estimation accuracy (SEA) by comparing patients’ ratings of contralesional limb strength to those of a physician and correlating them with power function exponents derived Figure 7.10 A hemorrhagic LHL in a 39-year-old, right-handed man who had neglect acutely, without aphasia.
www.Ebook777.com
for length estimation on a bisection task. Line bisection was chosen because of previous experience with the test in stroke patients ( Mennemeier et al., 2005 ).
Strength perception of the contralateral limb was highly correlated with mag- nitude estimation for line length in both patient groups. Additionally, a subset of patients in the RHL group who had decreased power function exponents for length estimation overestimated contralateral limb strength, whereas a subset of patients in the LHL group who had increased exponents underestimated limb strength ( Figure 7.11 ).
Th e results of this study suggest that altered strength perception may be quite common aft er a stroke, even in patients without frank neglect and anosognosia.
Further, perception of strength and stimulus magnitude may be governed by a common neural mechanism. Finally, altered perception of strength aft er a stroke might help explain the abnormal reactions that patients sometimes have to stroke, such as a lack of concern shown by some patients with right hemisphere lesions ( Starkstein, Federoff , Price, Leiguarda, & Robinson, 1992 ) and an over-concern or an extreme emotional response shown by some patients with left hemisphere lesions ( Starkstein, Federoff , Price, Leiguarda, & Robinson, 1993 ). Together with functional imaging studies ( Pinel, Piazza, LeBihan, & Dehaene, 2004 ), these patient studies indicate that both cerebral hemispheres generate magnitude esti- mates of stimulus intensity; however, lateralized brain injury may disrupt magni- tude estimation in diff erent ways. For example, left and right hemisphere lesions can have diff erent aff ects on arousal ( Heilman et al., 2003 ) which could alter magnitude estimation diff erently.
1.2
1.1
Power Function Exponent
1.0
0.9
–4 –2 0 2 4 6 8 10
Strength Estimation Accuracy Scores
1.2
1.1
Power Function Exponent
1.0
0.9
–4 –2 0 2 4 6 8 10
Strength Estimation Accuracy Scores Figure 7.11 Strength estimation accuracy (SEA) scoring. A negative SEA score corresponded to underestimation of contralateral limb strength, and a positive score to overestimation, with a score of 0 indicating agreement between the estimations of patients and physician. The SEA score increased as the power function exponent for patients with right hemisphere stroke decreased (graph on left ); in contrast, the SEA score decreased as the power function exponent increased in patients with left hemisphere stroke (graph on right : from Taylor-Cooke, et al 2006, figures 1 & 2, pages 1445 & 1446 respectively.
www.Ebook777.com