children and adolescents with chronic cerebellar lesions show no clinically relevant signs of...

Cognitive deficits in cerebellar children 1

Children and adolescents with chronic cerebellar lesions

show no clinically relevant signs of aphasia or neglect

S. Richter1,5, B. Schoch2, O. Kaiser1, H. Groetschel1, C. Hein-Kropp1, M. Maschke1, A.

Dimitrova1, E. Gizewski3, W. Ziegler4, H.-O. Karnath5, D. Timmann1

1 Departments of Neurology, 2 Neurosurgery, and 3 Neuroradiology, University of

Duisburg-Essen, 4 EKN – Clinical Neuropsychology Research Group, Department of

Neuropsychology, City Hospital Bogenhausen, Munich,

5 Center of Neurology, Hertie-Institute for Clinical Brain Research, University of

Tuebingen, Germany

Correspondence to:

Dr. Stefanie Richter

Department of Neurology / University of Duisburg-Essen

Hufelandstr. 55 / 45122 Essen / Germany

phone +49 201 723 3816

fax +49 201 723 5901

email [email protected]

Articles in PresS. J Neurophysiol (July 20, 2005). doi:10.1152/jn.00611.2005

Copyright © 2005 by the American Physiological Society.

mailto:[email protected]


ABSTRACT

We studied language and visuo-spatial functions of twelve children and adolescents

who had undergone surgery for cerebellar astrocytoma without subsequent radiation or

chemotherapy and compared them with 27 age-, gender- and education-matched healthy

control subjects. In order to study possible lateralization of the functions of the left and

right cerebellar hemispheres, subjects performed several language tasks including a verb

generation task as well as standard neglect and extinction tests. 3D-MR images

confirmed that lesions affected cerebellar hemispheres in all children but one who had a

pure vermal lesion. The right cerebellar hemisphere was affected in six, the left

hemisphere in four children and both hemispheres in one child. There were no signs of

aphasia in the children or adolescents with cerebellar lesions. Language abilities did not

differ between cerebellar patients and control subjects except for small increases in

reaction times in verb generation in patients with left-sided lesions. Visuospatial

functions were also intact in cerebellar subjects except for minor group differences in

neglect tasks. In sum, chronic focal cerebellar lesions acquired in childhood or youth do

not result in persistent language disorders or clinically significant signs of spatial

neglect or extinction.

Keywords: cerebellum, language, visuo-spatial neglect, localization


INTRODUCTION

The cerebellum is alleged to be involved in non-motor functions (Schmahmann 1997;

Thach 1998). According to the ”cerebellar cognitive affective syndrome” introduced by

Schmahmann and Sherman (1998), the vermis takes the role of the cerebellar limbic

system, while the lateral hemispheres are said to be associated with executive functions,

visuospatial processing and language.

The cerebellum is massively connected to the contralateral cerebral cortex

(Middleton and Strick 2001). The right cerebellar hemisphere has been associated with

language functions in brain-imaging (Petersen et al. 1989; Raichle et al. 1994;

Ackermann et al. 2004) and human lesion studies (Fiez et al. 1992; Silveri et al. 1994).

As far as verb generation tasks are concerned, findings are inconsistent. Some studies

showed that patients with right-sided cerebellar lesions were unable to reduce their

reaction times over blocks of trials and/or produced incorrect answers (Fiez et al. 1992;

Gebhart et al. 2000) while no impairments were found in other studies (Helmuth et al.

1997; Richter et al. 2004).

Visuospatial functions are alleged to be associated with the left cerebellar

hemisphere according to brain imaging studies in healthy adults (Fink et al. 2000; Fink

et al. 2001) and patient studies (Riva and Giorgi 2000). There is also evidence for non-

lateralized visuospatial deficits in cerebellar patients (Malm et al. 1998; Schmahmann

and Sherman 1998; Neau et al. 2000; Molinari et al. 2004), but impairments of

visuospatial functions were not confirmed in other studies (Dimitrov et al. 1996; Gomez

Beldarrain et al. 1997).

Spatial neglect occurs mainly after right-sided cerebral lesions and is

characterized by an inability to explore and react to stimuli within the contra-lesional

hemifield (Mesulam 2002; Karnath et al. 2001; Karnath et al. 2003). Based on the


contralateral projections from the cerebellum to the cerebral cortex, one would expect

ipsilateral neglect in cerebellar patients with left-sided lesions. Neglect has only rarely

been assessed in cerebellar patients. Previous results are inconsistent. Neglect to the

right was found to be associated with both right and left cerebellar lesions (Silveri et al.

2001; Hildebrandt et al. 2002; Aarsen et al. 2004).

Studies by Riva and Giorgi (2000) and Scott et al. (2001) in children after

cerebellar tumor surgery suggest a lateralization of cognitive functions in the cerebellar

hemispheres in childhood. Results of the studies by Levisohn et al. (2000), Steinlin et

al. (2003) and Aarsen et al. (2004), however, found no evidence to support this

conclusion. Finally, while deficits in the Riva and Giorgi study seem to be of transient

nature, cognitive deficits were found to be more permanent in other studies (Levisohn et

al. 2000; Scott et al. 2001; Steinlin et al. 2003; Aarsen et al. 2004).

The aim of the present study was to further investigate if language functions are

associated with the right cerebellar hemisphere and visuospatial functions with the left

cerebellar hemisphere in children. Children who had previously received surgery for

cerebellar astrocytoma performed several language tasks and a series of standard tests

for spatial neglect and extinction. Impairments were studied in relation to the site and

extent of cerebellar lesion as assessed by means of individual 3D-MRI-scans.


MATERIAL AND METHODS

Subjects

Twelve children who had previously received surgery to remove a cerebellar

astrocytoma at the Department of Neurosurgery, University of Duisburg-Essen,

Germany, participated in the study (7 female, 5 male; mean age 14.6 years ± 2.8 years,

range 9 -19 years). Three of these (Cb1, Cb5, Cb10) had taken part in a previous study

of our group on speech motor deficits in children with acute cerebellar lesions (Richter

et al. 2005). Twenty-seven healthy children, 12 girls and 15 boys, served as control

subjects. Mean age was 14.4 years (± 2.8 years) in the healthy control group (range: 8-

19 years). Patients and control subjects were matched for educational level. One patient

(Cb1) was left-handed, another patient (Cb6) was right-handed, but preferred to use the

left hand after the operation. Two control subjects (Con10, Con27) were left-handed, all

other control subjects were right-handed. Healthy control subjects revealed no

neurological or psychiatric diseases according to clinical history and neurological

examination.

Children were included in the language analysis, if they were German native

speakers, at least 8 years old and at least in the 3rd grade. The latter criterion was chosen

to assure some basic level of orthographic knowledge. Results of the language tests of

one patient (Cb6) were excluded from analysis since she was not a German native

speaker. One control subject (Con21) was in the 2nd grade only. Results of the language

tests of this child were also excluded.

The cerebellar operation had been performed at least one year ago (mean 5.7

years ± 3.4 years, range 1 – 13.4 years). All patients had presented with astrocytoma.

One child revealed an astrocytoma grade II, the others astrocytomas grade I. None of


the children had received radiation and/or chemotherapy or was taking centrally acting

medication at the time of testing. The cerebellar patients showed no evidence of

extracerebellar lesions based on clinical examination and MRI-data.

Subjects underwent a standard neurological examination according to the

International Cooperative Ataxia Rating Scale (ICARS; Trouillas et al. 1997). Total

ICARS-score ranges from 0 (no ataxia) to 100 (most severe ataxia). Mean total ICARS-

score was 3.8 ± 5.0, ranging from 0 to 14.5.

Written informed consent was obtained from all children and parents, who were

recompensed for their travel expenses. The local committee of research ethics approved

the study. Basic characteristics of patients with cerebellar lesions are summarized in

Table 1.

------------------------------

Insert Table 1 around here

-------------------------------

Experimental procedure

The investigation consisted of three main parts, assessment of (1) language, (2)

visuospatial functions and (3) localization of the cerebellar lesion on the basis of

individual 3D-MRI-scans.

Language

Verb generation

Subjects sat comfortably in a chair and looked at a central fixation cross on a computer

monitor in front of them. Coloured photographs of real objects (e.g. a car) were


presented for 4 seconds with an interval of 3 seconds. Two different types of responses

were tested. In one, the naming condition, subjects were asked to name the pictured

object (e.g. “car”). In the verb generation condition, they were instructed to say what

one can do with the object, i.e. to produce the corresponding verb (“drive”). There were

2 naming blocks and 4 verb generation blocks. The order of testing was: naming block 1

– verb generation blocks 1 to 4 – naming block 2. There was a training block of 5 trials

before the first naming block and a training block of 9 trials before the first verb

generation block using objects that differed from those used in the actual experiment.

The objects shown (see Table 4) were the same in both conditions and the sequence of

objects was varied between blocks.

The time between the object presentation and verbal responses (i.e. vocal

reaction times) was measured. To this end, the responses were segmented semi-

automatically with the help of a MATLAB-routine based on loudness functions (Ruske

and Beham 1992; Merk 2002). For further details of the verb generation task and

segmentation routine see Richter et al. (2004).

We analyzed for each presented object the answers given by the subjects based

on a thesaurus of the German language (http://wortschatz.uni-leipzig.de; Table 4). In

addition to giving synonyms for nouns and verbs, the thesaurus also indicated which

verbs are typically associated with specific objects.

Verbal response onsets were excluded from statistical analysis, when 1. answers

were wrong (inappropriate verb or noun, inappropriate word category); 2. no answer

was given; 3. the answer was appropriate only after a correction of the subject. The

latter answers were not considered “wrong” based on the quality of responses.

Only a very few answers were excluded from the statistical analysis. Across the

total of 1152 trials in the patient group (11x96), 7 trials were excluded due to self-

http://wortschatz.uni-leipzig.de/


correction (0.6 %). In the 2496 trials of the control group (26x96), a total of 18 trials

(0.7 %) were excluded (self-correction: 4 trials, noun instead of verb: 7 trials, verb

instead of noun: 1 trial; no answer: 4 trials, inappropriate verb: 2 trials).

Aphasia test (Aachener Aphasietest, AAT)

The AAT is a standardized aphasia test widely used in German speaking countries

which has been validated in a large group of aphasic patients (n = 376). The

combination of two subtests of the AAT, the token test and the written language subtest,

has been shown to detect aphasic symptoms with a high sensitivity (Huber et al. 1983).

In the token test the subject is asked to point to different tokens, touch them or pick

them up according to instructions of increasing complexity. The written language

subtest includes word reading, spelling, and writing to dictation. Additionally, an

experienced speech therapist (C. H.-K.) evaluated recordings of spontaneous speech

regarding signs of aphasia or delays in speech development based on AAT rating scales.

Developmental language test (Heidelberger Sprachentwicklungstest, HSET)

The HSET is a German developmental language test for children aged three to nine

years, validated in 791 healthy children (Grimm and Schöler 1978, 1991). It is explicitly

stated by the developers of the test that HSET might be readily applied also in older

children. In the sentence construction subtest a sentence has to be constructed from two

or three words (e.g. girl – play – doll). In the plural-singular formation subtest the task

is to generate the plural form from the singular form or vice versa of real (sparrow –

sparrows) or artificial words (“Mattau” – “Mattaus”; “Plabeln” – “Plabel”).

Quantitative and qualitative speech analysis


To assess speech motor deficits which might influence the above measures of language

function, a quantitative and qualitative analysis of dysarthric speech symptoms was

performed.

In a syllable repetition task subjects were asked to repeat six different

consonant-vowel syllables involving (co-)articulations of major articulatory organs.

Subjects were advised to produce syllable chains as fast and as regularly as possible for

about 5 s after a brief demonstration by the experimenter. In a sentence production task

subjects had to repeat 12 test sentences which differed only with respect to a target word

belonging to one out of three categories. There were two-syllabic words of a simple

consonant-vowel-structure, two-syllabic words containing consonant clusters and three-

syllabic words. The sentences were read aloud by the experimenter and repeated by the

subject.

Measurement of syllable durations in target words and repeated syllables was

based on a syllable-segmentation algorithm operating on loudness contours of the

speech signal (see Merk and Ziegler 1999).

For the qualitative speech analysis the speech therapist (C. H.-K.) rated

dysarthria symptoms in recordings of spontaneous speech on a four-point-scale

according to Kluin et al. (1988) and Darley et al. (1969a,b). For more detailed

information on the quantitative and qualitative speech analysis see Merk and Ziegler

(1999) and Richter et al. (2005).

Visuospatial functions

Letter cancellation test

In the letter cancellation test by Weintraub and Mesulam (1985), thirty target letters 'A'

are distributed amid distractors both on the left and right half of a horizontally oriented


sheet of paper. Patients were asked to cancel all of the targets. The test was finished

when subjects indicated twice that they believed they had cancelled all target stimuli.

Results are presented as percentage omissions on the left and right side separately with

reference to the total number of target stimuli. In adults, omissions of more than five

left-sided targets (i.e. > 17 %) are taken to diagnose spatial neglect (Weintraub and

Mesulam 1985).

Line bisection

Six sheets of paper with one line each (length 24.1 cm, width 0.03 cm) were presented

one after the other and the subject was asked to mark with a pen the middle of the line.

The line was not positioned exactly in the middle of the sheet so that subjects could not

use the edges of the paper to orient their response. The side with the smaller margin was

presented three times each on the left and on the right to prevent adaptation. For each

trial the difference between the subject’s estimation of the line’s middle and the actual

central point was calculated, both as absolute and signed values. Deviations to the right

were arbitrarily assigned a positive sign, those to the left a negative one. Differences

were expressed as percentage of half the length of the line and averaged across the 6

trials.

Extinction

Patients with visual extinction may react normally to stimuli when presented on one

side of the visual field. However, if two stimuli are presented simultaneously, the

stimulus on the side contralateral to the cerebral lesion may not be noticed (e.g. Heilman

et al. 1997). In adults, visual extinction is diagnosed if a subject fails to perceive the left

stimulus during bilateral presentation in more than 50% of the trials although the same


stimuli are reported at least 90% correctly with unilateral stimulus presentation (Karnath

et al. 2003).

Four geometrical figures (square, circle, triangle, diamond), each about 0.7° in

size, were presented for 180ms in random order 4° left and/or right of a central fixation

point with the help of the computer program ”Mel“ (Psychology Software Tools, Inc.,

USA). There were ten trials with bilateral and 20 trials with unilateral presentations. We

observed whether subjects perceived the stimuli, and also if they recognized their form.

Results are presented as percentage of stimuli perceived/recognized on the left and right

side in bilateral compared to unilateral presentations.

MRI analysis

In cerebellar patients extent of surgical lesions was defined based on individual 3D

MRI-data sets. First, a 3D sagittal volume of the entire brain was acquired using a T1-

weighted magnetization prepared rapid acquisition gradient echo (MPRAGE) sequence

(FOV = 256 mm, number of partitions = 160, voxel size = 1 x 1 x 1 mm3, TR/TE =

2400/4.38 ms, flip angle = 8 degrees) on a Siemens Sonata 1.5 Tesla MR. In addition,

axial and sagittal 2D-T2 weighted images of the entire brain were acquired. 3D-

MPRAGE and 2D-T2 weighted images were visually examined to reveal possible

extracerebellar pathology.

Surgical lesions were manually traced on axial and sagittal slices of the non-

normalized 3D-MRI data set and saved as region of interest using MRIcro

(http://www.mricro.com; Rorden and Brett 2000). The same software was used to

calculate individual lesion volumes from non-normalized ROIs.

Individual volume of the lesion and the complete 3D-MRI data set were then

spatially normalized into a standard proportional stereotaxic space (the MNI 152-space)

http://www.mricro.com/


using SPM99 (http://www.fil.ion.ucl.ac.uk/spm/; Wellcome Department of Cognitive

Neurology, London, UK). The MPRAGE volume was registered and resampled to 2 x 2

x 2 mm3 voxel size. Because MRI atlases of the cerebellum are available for adult

brains only, lesions of all subjects regardless of age were normalized to the adult MNI

152 brain average template.

The MNI coordinates of cerebellar lesions were determined in the horizontal (x),

sagittal (y) and vertical (z) directions. Based on these coordinates the corresponding

cerebellar lobules (Schmahmann et al. 2000) and affected cerebellar nuclei were defined

(Dimitrova et al. 2002). Sagittal divisions were defined according to Luft et al. (1998).

Both lesions of the paravermis (–24mm ≤ x ≤ –10mm, +10mm ≤ x ≤ +24mm) and

lateral hemispheres (x ≤ –24mm, x ≥ +24mm) were considered hemisphere lesions.

The process of spatial normalization is likely to introduce some errors in

individual anatomy. Extent of individual normalized lesions were manually adjusted

based on non-normalized data using characteristic anatomical landmarks by an

experimenter (B.S.) blinded for behavioral data.

Degree of preoperative hydrocephalus was assessed based on diagnostic 2D-

MRI scans using the bicaudate index (Aarsen et al. 2004) and the Evans score (Evans

1942). The bicaudate index is equal to the width of the lateral ventricles on the level of

the caudate nuclei, divided by the width of the brain on the same level. A value larger

than 0.16 is considered pathological up to the age of 30 years (van Gijn et al. 1985). The

Evans score equals the maximum distance between the frontal horns of the lateral

ventricles as percentage of the maximum diameter of the parietal lobes. A percentage

larger than 30 is considered pathological.

Six patients revealed a preoperative hydrocephalus according to both indices

(Cb2, Cb3, Cb4, Cb6) or according to the bicaudate index (Cb8) or Evans score (Cb11)

http://www.fil.ion.ucl.ac.uk/spm/


alone. In two patients, preoperative hydrocephalus could not be assessed because

diagnostic 2D-MRI scans were not available (Cb5, Cb12). At the time of testing, none

of the cerebellar patients revealed a hydrocephalus.

The cerebellar lesion sites according to MRI analysis are summarized in Table 2

and illustrated in Figure 1. Cerebellar children were assigned to groups based on the

affected hemisphere. There were six children with right-sided lesions (Cb1-Cb6), three

of them involving the right dentate (Cb3, Cb5, Cb6). Four patients presented with a left-

sided lesion (Cb7-Cb10), involving the left dentate in one case (Cb10). Two children

(Cb4 and Cb10) showed small extensions of the lesion to the contralateral side which

were not further considered. One child revealed a lesion of both hemispheres involving

the right dentate nucleus (Cb11). Cb12 was the only child with a pure vermal lesion;

lesions in another eight cases included the vermis (Cb2-Cb6, Cb9-Cb11). In cases in

which the dentate was included it was typically the posterior region of the nucleus that

was affected.

------------------------------

Insert Table 2 around here

-------------------------------

------------------------------

Insert Figure 1 around here

-------------------------------

Statistical analysis

Main statistical analyses considered the performance in the three groups of children

(right-sided lesions vs. left-sided lesions vs. control subjects). Post-hoc analyses were

conducted comparing the patients’ subgroups and the group of all cerebellar children


with the control group separately (see Table 3). Significance levels in post-hoc tests

were Bonferroni-corrected for multiple comparisons (p=0.017).


RESULTS

Language

Verb generation

Reaction time in the verb generation blocks

Figure 2A shows the means and standard deviations of the reaction times [s] of each of

the 16 trials in the six subsequent blocks for the two cerebellar subgroups with right-

and left-hemisphere lesions and the control subjects. The first and the last block are

naming blocks, the second to fifth block are verb generation blocks.

In both patient groups and the control subjects reaction time was equally reduced

across the four verb generation blocks. Analysis of variance with GROUP (right-sided

lesions vs. left sided-lesions vs. control subjects) as between-subjects factor and BLOCK

as within-subject factor revealed a significant effect of BLOCK (p<0.0001), but no

significant GROUP effect (p=0.692) or BLOCK by GROUP interaction (p=0.398).

As there was a trend for the children with left-hemisphere lesions to give their

answers more slowly, post-hoc analyses of variance were conducted. Results were

similar to those in the main analysis (see Table 3 and Figure 2B).

---------------------------------------

Insert Figure 2A,B about here

---------------------------------------

---------------------------------------

Insert Table 3 about here

---------------------------------------

Reaction time in the naming blocks


Comparison of the first and last block in Figure 2A shows that reaction time in the

naming condition did not change from the first to the second naming block in the patient

and control groups. Moreover, control subjects were slightly slower than patients both

in the first and second naming block. Analysis of variance with GROUP as between-

subjects factor and BLOCK as within-subject factor revealed neither a main effect of

GROUP (p=0.787), nor a BLOCK effect (p=0.632) or BLOCK by GROUP interaction

(p=0.714).

In the post-hoc analyses similar results were obtained (Table 3).

Comparison of naming and verb generation

Reaction times were averaged in the two naming and four verb generation blocks.

Analysis of variance with TASK as within-subject factor and GROUP as between-subjects

factor revealed that reaction time was significantly longer in the verb generation

compared to the naming condition in the patients’ subgroups with right and left

cerebellar lesions and control subjects (TASK effect p<0.0001). There was no significant

GROUP effect (p=0.852). The TASK by GROUP interaction was not significant (p=0.077).

However, the small p-value reflected slightly slower average responses of the left-

hemisphere patients.

Post-hoc analyses revealed no significant interaction between TASK and

GROUP comparing right-hemisphere patients and control subjects (p=0.581) and

comparing the group of all cerebellar patients and control subjects (p=0.135). The

interaction was significant, however, with left-hemisphere patients and control subjects

(p=0.015). The remaining results were similar as in the main analysis (Table 3).

In order to determine whether, practice effects aside, the patients were

disproportionately slow on the verb generation task, additional analysis was performed


comparing the means of the two naming and the first two verb generation blocks.

Analysis of variance revealed a significant TASK (p<0.0001) and TASK by GROUP

effect (p=0.042). The GROUP effect was not significant (p=0.907).

Post-hoc analyses revealed no significant interaction between TASK and

GROUP comparing right-hemisphere patients and control subjects (p=0.254). The

interaction was significant, however, in the comparison between left-hemisphere

patients and control subjects (p=0.015). A close-to-significant interaction was found

comparing the group of all cerebellar patients and control subjects (p=0.035). The

remaining results were similar as in the main analysis (Table 3).

Quality of responses

In the naming condition, all answers were correct both in patients and control subjects

except for one wrong answer in a control subject who gave a verb instead of noun

(sweep instead of broom). In the verb generation condition, patients gave no wrong

answers. In the control subjects, a noun was given instead of a verb in seven cases

(children’s drum – music (4 trials), pocket lamp – light (2 trials), doll - game) and an

inappropriate verb was given in two cases (cooking pot – eat (2 trials)).

The most frequent answer for each object was the same for both the patients’

subgroups and control subjects (Table 4). One single verb was generated in response to

scissors, two different answers each were produced in children’s bicycle, car, mobile

phone, and children’s drum. The remaining objects evoked three or more answers, with

a maximum of twelve answers for backpack. The results were the same for the patients’

subgroups.

------------------------------------



------------------------------------

Aphasia test (Aachener Aphasietest, AAT)

Although the percentile ranks given for the AAT refer to an aphasic norm group of

adults aged 21 to 70 years and older (Huber et al. 1983), application of the AAT is

recommended in adolescents and adults aged 14 years and older (Huber et al. 1983). In

the AAT, a percentile rank larger than 90 indicates that aphasia is absent. Percentile

ranks between 60 and 90 indicate weak, between 30 and 60 medium and lower than 30

severe aphasia. The results are summarized in Table 5. In the token test, the mean

percentile rank was 96.4 ± 1.95 in the patient subgroup with right-hemisphere lesions,

96.75 ± 2.06 in the subgroup with left-hemisphere lesions and 97.27 ± 2.0 in the control

group. A univariate analysis of variance comparing children with right-sided lesions,

left sided-lesions and control subjects revealed a non-significant GROUP-effect

(p=0.634).

Regarding written language, mean percentile ranks did not differ between

patients’ subgroups and control subjects. Average percentile ranks were 98.6 ± 1.5 in

the patients with right-hemisphere lesions, 98.0 ± 1.83 in the patients with left-

hemisphere lesions and 97.73 ± 4.2 in the control subjects. Again, univariate analysis of

variance revealed a non-significant GROUP-effect (p=0.892).

T-tests comparing both patients’ subgroups and the group of all cerebellar

patients to the control group separately were not significant for both AAT-subtests (see

Table 3).

The mean ± 2 SD was calculated for the control children to define a normal

range of performance based on the present data. In Table 5 individual values are marked

in bold which fall below the normal range. In the token test none of the cerebellar


patients and one control subject (Con 20, 3.7 %) fell below the normal range. In the

written language test none of the cerebellar patients and two of the control subjects (Con

20, Con 25, 7.4 %) fell below the normal range.

Compared with German norms for the token test acquired in children younger

(Gutbrod and Michel 1986) and older than 14 years (De Renzi and Vignolo 1962;

Orgass 1982), none of our children aged eight to 14, and none of the children older than

14 revealed pathological results, except one control child (Con20).

According to Huber et al. (1983) the relation between the results of the token

test and the test of written language sheds light on the presence of aphasia. Three

control subjects (Con5, Con20, Con25) revealed such a critical constellation of results

in the two tests. However, one of these was very young (Con5: 8 years).

Finally, none of the cerebellar or control children revealed perceptible signs of

aphasia. No problems with word finding, paraphasia, neologism, agrammatism, jargon

or verbal stereotypes were observed in spontaneous speech samples.

------------------------------------


------------------------------------

Developmental language test (Heidelberger Sprachentwicklungstest, HSET)

Performance in our patients and control subjects was refered to the data of the oldest

norm group (8 to 9.9 years; n=92) (Grimm and Schöler 1978). In general, the definition

of a percentile rank as “average” is based on stanine scores. A percentile rank < 22

reflects “below average” (stanine 1-3), between 23 and 76 “average” (stanine 4-6), and

larger than 76 “above average” performance (stanine 7-9). In the present study a further


differentiation in “low” (PR 4-10, stanine 2), “below average” (PR 11-22, stanine 3),

“low average” (PR 23-39, stanine 4), and “average” (PR 40-59, stanine 5) was applied.

In the plural-singular formation task, mean percentile rank was somewhat

smaller in the patients with left-sided lesions (45.0 ± 35.8) and in the control subjects

(48.4 ± 29.2) than in the children with right-hemisphere lesions (58.4 ± 37.1). A

univariate analysis of variance with GROUP (right-sided lesions vs. left sided-lesions vs.

control subjects) as between-subjects factor revealed a non-significant effect (p=0.769).

Three out of 11 cerebellar children (27.3%) revealed a percentile rank of 14,

which was below average. Two children children had left-hemisphere lesions (Cb8,

Cb10; 50 %) and one a right-hemisphere lesions (Cb3; 16.7 %). In the control group

there were four subjects with a low (15.4 %), three children with a below average (11.5

%) and two children with a low average percentile rank (7.7 %).

In sentence construction, the group means were: right cerebellar patients: 75.4 ±

20.71; left cerebellar patients: 44.0 ± 36.56; control subjects: 70.3 ± 26.4. The lower

mean in the left-cerebellar patients was due to the fact that one patient had a low

average (Cb9), and two a below average percentile rank (Cb8, Cb10). Two control

subjects performed below average (7.7 %) and three revealed a low average percentile

rank (11.5 %). Univariate analysis of variance showed a non-significant GROUP-effect

(p=0.171).

Post-hoc comparisons (t-tests) for the plural-singular formation and sentence

construction tasks revealed similar results as in the main analysis.

Variability of results was comparatively high in both HSET subtests already in

control subjects (Table 5, Figure 3). This is further demonstrated by calculation of a

normal range based on the mean ± 2 SD in the control children. Lower normal range

resulted in percentiles <0 in plural-singular formation and <10.8 in sentence


construction. Based on the present findings in control subjects, none of the control and

cerebellar subjects revealed abnormal results in the plural-singular and the sentence

construction task (Table 4). Overall HSET subtests did not appear effective to

distinguish between control and patient groups.

------------------------------------

Insert Figure 3 about here

------------------------------------

Quantitative and qualitative speech analysis

Syllable duration was longer in the sentence production task than in the oral

diadochokinesis task in patients and control subjects (main effect of TASK: p<0.0001).

There was neither a GROUP effect (p=0.555), nor an interaction between TASK and

GROUP (p=0.795). The means in the individual groups and tasks were: sentence

production, right cerebellar patients: 193.66 ms ± 25.21 ms, left cerebellar patients:

189.91 ms ± 38.34 ms, control subjects: 204.43 ms ± 31.58 ms; syllable repetition, right

cerebellar patients: 167.76 ms ± 16.33 ms, left cerebellar patients: 163.39 ms ± 26.20

ms, control subjects: 171.19 ms ± 16.53 ms. (Post-hoc comparisons: TASK-effects, all p-

values ≤ 0.0001; GROUP-effects, all p-values ≥ 0.361; TASK by GROUP interactions, all

p-values ≥ 0.589).

None of the cerebellar or control children revealed perceptible signs of

cerebellar dysarthria. Specific articulation problems (e.g. sigmatism, nasality) were

present in three cerebellar (Cb1, Cb4, and Cb9) and in five control children (Con4,

Con10, Con19, Con23, Con24).


Visuospatial functions

Cancellation test

None of the subjects showed spatial neglect, i.e. had more than five omissions (17 %)

on the left side (see Table 6). In patients’ subgroups, average percentage of omissions

was 3.3 % ± 3.8 % on the left and 0.8 % ± 1.7 % on the right in the group with left-

sided lesions. In the subgroup of patients with right-hemisphere lesions, average

percentage of omission was 0.6 % ± 1.4 % on the left and 1.1 % ± 1.7 % on the right.

Control subjects omitted 0.5 % ± 2.0 % targets on the left and 0.7 % ± 1.7 % on the

right. Average percentage of omissions on the left was larger in the subgroup of patients

with left-sided lesions compared to patients with right-sided lesions or control subjects.

This was due to Cb8 and Cb9, who missed two A’s on the left each. A Kruskal-Wallis

H-test comparing the three groups of subjects revealed no significant differences for

omissions on the right (p=0.76) and a difference close to significance on the left

(p=0.06). Nearly significant differences between left-hemisphere patients and the other

groups for omissions on the left were also found in post-hoc comparisons (left-sided

lesions vs. control subjects p=0.02; all patients vs. control subjects p=0.04). The

remaining tests were not significant (see Table 3).

The mean ± 2 SD was calculated for the control children to define a normal

range of performance. In the letter cancellation task three of the cerebellar patients (25

%) and one of the control subjects (3.8 %) fell below the normal range (Table 6). Two

patients had a left-sided (50 %) and one patient a vermal lesion (8.3 %).

-------------------------------------------


------------------------------------


Line bisection

A univariate analysis of variance showed that absolute percentage deviation from the

center did not differ between both cerebellar patient groups and control subjects (right-

hemisphere lesions: 5.2 % ± 1.7 %, left-hemisphere lesions: 3.5 % ± 2.0 %, control

subjects 4.2 % ± 1.8 %; p=0.26).

Regarding signed deviation, Table 6 reveals that both patients with left- and

right-sided lesions tended to define the middle of the line leftward from the actual

middle (right-hemisphere lesions: -2.2 % ± 3.2 %, left-hemisphere lesions: -0.8 % ± 3.2

%), while control subjects set the middle of the line rightward from the actual middle

(0.7 % ± 3.0 %). A univariate analysis of variance with GROUP as between subjects

factor missed significance (p=0.07).

Differences between the group of patients with right-sided lesions and control

subjects were close to significance for signed, but not for absolute deviations (absolute

p=0.14, signed p=0.03). The difference between the group of all cerebellar patients

(signed: -1.7 % ± 2.8 %, absolute: 4.5 % ± 1.8 %) and the control group was not

significant for absolute (p=0.43), but significant for signed deviation (p=0.01). Patients

with left-sided lesions did not significantly differ from control subjects (absolute

p=0.30, signed p=0.56).

In the line bisection task one of the cerebellar patients (8.3 %) and one of the

control subjects (3.7 %) fell below the normal range of performance based on the results

in the control group (mean ± 2 SD) (Table 6). The patient had a right-sided lesion.

Extinction

Detection


All cerebellar and control children perceived all stimuli on either side both in trials with

one stimulus, and in the critical trials with two stimuli. Thus, no patient revealed

significant signs of extinction.

Identification

Identification of stimuli was also close to perfect (see Table 6). Percentage of correctly

identified items on the left side under unilateral and bilateral stimulus presentation was

100 % and 100 % correct in children with left-sided lesions, 100 % and 96.7 % ± 5.2 %

in children with right-sided lesions, and 100 % and 98.1 % ± 4.8 % in control children.

Percentage of correctly identified stimuli on the right side under unilateral and bilateral

stimulus presentation was 97.5 % ± 5.0 % each in children with left-sided lesions, 100

% and 95 % ± 5.5 % in children with right-sided lesions, and 99.6 % ± 1.9 % and 96.3

% ± 7.4 % in control children.

The one cerebellar child with a bilateral lesion (Cb11) did not identify three out

of 10 presentations of bilateral stimuli on the left as well as on the right side (70 %

detection on each side).

A Kruskal-Wallis H-test comparing the percentage of correctly identified stimuli

in bilateral trials in the three groups of subjects revealed no significant differences for

both stimuli on the right (p=0.60) and left (p=0.39). Moreover, post-hoc tests comparing

each patients’ subgroup and the group of all cerebellar patients to control subjects were

non-significant (see Table 3).

One of the cerebellar patients (8.7 %) and one of the control subjects (3.8 %) fell

below the normal range of performance in control children (mean ± 2 SD) (Table 6).

The patient had a bilateral lesion (Cb11).


DISCUSSION

Children and adolescents with chronic focal cerebellar lesions did not show clinically

relevant signs of aphasia, spatial neglect or visual extinction. There were minor group

differences in verb generation and in neglect tasks. No clear evidence was found for the

assumption of a lateralization of language functions to the right cerebellar hemisphere

and visuospatial function to the left cerebellar hemisphere. Whereas minor differences

in verb generation and letter cancellation were present in patients with left-sided lesions

compared to control subjects, patients with right-sided lesions tended to be impaird in

line bisection. The present results suggest that no clinically significant disorders in

language and visuospatial function are present in children and adolescents with chronic

focal cerebellar lesions.

Verb generation

In the verb generation task, there were no differences between patients and control

subjects concerning quality of responses and reduction of verbal reaction times

(learning). The results are in accordance with previous studies in adult cerebellar

patients (Helmuth et al. 1997; Richter et al. 2004). Preserved learning and correct

answers contradict other studies where learning and/or quality of answers were impaired

in adult cerebellar subjects (Fiez et al. 1992; Gebhart et al. 2000; Fabbro et al. 2000).

The results of the Fiez and Gebhart studies may differ from the present findings

for several reasons. First, the Fiez patient had an acute infarct, while our patients

revealed a chronic disease. However, the authors report that there were no marked

changes in the patient’s performance over a period of 10 months. Second, neither Fiez et

al. (n=1) nor Gebhart et al. (n=4) based their findings on a larger group of cerebellar


patients. Finally, pictures of objects were used in the present study and written nouns in

previous experiments. The association between pictures of objects and appropriate verbs

may be stronger than between written nouns and appropriate verbs. Therefore, patients

may be less impaired in the present study. However, Helmuth et al. (1997) used written

nouns and showed no impairment in verb generation in adult cerebellar patients.

In our previous study of adult cerebellar patients there was a tendency of

reaction times in the verb generation condition to be proportionately more prolonged

relative to naming in the cerebellar patients compared to control subjects. The same

tendency was found in the present study in cerebellar children with left-sided lesions. It

cannot be excluded that the effect reflects inherent difficulties in verb processing, i.e.

grammatical or semantic deficits (Cappa et al. 1998; Peran et al. 2003).

Apart from linguistic deficits, dysarthria-related problems with the motor parts

of the verb generation task have to be considered. First, the lexical semantic

representation of actions is expected to be accomplished in motor regions (Cappa et al.

2002). Next, movement imagination has generally been shown to activate similar

cerebellar regions as actual movements in brain imaging studies (Hanakawa et al. 2003).

Moreover, inner speech has been found to activate cerebellar regions (Ackermann et al.

1998; 2004). These motor processes might play a greater role in verb generation than in

naming, e.g. when subjects think about what one can do with the object or silently

search for the correct verb association. They may be the reason for the increase in

reaction times in verb generation compared to naming in patients with left-sided lesions.

This interpretation is supported by the fact that increases of verbal reaction times

were generally more pronounced in the patients with left-sided lesions. Cerebellar

dysarthria has been related to left-sided cerebellar lesions by Lechtenberg and Gilman

(1978). It has to be noted, however, that increased mean reaction time in the group of


left-sided cerebellar patients was mainly due to findings in one child (Cb9, mean verbal

reaction time 1200 ms), who showed also comparatively long syllable durations in the

syllable repetition (mean 201 ms) and sentence production tasks (242 ms). Cb9 was the

youngest child in the cerebellar group and age effects have to be considered too.

Aphasia

The token test and the written language subtest of the Aachener Aphasietest (AAT)

were used since performance in these two tests is especially sensitive to the effects of

aphasia (Huber et al. 1983). However, neither these tests nor a perceptual analysis of

spontaneous speech revealed signs of aphasia in the cerebellar group. Preserved

performance of cerebellar children in the token test is in accordance with the results of

Steinlin et al. (2003) in chronic patients after astrocytoma surgery. In contrast, deficits

were found in the token test in children with acute right cerebellar affection due to

astrocytoma surgery by Riva and Giorgi (2000). Results agree with their hypothesis that

language deficits might be of transient nature.

More specific grammatical functions were examined in two subtests of a

developmental language test (HSET). No significant group differences were found, but

there was a tendency of children with left-sided lesions to perform worse in the sentence

construction task. Although language dysfunctions were expected in patients with right-

sided lesions, abnormalities in patients with left-sided lesions do not exclude a role of

the cerebellum in language. Only in the lesion studies of Riva and Giorgi (2000) and

Scott et al. (2001) language impairments were related to right-hemisphere lesions. In the

latter, however, all children but one received cranial radiation. In the Riva and Giorgi

study children with acute cerebellar lesions were investigated.


It has to be noted, however, that specificity of individual findings in the HSET

subtests is limited. Both in the plural-singular formation and sentence construction task

variability was high already in control subjects. Moreover, in the plural singular-

formation task, 10 (38.5 %) of the control children performed below the normal test

range or presented with a low percentile rank compared to 3 (27.3 %) of the group of all

cerebellar patients. Especially the building of the plural form of artificial words appears

to be difficult for both patients and control children likely due to its abstract nature.

The observation of mainly preserved language functions seems at variance with

previous findings in children following cerebellar surgery. One reason might be that in

these studies different language tests were used, e.g. the verbal subtests of the Wechsler

intelligence scale (e.g. vocabulary, comprehension; Levisohn et al. 2000; Scott et al.

2001; Steinlin et al. 2003). Differences in pathology as well as lesion size and

localization have also to be taken into account. Whereas the present and two previous

studies (Steinlin et al. 2003; Aarsen et al. 2004) included children with astrocytoma

only, other studies examined patients with both astrocytoma and medulloblastoma (Riva

and Giorgi 2000; Levisohn 2000; Scott et al. 2001). Direct comparison of lesion

localization is difficult because 3D MR data sets were available in the present study

only. In the present study, however, patients were included with lesions affecting

primarily the posterior cerebellar lobe which is supposed to be involved in non-motor

functions (Schmahmann and Sherman 1998). Furthermore, the results of all previous

studies in children were not compared to control groups, but to test norms found in the

literature. Norms acquired in a larger group of children have the advantage that the

interindividual variability is small. However, norms are often not available for each

investigated age group and language, and it is not clear if they are applicable if a test is

performed as part of a two to four hour test battery.


In sum, no significant signs of aphasia were observed in the present group of

children and adolescents with chronic cerebellar lesions. It cannot be excluded,

however, that more subtle language deficits may be revealed in a larger group of

children with chronic cerebellar lesions and in more challenging tasks. Specific

problems in grammar function, verbal working memory, temporal aspects of both

speech production and perception have been shown in adult cerebellar patients

(Desmond et al. 1997; Zettin et al. 1997; Gasparini et al. 1999; Marien et al. 1996,

2000; Ackermann et al. 2004; Justus 2004).

Visuo-spatial functions

None of the cerebellar patients showed clinically relevant signs of spatial neglect in the

letter cancellation and line bisection tasks, or any signs of visual extinction as they are

typically observed after cerebral lesions in the right hemisphere.

In the letter cancellation task, absolute number of missed targets was small in the

patients. Not a single patient missed more than two left-sided targets. The percentage of

children missing left-sided targets, however, was larger in the group of patients with

left-sided lesions (2/4) than in the patients with right-sided lesions (2/6) and group

differences were close to significance. The result is in agreement with the assumption of

visuospatial functions lateralized to the left cerebellum. Findings, however, need to be

confirmed in a larger group of patients with left-sided lesions. Left-sided targets were

also missed by the patient with a pure vermal lesion (Cb12) and the two youngest

control children (Con5, Con21). One of the patients with left-sided lesions was only

nine years old. Age effects cannot be excluded.


Deviations in the line bisection task were small in cerebellar patients. There

were no significant differences in the absolute deviations comparing control subjects

and patients with left- or right-sided lesions. Patients with right cerebral lesions bisect

horizontal lines significantly to the right of the veridical center (Heilman et al. 1997;

Ferber and Karnath 2001). A similar finding was expected in chronic cerebellar patients

with left-sided lesions. Although this assumption was not confirmed, subtle

impairments could not be excluded in line bisection irrespective of lesion side. There

was a tendency of cerebellar patients, in particular patients with right-sided lesions, to

define the middle of the line more to the left whereas control subjects defined the

middle of the line more to the right. While neglect of the right hemispace after right-

sided cerebellar lesions contradicts the assumption of a left-cerebellar involvement in

visuospatial functions, it is in agreement with the results of previous studies (Silveri et

al. 2001).

Finally, cerebellar children perceived all stimuli in the extinction task. In the

present study it was also assessed if subjects were able to identify the stimuli. Few

errors were observed in bilateral trials with no significant group differences. Errors were

not confined to the ipsilateral space, i.e. children with right-hemisphere lesions did not

preferentially misperceive targets on the right and children with left-hemisphere lesions

targets did not misperceive more targets on the left. Performance was worst in the one

patient with a bilateral lesions (Cb11) who misperceived stimuli on the left and right

side in three bilateral trials. Cb11 presented with preoperative hydrocephalus which may

explain part of the findings.

Former group studies in children after cerebellar surgery assessed visuospatial

functions by means of several subtests of the Wechsler intelligence scale (block design,

object assembly) or copying and recall of the Rey-Osterrieth complex figure. Only in


the Aarsen study (2004) line bisection was applied. In all of these studies, deficits in

visuospatial functions were found. While two of the four studies could associate deficits

with left-hemisphere lesions (Riva and Giorgi 2000; Scott et al. 2001), in the study of

Steinlin et al. (2003), they were associated with vermal affection.

Interestingly, in the study of Aarsen et al. (2004), visuo-spatial deficits were

associated with preoperative hydrocephalus. In the present study, no significant

correlations could be observed. However, the number of children was small,

preoperative MRI scans were not available in two of the twelve children and overall

deficits were weak. Finally, effects of disordered eye movement cannot be excluded.

Although there was no significant relation to oculomotor signs in clinical testing, eye

movements have not been quantified during testing.

In sum, minor group differences in visuospatial functions were observed. While

mild differences in the letter cancellation task were found in left-sided lesions,

deviations in line bisection were more pronounced in right-sided lesions.

Conclusion

In conclusion, children and adolescents with chronic focal cerebellar lesions did not

show clinically relevant signs of aphasia, spatial neglect or visual extinction. Minor

group differences were present in verb generation and neglect tasks. Findings were not

consistent with a lateralization of language functions within the right cerebellum and

visuospatial function within the left cerebellum. Lack of clinically significant findings

need to be confirmed in a larger group of children with chronic cerebellar lesions and in

children with acute cerebellar lesions.


ACKNOWLEDGEMENT

We like to thank the participating patients and control subjects for their time and effort.

The study was supported by grants of the Deutsche Forschungsgemeinschaft (DFG Ti

239/5-1, Ti 239/5-2).


REFERENCES

Aarsen FK, Van Dongen HR, Paquier PF, Van Mourik M, and Catsman-

Berrevoets CE. Long-term sequelae in children after cerebellar astrocytoma surgery.

Neurology 62(8): 1311-1316, 2004.

Ackermann H, Wildgruber D, Daum I, and Grodd W. Does the cerebellum

contribute to cognitive aspects of speech production? A functional magnetic resonance

imaging (fMRI) study in humans. Neurosci Let 247: 187-190, 1998.

Ackermann H, Mathiak K, Ivry RB. Temporal organization of “Internal

speech” as a basis of cerebellar modulation of cognitive functions. Beh Cog Neurosci

Rev 3: 14-22, 2004.

Cappa SF, Binetti G, Pezzini A, Padovani A, Rozzini L, and Trabucchi M.

Object and action naming in Alzheimer’s disease and frontotemporal dementia.

Neurology 50: 351-355, 1998.

Cappa SF, Sandrini M, Rossini PM, Sosta K, Miniussi C. The role of the

frontal lobe in action naming. rTMS evidence. Neurology 59: 720-723, 2002.

Darley FL, Aronson AE, and Brown JR. Clusters of deviant speech

dimensions in the dysarthrias. J Speech Hear Res 12: 462-496, 1969a.

Darley FL, Aronson AE, and Brown JR. Differential diagnostic pattern of

dysarthria. J Speech Hear Res 12: 246-269, 1969b.

De Renzi E, and Vignolo LA. The token test: a sensitive test to detect receptive

disturbances in aphasics. Brain 85: 665-678, 1962.

Desmond JE, Gabrieli JD, Wagner AD, Ginier BL, and Glover GH. Lobular

patterns of cerebellar activation in verbal working-memory and finger-tapping tasks as

revealed by functional MRI. J Neurosci 17(24): 9675-9685, 1997.


Dimitrov M, Grafman J, Kosseff P, Wachs J, Always D, Higgins J, Litvan

I, Lou JS, and Hallett M. Preserved cognitive processes in cerebellar degeneration.

Behav Brain Res 79(1-2): 131-135, 1996.

Dimitrova A, Weber J, Redies C, Kindsvater K, Maschke M, Kolb FP,

Forsting M, Diener HC, and Timmann D. MRI atlas of the human cerebellar nuclei.

Neuroimage 17: 240-255, 2002.

Evans WA. An encephalographic ratio for estimating ventricular enlargement

and cerebral atrophy. Arch Neurol Psych 47: 931-937, 1942.

Fabbro F, Moretti R, and Bava A. Language impairments in patients with

cerebellar lesions. J Neuroling 13: 171-188, 2000.

Ferber S, and Karnath H-O. How to assess spatial neglect – Line bisection or

cancellation tasks? J Clin Exp Neuropsychol 23(5): 599-607, 2001.

Fiez JA, Petersen SE, Cheney MK, and Raichle ME. Impaired non-motor

learning and error detection associated with cerebellar damage. A single case study.

Brain 115 Pt 1: 155-178, 192.

Fink GR, Marshall JC, Shah NJ, Weiss PH, Halligan PW, Grosse-Ruyken

M, Ziemons K, Zilles K, and Freund HJ. Line bisection judgments implicate right

parietal cortex and cerebellum as assessed by fMRI. Neurology 54(6): 1324-1331, 2000.

Fink GR, Marshall JC, Weiss PH, and Zilles K. The neural basis of vertical

and horizontal line bisection judgements: an fMRI study of normal volunteers.

Neuroimage 14(1.2): 59-67, 2001.

Gasparini M, Di Piero V, Ciccarelli O, Cacioppo MM, Pantano P, Lenzi

GL. Linguistic impairment after right cerebellar stroke: a case report. Eur J Neurol

6(3): 353-356, 1999.


Gebhart AL, Petersen SE, and Thach WT. Evidence for functional

lateralization of language operations to the right posterior cerebellum. Soc Neurosci

Abstr 26: 1246, 2000.

Gomez Beldarrain M, Garcia-Monco JC, Quintana JM, Llorens V, and

Rodeno E. Diaschisis and neuropsychological performance after cerebellar stroke.

Europ Neurol 37(2): 82-89, 1997.

Grimm H, and Schöler H. Heidelberger Sprachentwicklungstest. Göttingen:

Hogrefe, 1991.

Gutbrod K, and Michel M. Zur klinischen Validität des Token Tests bei

hirngeschädigten Kindern mit und ohne Aphasie. Diagnostica 32: 118-128, 1986.

Hanakawa T, Immisch I, Toma K, Dimyan MA, van Gelderen P, and

Hallett M. Functional properties of brain areas associated with motor execution and

imagery. J Neurophysiol 89: 989-1002, 2003.

Heilman KM, Watson RT, and Valenstein E. Neglect: clinical and anatomical

aspects. In: Behavioral neurology and neuropsychology, edited by Feinberg TE, and

Farah MJ. New York: McGraw-Hill, 1997, p. 309-317.

Helmuth LL, Ivry RB, and Shimizu N. Preserved performance by cerebellar

patients on tests of word generation, discrimination learning, and attention. Learn Mem

3(6): 456-474, 1997.

Hildebrandt H, Spang K, and Ebke M. Visuospatial hemi-inattention

following cerebellar/brain stem bleeding. Neurocase 8(4): 323-329, 2002.

Huber W, Poeck K, Weniger D, and Willmes K. Aachener Aphasie Test.

Bern: Huber, 1983.


Justus T. The cerebellum and English grammatical morphology: evidence from

production, comprehension, and grammaticality judgments. J Cogn Neurosci 16(7):

1115-1130, 2004.

Karnath H-O, Ferber S, and Himmelbach M. Spatial awareness is a function

of the temporal not the posterior parietal lobe. Nature 411: 950-953, 2001.

Karnath H-O, Himmelbach M, and Küker W. The cortical substrate of visual

extinction. Neuroreport 14(3): 437-442, 2003.

Kluin KJ, Gilman S, Markel DS, Koeppe RA, Rosenthal G, and Junck L.

Speech disorders in olivopontocerebellar atrophy correlate with positron emission

tomography findings. Ann Neurol 23: 547-554, 1988.

Lechtenberg R, and Gilman S. Speech disorders in cerebellar disease. Ann

Neurol 3: 285-290, 1978.

Levisohn L, Cronin-Golomb A, and Schmahmann JD. Neuropsychological

consequences of cerebellar tumour resection in children – Cerebellar cognitive affective

syndrome in a paediatric population. Brain 123: 1041-1050, 2000.

Luft AR, Skalej M, Stefanou A, Klose U, and Voigt K. Comparing motion-

and imagery-related activation in the human cerebellum: a functional MRI study. Hum

Brain Map 6(2): 105-113, 1998.

Malm J, Kristensen B, Karlsson T, Carlberg B, Fagerlund M, and Olsson

T. Cognitive impairment in young adults with infratentorial infarcts. Neurology, 51,

433-440, 1998.

Marien P, Engelborghs S, Pickut BA, De Deyn PP. Aphasia following

cerebellar damage: Fact or fallacy? J Neuroling 13: 145-171, 2000

Marien P, Saerens J, Nanhoe R, Moens E, Nagels G, Pickut BA, Dierckx

RA, De Deyn PP. Cerebellar induced aphasia: Case report of cerebellar induced


prefrontal aphasic language phenomena supported by SPECT findings. J Neurol Sci

144: 34-43, 1996.

Merk M. Entwicklung und Implementierung PC-gestützter akustischer

Analyseverfahren für die klinische Diagnostik neurogener Sprechstörungen.

Dissertation, Universität der Bundeswehr München, 2002.

Merk M, and Ziegler W. MoDiaS - A PC-based system for routine acoustic

speech analysis of neurogenic speech disorders. In: Pathologies of speech and

language. Advances in clinical phonetics and linguistics, edited by Maasen B, and

Groenen P. London: Whurr, 1999, p. 315-321.

Mesulam M-M. Functional anatomy of attention and neglect: from neurons to

networks. In: The cognitive and neural bases of spatial neglect, edited by Karnath H-O,

Milner AD, and Vallar G. Oxford: Oxford University Press, 2002.

Middleton FA, and Strick PL. Cerebellar projections to the prefrontal cortex of

the primate. J Neurosci 21(2): 700-712, 2001.

Molinari M, Petrosini L, Misciagna S, and Leggio MG. Visuospatial abilities

in cerebellar disorders. J Neurol Neurosurg Psych 75(2): 235-240, 2004.

Neau JP, Arroyo-Anllo E, Bonnaud V, Ingrand P, and Gil R.

Neuropsychological disturbances in cerebellar infarcts. Acta Neurol Scand 102: 363-

370, 2000.

Orgass B. Token Test. Manual. Weinheim: Beltz Testgesellschaft, 1982.

Peran P, Rascol O, Demonet J-F, Celsis P, Nespoulous J-L, Dubois B, and

Cardebat D. Deficit of verb generation in nondemented patients with Parkinson’s

disease. Mov Dis 18(2): 150-156, 2003.


Petersen SE, Fox PT, Posner ML, Mintun M, and Raichle ME. Positron

emission tomographic studies of the processing of single words. J Cogn Neurosci 1:

153-170, 1989.

Raichle ME, Fiez JA, Videen TO, MacLeod AM, Pardo JV, Fox PT, and

Petersen SE. Practice-related changes in human brain functional anatomy during

nonmotor learning. Cereb Cortex 4(1): 8-26, 1994.

Richter S, Kaiser O, Hein-Kropp C, Dimitrova A, Gizewski E, Beck A,

Aurich V, Ziegler W, and Timmann D. Preserved verb generation in patients with

cerebellar atrophy. Neuropsychologia 42: 1235-1246, 2004.

Richter S, Schoch B, Ozimek A, Gorissen B, Hein-Kropp C, Kaiser O,

Hövel M, Wieland R, Gizewski E, Ziegler W, and Timmann D. Dysarthria is a rare

sign in children with cerebellar tumors. Brain Lang 92:153-67, 2005

Riva D, and Giorgi C. The cerebellum contributes to higher functions during

development – Evidence from a series of children surgically treated for posterior fossa

tumours. Brain 123: 1051-1061, 2000.

Rorden C, and Brett M. Stereotaxic display of brain lesions. Behav Neurol 12:

191-200, 2000.

Ruske G, and Beham M. Gehörbezogene automatische Spracherkennung. In:

Sprachliche Mensch-Maschine-Kommunikation, edited by Mangold H. München, Wien:

Oldenbourg Verlag, 1992, p. 33-47.

Schmahmann JD. The cerebellum and cognition. San Diego: Academic Press,

1997.

Schmahmann JD, Dojon J, Toga AW, Petrides M, and Evans CA. MRI Atlas

of the human cerebellum. San Diego: Academic Press, 2000.


Schmahmann JD, and Sherman JC. The cerebellar cognitive affective

syndrome. Brain 121: 561-579, 1998.

Scott RB, Stoodley CJ, Anslow P, Paul C, Stein JF, Sugden EM, and

Mitchell CD. Lateralized cognitive deficits in children following cerebellar lesions. Dev

Med Child Neurol 43: 685-691, 2001.

Silveri MC, Leggio MG, and Molinari M. The cerebellum contributes to

linguistic production: a case of agrammatic speech following a right cerebellar lesion.

Neurology 44(11): 2047-2050, 1994.

Silveri MC, Misciagna S, and Terrezza G. Right side neglect in right

cerebellar lesion. J Neurol Neurosurg Psych 71: 114-117, 2001.

Steinlin M, Imfeld S, Zulauf P, Boltshauser E, Lövblad K-O, Ridolfi Lüthy

A, Perrig W, and Kaufmann F. Neuropsychological long-term dequelae after

posterior fossa tumour resection during childhood. Brain 126: 1998-2008, 2003.

Thach WT. What is the role of the cerebellum in motor learning and cognition?

Trends Cogn Sci 2: 331-337, 1998.

Trouillas P, Takayanagi T, Hallett M, Currier RD, Subramony SH, Wessel

K, Bryer A, Diener HC, Massaquoi S, Gomez CM, Coutinho P, Ben Hamida M,

Campanella G, Filla A, Schut L, Timmann D, Honnorat J, Nighoghossian N,

Manyam B. International cooperative ataxia rating score for pharmacological

assessment of the cerebellar syndrome. J Neurol Sci 45: 205-211, 1997

van Gijn J, Hijdra A, Wijdicks EF, Vermeulen M, and van Crevel H. Acute

hydrocephalus after aneurysmal subarachnoid hemorrhage. J Neurosurg 63(3): 355-362,

1985.


Weintraub S, and Mesulam M-M. Mental state assessment of young and

elderly adults in behavioral neurology. In: Principles of behavioral neurology, edited by

Mesulam M-M, Philadelphia: F.A. Davis, 1985, p. 71-123.

Zettin M, Cappa SF, D’Amico A, Rago R, Perino C, Perani D, Fazio F.

Agrammatic speech production after a right cerebellar haemorrhage. Neurocase 3: 375-

380, 1997.


FIGURE LEGENDS

Figure 1: Individual lesions in children with right-hemisphere (Cb1-Cb6), left-

hemisphere (Cb7-Cb10), bilateral (Cb11) and vermal (Cb12) affection on superimposed

horizontal MR sections of a healthy adult brain. Most rostral sections on the right. Ri =

right, le = left.

Figure 2: (A) Means and standard deviations of the reaction times [s] across the 16

trials in the six subsequent blocks for the cerebellar subgroups with right- and left-

hemisphere lesions and control subjects. (B) Means of the reaction times [s] across the

16 trials in the six subsequent blocks for the group of all cerebellar patients and control

subjects. The first and the last block are naming blocks (nam1, nam2), the second to

fifth block are verb generation blocks (vg1-vg4).

Figure 3: Individual percentile ranks in the subtests of the developmental language test

(Heidelberger Sprachentwicklungstest, HSET), i.e. plural-singular formation (A) and

sentence construction (B), for each patient (triangles) and control subject (squares)

plotted against age.


Table 1: Patient data and pre- and postoperative clinical signs.

ICARS-ScorePreoperative

Hydrocephalus

Sub-

ject

Gen-

der

Age Handed

-ness

Dia-

gnosis

Lesion

Site

BI Evans

Time since

Surgery

(months)

PG KF

Upper

Limb

KF

Lower

Limb

SP OC total

Cb1 f 11 Left Astro I Right 0.1 20% 41 0.0 0.0 0.0 0.0 0.0 0.0

Cb2 f 12 Right Astro I Right 2.0 31% 75 0.0 0.5 0.0 0.0 0.0 0.5

Cb3 f 17 Right Astro I Right 0.26 35% 97 1.5 7.5 1.0 0.0 4.5 14.5

Cb4 m 16 Right Astro I Right 0.25 38% 97 0.0 0.0 0.0 0.0 0.0 0.0

Cb5 m 19 Right Astro II Right 55 3.0 0.0 0.0 0.0 0.0 3.0

Cb6 f 17 Right1 Astro I Right 0.18 36% 39 2.5 7.0 1.0 0.0 2.5 13.0

Cb7 f 16 Right Astro I Left 0.10 30% 18 0.0 1.0 0.0 0.0 1.0 2.0

Cb8 m 16 Right Astro I Left 0.18 29% 161 1.0 1.0 0.0 0.0 0.0 2.0

Cb9 m 9 Right Astro I Left 0.10 16% 12 0.0 0.0 0.0 0.0 0.0 0.0

Cb10 f 14 Right Astro I Left 0.04 23% 52 1.0 3.0 0.0 0.0 2.0 6.0

Cb11 m 14 Right Astro I Bilateral 0.15 32% 78 0.0 3.0 0.0 0.0 0.0 3.0

Cb12 f 14 Right Astro I Vermis 88 0.5 1.0 0.0 0.0 0.0 1.5

Cb = cerebellar; f = female, m = male; astro I/II = astrocytoma grade I/II; BI = bicaudate index, Evans = Evans Index, pathological values are

marked in bold; ICARS-score (Trouillas et al. 1997), total score: maximum = 100, Subscores: PG: Posture and Gait (maximum = 34), KF:

Kinetic function (maximum lower limb = 16, upper limb = 36), SP: speech (maximum = 8), OC: Oculomotor (maximum = 6). 1 Cb6 was right-

handed, but preferred to use the left hand after the operation due to remaining ataxia.


Table 2: Cerebellar lesion site according to MRI-analysis.

Sub-ject

Vermis Hemisphere Nuclei

Paravermis Lateral

White

matter

Vol-ume

(cc)

Right-sided lesions

Cb 1 r: VI 1,2

Cb 2 III, IV, V, VI, VIIAt, VIIB, VIIIA, VIIIB, IX, X

r: CRI, CRII, VIIB, VIIIA

r: (VI), CRI, CRII V 1,2,3(r: PV 2)

(NF b)(NI b)

26,2

Cb 3 IX, X, r: VII At, VIIIA, VIIIB, X

r: VIIB, VIIA V 2, 3r: PV 2, 3r: MH 2,3

NF rNI rND r: p, (a)

12,0

Cb 4 VI, VIIAt,VIIb, VIIIA r: VI, CRI, CRIIl: (CR II)

V 1,2r: PV 2,3

9,6

Cb 5 V, VI, VIIAt, VIIB r: VI, (CRI, IX) V 1,2(r: PV 2)

NF bNI bND r: p, (a)

8,4

Cb 6 III, VIIAt, VIIB, VIIIA, VIIIB, IX

r: IX V 1,2,3 r: PV 1

(NF b)(NI r)ND r: p,a

10,8

Left-sided lesions

Cb 7 l: CRI, CRII 2,7

Cb 8 l: VI, CRI, CRII, VIIB, VIIIA, VIIIB

l: VI, CRI, CRII, VIIB, VIIIA, VIIIB

(l: PV2,3)l: LH

48,6

Cb 9 VIIAt, VIIB l: CRII, VIIB l: CRII, VIIB 8,0

Cb 10 IX, X r: VII Bl: VIIB,VIIIA,VIIIB, IX

V 3 ND l: p 9,0

Bilateral and pure vermal lesions

Cb 11 III, IV, V, VI, VIIAt, VIIB, VIIIA, VIIIB, IX, X

r: (V, VI,VIIIA, IX)l: (V), VI,

l: CRI V 1,2,3r: PV 2,3l: PV 2,3

NF bNI b(ND r: p)(l: a, p)

15,5

Cb 12 (I, II, III), VIIAt, VIIB, VIIIA, VIIIB, IX, X

r: (VIIAt, VIIB, VIIIA)

V 1,2,3 (NF b)NI b

8,6

r = right side, l = left side, b = both sides; subdivisions of white matter according to Luft et al. (1998), with V1-3 = vermis 1-3, PV1-3 = paravermis 1-3 MH1-3 = medial hemisphere, and LH 1-3 = lateral hemisphere 1-3. NF = fastigial nuclei, NI = interposed nuclei, ND = dentate nuclei, p = posterior, a = anterior, cc = cubiccentimeters; affected regions in brackets mark those regions marginally affected.


Table 3: Results of post-hoc tests.

Task Right cerebellar vs. control subjects

Left cerebellar vs. control subjects

All patients vs. control subjects

Verb generation (analyses of variance)

Verb generation blocks

Block: p<0.0001Group: p=0.672Block x Group: p=0.287



Naming blocksBlock: p=0.941Group: p=0.540Block x Group: p=0.560

Block: p=0.254Group: p=0.700Block x Group: p=0.578

Block: p=0.509Group: 0.474Block x Group: p=0.939

Verb gen (1-4) vs. naming

Task: p<0.0001Group: p=0.590Task x Group: p=0.581



Verb gen (1-2) vs. naming




AAT (t-tests)

Token test p=0.38 p=0.63 p=0.36

Written language p=0.65 p=0.90 p=0.53

HSET (t-tests)

Plural-Singular p=0.51 p=0.83 p=0.79

Sentence construction

p=0.69 p=0.09 p=0.65

Visuospatial tests (Mann-Whitney U-test)

Letter cancellation

Right p=0.47Left p=0.51



Line bisection Absolute p=0.14Signed p=0.03

Absolute p=0.56Signed p=0.30

Absolute p=0.43Signed p=0.01

Extinction Right p=0.34Left p=0.33



Verb gen (1-4)= analyses of variance including all four verb generation blocks; Verb gen (1-2)= analyses of variance including only the first two verb generation blocks.


Table 4: Presented objects and answers (in %) in the naming and verb generation blocks in the cerebellar and control groups.Object Most frequent answer Naming Verb generation

Naming / Verb generation Cerebellar patients

Control subjects

Cerebellar patients

Control subjects

1. Backpack (“Rucksack”) Backpack / carry (“tragen”) 77 88 56 552. Bed (“Bett”) Bed / sleep (“schlafen”) 100 100 75 853. Children’s bicycle (“Kinderfahrrad”) Bicycle (“Fahrrad”) /ride (“fahren”) 100 98 100 984. Broom (“Besen”) Broom / sweep (“fegen”) 100 92 88 645. Car (“Auto”) Car /drive (“fahren”) 100 98 100 996. Cooking pot (“Kochtopf”) Pot (“Topf”) / cook (“kochen”) 85 80 100 977. Doll (“Puppe”) Doll / play (“spielen”) 64 71 100 918. Children’s drum („Kindertrommel“) Drum (“Trommel“) / drum („trommeln“) 100 94 79 869. Football (“Fußball”) Ball (“Ball”) /play (“spielen”) 82 71 95 8810. Glasses (“Brille”) Glasses / look (“sehen”) 100 100 34 3711. Mobile phone (“Handy”) Mobile phone / phone (“telephonieren”) 86 90 100 9712. Pocket lamp (“Taschenlampe”) Pocket lamp / shine (“leuchten”) 95 92 89 7313. Scissors (“Schere”) Scissors / cut (“schneiden”) 100 100 100 10014. Smoothie (“Softeis”) Ice-cream (“Eis”) / eat (“essen”) 100 100 100 9015. Sports shoes (“Turnschuhe”) Shoes (“Schuhe”) / Wear (“tragen”), 100 92 48 4216. Weight-scale (“Waage”) Weight-scale / weigh (“wiegen”) 100 100 93 94Further answers: 1 Backpack: satchel (“Tornister“), bag (“Tasche”); satchel (“Ranzen”, “Schulranzen”), pack (“packen”, ”einpacken“, ”reinpacken“), make a journey (“verreisen”),

transport (“transportieren”), put on (“anziehen“), wind up (“aufziehen”, “aufsetzen“), take along (“mitnehmen“), go away (“weggehen“); buckle on(“aufschnallen“)

2 Bed: lie (“liegen“), sit (“sitzen“), lay down (“hinlegen“); 3 Bicycle: children’s bicycle (“Kinderfahrrad“), play (“spielen“); 4 Broom: mop (“Schrubber”, “Feger“), brush (“Kehrbesen”), broomstick (“Besenstiel“) scavange (“kehren“), brush (“bürsten”), clean (“putzen“,

“saubermachen“, “säubern“), scrub (“schrubben“); 5 Car: Golf (German brand name for a specific car), look at (“angucken“); 6 Cooking pot: pot (“Topf”), cook (“garen”), warm (“wärmen”); 7 Doll: baby (“Baby”), babydoll (“Babypuppe”), babyborn (German brand name for a specific doll), cuddle (“kuscheln”, “knuddeln”); 8 Drum: drums (“Schlagzeug“), children’s drum (“Kindertrommel“), play (“spielen”); 9 Football: Football (“Fußball“), shoot (“schießen”), kick (“treten”), throw (“werfen”); 10 Glasses: look (“gucken”), wear (“tragen”), put on (“anziehen”), wind up (“aufziehen”, “aufsetzen“), read (“lesen”), clean (“putzen”); 11 Mobile phone: telephone (“Telefon“), write (an SMS) (“schreiben”);


12 Pocket lamp: lamp (“Lampe”), light (“Licht machen“), switch on (“anmachen”, “anschalten“), see (“sehen”), light up (“beleuchten”), illuminate (“anleuchten“, “ausleuchten“); shine (“strahlen“) bring light into the dark (“Licht ins Dunkel bringen“);14 Smoothie: lick (“lecken”, “schlecken“), suck (“lutschen”), enjoy (“genießen“); 15 Sports shoes: sports shoes (“Turnschuhe“), run (“laufen”), walk (“gehen”), tie (“binden”), air (“lüften“); put on (“anziehen”);16 Weight scale: measure (“messen”), stand on (“draufstellen”).


Table 5: Results of the AAT and HSET subtests (percentile ranks).Subject Age Lesion site Token

testWritten

languagePlural-

SingularSentence constr.

Cb1 11 Right 97 99 99 54Cb2 12 Right 95 99 95 79Cb3 17 Right 97 99 14 92Cb4 16 Right 99 96 42 98Cb5 19 Right 94 100 42 54Cb61 17 Right

Mean±SD 96.4±1.95 98.6±1.5 58.4±37.1 75.4±20.71

Cb7 16 Left 99 100 76 98Cb8 16 Left 97 99 14 21Cb9 9 Left 94 96 76 35Cb10 14 Left 97 97 14 22

Mean±SD 96.75±2.06 98.0±1.83 45.0±35.8 44.0±36.56

Cb11 14 Bilateral 97 100 50 79Cb12 14 Vermal 97 99 42 92

Total Mean±SD 96.6±1.7 98.5±1.5 51.3±31.4 65.8±29.9

Con1 16 97 100 89 79Con2 16 97 97 76 98Con3 16 99 100 42 98Con4 16 99 99 50 79Con5 8 97 91 50 35Con6 12 97 97 95 54Con7 12 95 99 89 54Con8 15 97 99 89 92Con9 13 95 99 31 54

Con10 12 94 99 21 79Con11 19 99 100 8 98Con12 18 95 99 62 79Con13 16 99 99 10 79Con14 16 97 100 14 98Con15 16 99 99 21 92Con16 17 97 99 42 54Con17 17 97 100 62 92Con18 15 99 100 88 34Con19 15 97 99 62 79Con20 11 91 89 50 51Con212 8Con22 15 97 100 62 14Con23 13 99 99 21 98Con24 15 99 100 31 90Con25 13 99 82 10 21Con26 11 99 97 76 34Con27 18 99 99 8 92

Mean±SD 97.19±2.0 96.96±5.7 46.9±29.7 68.0±28.6

93.19 86.56 <0 10.8Mean con-2SD


Constr.=construction; Cb=cerebellar; Con=control; SD=standard deviation; 1 data were not collected in Cb6 because she was not a German native speaker; 2 data were not collected in Cb21 because of the young age of the subject.


Table 6: Results of the visuospatial tests (in %). Letter

cancellationLine

BisectionExtinction

Unilat Bilat Unilat BilatSubject Age Site Left Right Left Right

Cb1 11 Right 3.33 3.33 -3.46 100 100 100 90Cb2 12 Right 0 3.33 -3.87 100 100 100 90Cb3 17 Right 0 0 3.04 100 90 100 100Cb4 16 Right 0 0 -6.09 100 100 100 100Cb5 19 Right 0 0 -2.90 100 100 100 90Cb6 17 Right 0 0 -0.14 100 90 100 100

Mean±SD 0.6±1.4 1.1±1.7 -2.2±3.2 100±0 96.7±5.2 100±0 95.0±5.5

Cb7 16 Left 0 0 3.60 100 100 100 100Cb8 16 Left 6.67 0 -1.52 100 100 100 90Cb9 9 Left 6.67 0 -1.11 100 100 90 100Cb10 14 Left 0 3.33 -4.01 100 100 100 100

Mean±SD 3.3±3.8 0.8±1.7 -0.8±3.2 100±0 100±0 97.5±5.0 97.5±5.0

Cb11 14 Bilat. 0 3.33 -2.35 100 70 100 70Cb12 14 Verm 6.67 0 -1.11 100 100 100 100

Total Mean±SD 1.9±3.0 1.1±1.6 -1.7±2.8 100±0 95.8±9.0 99.2±2.9 94.2±9.0

Con1 16 0 0 0.41 100 100 100 100Con2 16 0 0 -0.97 100 100 100 100Con3 16 0 0 5.81 100 100 100 100Con4 16 0 0 3.32 100 100 100 100Con5 8 10.00 6.67 3.60 100 100 100 90Con6 12 0 0 -0.97 100 100 100 100Con7 12 0 0 -1.24 100 100 100 100Con8 15 0 0 4.29 100 90 100 100Con9 13 0 0 3.32 100 100 100 100

Con10 12 0 0 -2.63 100 90 100 100Con11 19 0 3.33 -1.80 100 100 100 100Con12 18 0 3.33 -0.69 100 100 100 100Con13 16 0 3.33 3.04 100 100 100 100Con14 16 0 0 -2.90 100 100 100 70Con15 16 0 0 -1.94 100 100 100 100Con16 17 0 0 1.94 100 80 100 90Con17 17 0 0 0.69 100 100 100 100Con18 15 0 0 0.14 100 100 100 100Con19 15 0 0 -1.94 100 100 100 100Con20 11 0 0 4.98 100 100 100 100Con21 8 3.33 3.33 0.69 100 100 100 90Con22 15 0 0 0.55 100 100 100 90Con23 13 0 0 -8.02 100 100 90 80Con24 15 0 0 4.84 100 100 100 100Con25 13 0 0 1.66 100 100 100 90Con26 11 0 0 2.49 100 100 100 100Con27 18 0 0 0.97 100 90 100 100

Mean±SD 0.5±2.0 0.7±1.7 0.7±3.0 100±0 98.1±4.8 99.6±1.9 96.3±7.4Mean con +/- 2SD

4.5 4.1 -5.3; 6.7 Const 88.5 95.8 81.5

Site=lesion site; unilat=unilateral; bilat=bilateral; Cb=cerebellar; SD=standard deviation; Con=control.


Figure 1


Figure 2

Figure 3

nam1 vg1 vg2 vg3 vg4 nam2

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

nam1 vg1 vg2 vg3 vg4 nam2

Controls Right Hemisphere Left Hemisphere

Rea

ctio

n Ti

me

[sec

]

BA

All Patients Controls

8 9 10 11 12 13 14 15 16 17 18 19

25

50

75

100

8 9 10 11 12 13 14 15 16 17 18 19

BA Sentence ConstructionPlural-Singular Formation

Per

cent

ile R

ank

Age (years)

"low average"

"below average"

"low"

Patients Controls

Age (years)

children and adolescents with chronic cerebellar lesions show no clinically relevant signs of...

Documents