Neurol Sci

DOI 10.1007/s10072-015-2203-4

ORIGINAL ARTICLE

STIMA: a short screening test for ideo-motor apraxia, selective

for action meaning and bodily district
Alessia Tessari1 • Alessio Toraldo2 • Alberta Lunardelli3 • Antonietta Zadini3 •

Raffaella Ida Rumiati4

Received: 10 January 2015 / Accepted: 1 April 2015

Ó Springer-Verlag Italia 2015

Abstract We propose STIMA, a short test for ideo-motor equivalent scores and main percentile scores were com-
apraxia, allowing us to quantify the apraxic deficit ac- puted for each subscale. Participants imitated better known
cording to action meaning and affected body segment. than new gestures, and proximal better than distal gestures.
STIMA is based on a neurocognitive model holding that Age influenced performance on all subscales, while
there are two processes involved in action imitation (i.e., a education only affected one subscale. STIMA is easy and
semantic route for recognizing and imitating known ges- quick to administer, and compared to previous tests, it
tures, and a direct route for reproducing new gestures). The offers important information for planning adequate reha-
test allows to identify which imitative process has been bilitation programs based on the functional locus of the
selectively impaired by brain damage (direct vs. semantic deficit.
route) and possible deficits depending on the body segment
involved (hand/limb vs. hand/fingers). N = 111 healthy Keywords Ideomotor apraxia  Imitation  Gestures 
participants were administered with an imitation task in Neuropsychological assessment
two separated blocks of known and new gestures. In each
block, half of the gestures were performed mainly with the
proximal part of the upper limb and the remaining half with Introduction
the distal one. It resulted in 18 known gestures (nine
proximal and nine distal) and 18 new gestures (nine Ideomotor apraxia (IMA) is a deficit of execution of vol-
proximal and nine distal) for a total of 36. Each gesture was untary motor programming, unrelated to deficits of primary
presented up to a maximum of two times. Detailed criteria motor or sensitive areas, task instructions understanding,
are used to assign the final imitation score. Cut offs, object recognition or frontal inertia [1]. It affects ap-
proximately one-third of left-hemisphere (LH) stroke pa-
tients, independently of stroke type, age and gender [2],
Electronic supplementary material The online version of this
article (doi:10.1007/s10072-015-2203-4) contains supplementary
and often co-occurs with other, severe cognitive deficits
material, which is available to authorized users. such as aphasia. IMA affects the performance of both
known and new gestures, typically on imitation, but also
& Alessia Tessari when gestures are elicited through other modalities (e.g.,
[Link]@[Link]
on verbal command or visual presentation of objects), and
1
Dipartimento di Psicologia, Università degli Studi di it differs from ideational apraxia, which refers to a loss of
Bologna, Viale Berti Pichat 5, 50127 Bologna, Italy the conceptual representation of a known gesture [3].
2
Department of Brain and Behavioural Sciences, Università The imitation deficits are explained on the basis of a
degli Studi di Pavia, Pavia, Italy dual-route model (originally proposed by [4], and devel-
3
Struttura Complessa di Medicina Riabilitativa, Ospedali oped by [5–8]) assuming the existence of two pathways for
Riuniti di Trieste, Trieste, Italy transforming the visual input—the gesture to be imitated,
4
Settore di Neuroscienze Cognitive, Scuola Internazionale performed by the examiner, in a motor act—the gesture
Superiore di Studi Avanzati, Trieste, Italy performed by the patient (Fig. 1). If, after visual analysis,

123
 Neurol Sci

on healthy individuals performing both gesture types. The

two routes are associated with separate brain areas: the
semantic route mainly relies on LH areas (inferior tempo-
ral, parahippocampal, and angular gyri); the direct route
includes a more extensive network of cortical areas (i.e.,
superior parietal cortex bilaterally, right parieto-occipital/
occipito-temporal junctions and left superior temporal
cortex [12, 15, 16]). Moreover, the composition of the list
of actions to be imitated—new and known gestures inter-
mixed in a same list vs. presented in separate lists— has a
role [5, 12, 17, 18]. With mixed lists, the direct route is
used for imitating both types of action; with separate lists,
the semantic route is selected for imitating known gestures
and the direct route for new gestures [12, 18]. This strategy
allows the participant to minimize the number of switches
between the two routes, hence reducing cognitive load
[18].
The most widespread tests for IMA [3, 19, 20] can de-
tect severe ideomotor deficits. However, they were not
standardized to identify selective or disproportionate
damage to one of the two routes, which would be critical
for tailoring the rehabilitation technique for each specific
patient (see [21] for a review of rehabilitation approaches).
Patients with direct-route damage are impaired at learning
Fig. 1 Modified version of the dual-route model for action imitation new gestures by imitation, even though in a domestic
proposed by Rumiati and Tessari [7]. After early visual processing, context they can properly use objects and tools. By con-
shared by both routes, known gestures automatically activate the trast, patients with semantic-route damage can learn new
semantic route, using information stored in long-term memory motor skills, but are impaired in a domestic context, be-
(LTM). By contrast, new gestures are imitated via the direct route,
which decomposes the seen gesture into smaller motor components cause they cannot retrieve motor information associated
which are stored in working memory (WM) till they are reproduced. with known objects. Hence, identifying these two patient
The LTM–WM connections allow learning of new gestures types would much improve the effectiveness of reha-
bilitation programs.
New IMA batteries have been proposed (e.g., [19, 20])
the gesture is recognized, i.e., it belongs to the motor that evaluate gesture recognition, identification and pro-
repertoire of the individual, it is processed via the ‘‘se- duction in detail. However, administration time is usually
mantic route’’ (enabling only imitation of known gestures). so long as to advise their use just in a post-screening phase,
If the gesture is new, after visual processing, it is decom- after patients received an IMA diagnosis. Some of the tests
posed into simpler components, which are held in working (e.g. [23]) require gesture production only on verbal
memory till they are physically reproduced (‘‘direct route’’ command, thus providing ambiguous information (most
[5, 7]). LH patients have language comprehension deficits). Ad-
Regardless of gesture type, lesions of inferior parietal ditionally, some of these tests do not analyze the known/
cortex, subcortical structures, and premotor cortex in the new dissociation and the distinction between distal (fingers
LH are most frequently associated with IMA. Cortical le- and hand) and proximal (arm) components of gesture
sions tend to be associated with sequence errors, body-part- production, relating more to grasping and reaching, re-
as-a-tool errors or unrecognizable gestures while subcor- spectively [24]. However, distal and proximal components
tical lesions tend to be associated with postural or timing show different vulnerability after brain damage [8, 24–31].
errors [9–14]; see 13 for a review]. Right-handed indi-
viduals with LH damage show IMA of both upper limbs
[3]. However, the right limb is often plegic, so IMA is Aim of the study
usually tested only with the left limb. The anatomo-func-
tional correlates of IMA have been analyzed in brain- We wish to propose a new short IMA test to be used in the
damaged patients with selective deficits in imitating known screening phase, and which is able to separately test (1)
or new gestures [8, 12, 14] and in neuroimaging research direct-route from semantic-route deficits, and (2) deficits of

123
Neurol Sci

the proximal vs. distal movement components. This would and arm to reproduce their position correctly with respect to
help fast and accurate IMA diagnosis and classification of either other body parts or between them.
patients, allowing for tailored rehabilitation. Longer, in- Correct imitation on first presentation was granted 2
depth assessment might then be performed with ad hoc points. If a participant failed to reproduce the gesture
batteries (e.g. [19, 20, 32]). correctly on first presentation, the experimenter presented it
a second time; correct imitation after second presentation
was granted 1 point. A double failure was scored 0. The
Method maximum test score was 72/72. Each participant’s perfor-
mance was videotaped and later analyzed by a second in-
Participants dependent judge. If there was no agreement between the
examiner and a second judge (A. Tessari, who later wat-
We recruited 111 participants (55 females, ched the video-recorded performance of all participants),
age = 60.2 ± 15.5, range 30–84, education = 9.8 ± 4.04, the participant was discharged by the study (only 1 par-
range 4–201). Inclusion criteria were: (1) aged ticipant, out of an original sample of 112, was excluded).
30–90 years; (2) not showing anamnestic or clinical evi- After the imitation task, each participant was asked to
dence of neurological disease, head trauma, psychiatric recognize the 18 known gestures. This will be critical for
disorders requiring pharmacological intervention, evidence telling pre-semantic/semantic from post-semantic deficits
of alcoholism or drug addiction; (3) being right-handed on in patients: impaired recognition with intact imitation of
the Edinburgh Test [33]. Each participant signed a state- known gestures suggests a pre-semantic or semantic deficit
ment of informed consent. along the semantic route; impaired imitation with intact
recognition of known gestures would suggest post-seman-
Procedure tic damage (Fig. 1).
The test normally takes 2–3 min for a non-apraxic
Ten experts not directly involved in the research project person. It can take up to 4–5 min when administered to
selected 18 known gestures (easily recognizable) and 18 severe apraxic patients.
non-recognizable gestures (see Appendix). Half the known
gestures mainly involved the hand (e.g., OK sign), while
Statistical methods
the others mainly involved the arm (e.g., military salute).
Known and new gestures were presented in separate
Collinear predictors, distribution shapes and statistical
blocks, known gestures first, to avoid the participant from
models
selecting the direct route as a default strategy.
The examiner, previously trained by an investigator
Education showed the typical correlation profile due to
through a demo ([Link]
social evolution in the last decades in Italy: age and
[Link]; the video is for demonstration purposes for the
education were anticorrelated (Spearman’s q = -.473,
examiner only. During the test, it is recommended that the
p \ .001), and women showed a slightly lower education
examiner to stand, next to the patient, in order to be able to
level than men (Mann–Whitney z = 2.42, p = .016), an
easily perform the proximal gesture and to perform the distal
effect emerging from the oldest individuals. Hence colli-
new gestures resting his/her hand on a table), presented each
nearity affected our demographic predictors. To disentan-
stimulus up to two times. The examiner demonstrated each
gle their effects on imitation performance, we had to
gesture with his/her right hand and the participant imitated it
introduce them simultaneously in a single analysis. We
in a mirror fashion by his/her left limb.2 Participants were
used generalized linear model (GzLM) with Tweedie dis-
instructed to imitate the gesture in a mirror-like configura-
tribution (1.5) and Log-link function. Indeed GzLM
tion and to pay attention to the exact position of both hand
Tweedie distribution can accurately model markedly non-
normal score distributions: on our test, most scores lay at,
1
Some of the oldest participants had not completed the elementary or close to ceiling, with a long tail towards lower values
school (i.e. they had 4 years of education); some others carried out a (Skewness ranged -1.08 to -1.99 in different subscales;
2-year post-graduate master (SSIS—‘‘Scuola di Specializzazione Kurtosis ranged 1.47–5.00). After having detected sig-
all’Insegnamento Secondario’’) which used to be necessary to become
a teacher in Italy.
nificant predictors, we modeled their effects on the scores,
2
We tested imitation with the left limb only for two reasons (1)
hence providing correction equations and tables. Overall,
because IMA is much more frequent and severe after left hemisphere the procedure was as follows.
damage, and (2) because previous studies reported no difference
between right upper limb and left upper limb in gesture imitation [3, 1. We computed Score minus MaxScore (so that ceiling
19]. values became 0, a necessary condition for the

123
 Neurol Sci

Tweedie model). We then applied GzLM to identify t(110) = 4.836, p \ .001) were found, with a significant
critical predictors, with a backward selection tech- interaction [t(110) = 4.702, p \ .001]. Post hoc tests
nique: on a first step age, education and gender were showed no body-segment effect within known gestures
introduced in the analysis; then variables surviving a [t(110) = 1.205, p = .231), while such effect appeared
p \ .05 threshold, one-tailed in the expected direction within new gestures [t(110) = 5.745, p \ .001). meaning
for age and education, two-tailed for gender, accessed had a significant effect both for proximal [t(110) = 3.425,
a second step, and so on, until only p \ .05 predictors p \ .001] and distal [t(110) = 8.528, p \ .001] gestures,
survived (Table 1). even though it was markedly higher in the latter. The
2. Scores were corrected for the predictors surviving step overall profile is visible in Fig. 2: proximal and distal
(i) (i.e., only age, in all cases). We fitted a two- gestures were imitated at a similar level when they were
parameter quadratic model, raw score known (mean proximal = 17.04 vs. mean distal = 16.85);
R = i ? q(Age - 30)2, with i = intercept and q = s- when gestures were new, distal were imitated worse (16.47
lope of quadratic component, and derived corrected vs. 15.15). The known vs. new advantage was clear among
scores Cage30 for minimal age in the sample (30 years). proximal (17.04 vs. 16.47) and distal (16.85 vs. 15.15).
The linear component was omitted because it was not
significant (see ‘‘Results’’). Overall score (0–72): model and correction table
3. We tested whether corrected scores Cage30 were really
independent of other predictors (education and gen- Age was the only significant predictor of the overall score on
der); if so, the corrected scores were used as the final GzLM analysis (see Table 1). Score drops with age (Fig. 3).
standardization outcome; if not, a further second-level When fitting a standard second-order polynomial a significant
correction was applied. In both cases, correction quadratic component was detected [t(108) = 2.268,
equations and tables were provided. p = .025] without linear component [t(108) = 1.503,
p = .136]. Such non-linear pattern was not due to ceiling (72/
We repeated this procedure separately for the overall
72, achieved by seven young individuals), as the exclusion of
score (0–72), for the subscales known (0–36), new (0–36),
an identical proportion of top-scoring individuals from the
proximal (0–36), distal (0–36) gestures, and for the four
older age classes did not change the profile [quadratic:
atomic subscales known proximal (0–18), known distal
t(84) = 2.165, p = .033; linear: t(84) = 1.428, p = .157].
(0–18), new proximal (0–18), new distal (0–18) gestures.
We implemented a model with only intercept i and
quadratic q components: raw score R = i ? q(Age -
30)2. Given that variance increases with age [the four age
Results
classes 30–46, 47–62, 63–73, 74–84, yielded a significant
Levene (3107) = 3.126, p = .029], we included a linear
All participants recognized each and every ‘‘known’’ ges-
link between intercept i (=performance at age = 30) and
ture (100 % accuracy). Imitation performance was ana-
quadratic decrement q, to account for this variance in-
lyzed as detailed in the following paragraphs.3
crease. The final equation providing an age-corrected
score, standardized for age = 30, was:
Meaning and body segment effects h i
Cage30 ¼ Raw score þ 0:02068ðAge  30Þ2 =
Between-subscale differences showed close-to-normal h i
distributions (Skewness ranged 0.04–0.90, Kurtosis -0.37 1 þ ðAge  30Þ2 =3936
to 3.70) so paired-samples t tests were used. Main effects
of meaning (known vs. new gestures, t(110) = 8.178, If Cage30 [ 72; make it ¼ 72:
p \ .001) and body-segment (distal vs. proximal, This age-corrected score correlated neither with educa-
tion (q = .119, p = .215) nor with gender (Mann–Whit-
3
We also analyzed the imitation errors made by the participants.
ney, z = 1.42, p = .156). Hence, no further correction was
They only consisted of very few spatial errors of the hand or limb needed. Table 2 allows to find the raw scores correspond-
configuration. The task, indeed, is very simple for healthy par- ing to percentiles 5th, 10th, 25th, 50th, 75th, and to
ticipants, as showed by the ceiling effect in all conditions but the new equivalent scores [34] 0–4, given the patient’s age.
distal movements, where older participants produced some more
spatial errors in hand configuration. A different error pattern may be
expected with ideomotor apraxic patients: both hand and limb spatial Subscales
configuration errors should be found, together with omissions,
unrecognizable movements, orientation errors, visuo-semantic errors,
perseverations, and substitutions (see [12] for a detailed description of For the sake of consistency, we applied the same general
all errors categories). model as that used with overall-score to all subscale-

123
Neurol Sci

Table 1 Generalized linear Step 1 Step 2

model results for the effects of
age, education, gender Age Education Gender Age
Wald 1-tailed p Wald 1-tailed p Wald 2-tailed p Wald 1-tailed p

Overall score 27.836 \.001*** .872 .175 1.069 .301 41.248 \.001***
Known 11.07 \.001*** 1.25 .132 .117 .732 19.716 \.001***
New 21.158 \.001*** .164 .3425 1.095 .295 28.686 \.001***
Proximal 8.559 .0015** .92 .1685 .848 .357 15.899 \.001***
Distal 20.361 \.001*** .252 .308 .397 .529 27.701 \.001***
Known proximal 3.778 .026* .485 .243 .106 .745 8.062 .0025**
Known distal 8.275 .002** .933 .167 .026 .871 13.188 \.001***
New proximal 5.133 .0115* .421 .258 .855 .355 8.738 .0015**
New distal 15.453 \.001*** .002 .483 .369 .544 19.335 \.001***
Step 1: analyzed all 3 predictors simultaneously. Step 2: only variables surviving one-tailed p \ .05
selection on step 1, i.e., age, were included. Wald statistics (df = 1), two-tailed p values for gender and
one-tailed (expected direction) p values for education and age are reported
* p \ .05, ** p \ .01, *** p \ .001

Fig. 2 Mean performance (SE) of 111 participants (0–18 scale) as a

function of gesture meaning (known–new) and body segment
(proximal–distal)

scores. Table 3 reports the fitted quadratic models. The

age-corrected Cage30 scores did not correlate with
education or gender, so the models were taken as the Fig. 3 Overall scores (range 0–72) by the 111 participants as a
function of age. Curves show percentiles 3.1 (the boundary between
final ones. One exception was Cage30 of the known distal equivalent scores 0 and 1), 10.7 (between 1 and 2), 26.8 (between 2
subscale, which correlated with education. This effect and 3), and median (between 3 and 4) according to the quadratic
was modeled by a simple linear regression, leading to a model detailed in the text
further correction. The final score is corrected both for
age (standardized at 30) and education (standardized at Discussion
20), Cage30/ed20.
Correction tables with equivalent scores and per- The purpose of this study was to provide a new, short test
centiles for all eight subscales (known, new, proximal, for detecting IMA deficits that specifically affects the
distal, known proximal, known distal, new proximal, new imitation of known/new gestures or different body
distal gestures) are reported in the supplementary mate- segments.
rial. This makes all subscales ready to use in clinical Results showed that known gestures are imitated more
practice without using the complex correction formulae accurate than new gestures, and that gestures involving
reported in Table 3. proximal segments are imitated better than those involving

123
 Neurol Sci

Table 2 Overall score Age Equivalent Percentile

correction (range 0–72). Raw
score and age are the entries, 0 1 2 3 4 5 10 25 50 75
and percentiles and equivalent \ From To From To From To [
scores are the output
Corrected
61.2 61.2 65.8 65.8 68.0 68.0 69.6 69.6 62.8 65.6 67.8 69.6 71.2
Raw scores
30 61.2 61.2 65.8 65.8 68 68 69.6 69.6 62.8 65.6 67.8 69.6 71.2
33 61.2 61.2 65.8 65.8 68 68 69.5 69.5 62.8 65.5 67.8 69.5 71.2
36 61 61 65.7 65.7 67.9 67.9 69.5 69.5 62.7 65.4 67.7 69.5 71.1
39 60.8 60.8 65.5 65.5 67.7 67.7 69.3 69.3 62.5 65.3 67.5 69.3 71
42 60.5 60.5 65.2 65.2 67.5 67.5 69.1 69.1 62.2 65 67.3 69.1 70.8
45 60.1 60.1 64.9 64.9 67.2 67.2 68.9 68.9 61.8 64.7 67 68.9 70.6
48 59.6 59.6 64.5 64.5 66.9 66.9 68.6 68.6 61.3 64.3 66.7 68.6 70.4
51 59 59 64 64 66.5 66.5 68.2 68.2 60.8 63.8 66.3 68.2 70.1
54 58.3 58.3 63.5 63.5 66 66 67.8 67.8 60.1 63.3 65.8 67.8 69.7
57 57.5 57.5 62.9 62.9 65.5 65.5 67.4 67.4 59.4 62.6 65.3 67.4 69.3
60 56.6 56.6 62.2 62.2 64.9 64.9 66.9 66.9 58.6 62 64.7 66.9 68.9
63 55.7 55.7 61.5 61.5 64.3 64.3 66.3 66.3 57.7 61.2 64 66.3 68.4
66 54.6 54.6 60.7 60.7 63.6 63.6 65.7 65.7 56.7 60.4 63.3 65.7 67.9
69 53.4 53.4 59.8 59.8 62.8 62.8 65 65 55.7 59.5 62.5 65 67.3
72 52.2 52.2 58.8 58.8 62 62 64.3 64.3 54.5 58.5 61.7 64.3 66.6
75 50.9 50.9 57.8 57.8 61.1 61.1 63.5 63.5 53.3 57.4 60.8 63.5 66
78 49.4 49.4 56.7 56.7 60.1 60.1 62.6 62.6 52 56.3 59.8 62.6 65.2
81 47.9 47.9 55.5 55.5 59.1 59.1 61.8 61.8 50.6 55.1 58.8 61.8 64.5
84 46.3 46.3 54.2 54.2 58 58 60.8 60.8 49.1 53.9 57.7 60.8 63.7
E.g., to assess a 67-year-old patient who obtains a 55/72 score, look at the row reporting the closest age
(66 years) and read the equivalent score, which is 1 in this case (between 54.6 and 60.7), and percentile: 55
is below the 5th, which is 56.7

distal segments. These two difficulties interact: new ges- than other tests presenting known and new actions in mixed
tures involving distal segments are over-additively difficult lists (e.g. [3, 20]). With mixed lists, participants are likely
(Fig. 2). Unlike gender, age has a significant impact on all to rely on the direct route only, as this can imitate both
subscales; education had a marginal effect on one subscale gesture types, thus avoiding the cognitive load of fre-
(distal known gestures). quently switching between the two routes. However, this
We also estimated and subtracted the effects of age on strategy would swamp any experimental difference be-
all subscales, and provided tables for converting raw scores tween known and new gestures. By contrast, separate-
into equivalent scores [34] and percentiles. While equiva- blocks presentation minimizes the cognitive load (no
lent scores have well-known meaning in clinical practice, switch is required within each block), hence prompting the
the fifth percentile is conventionally accepted as cut off for use of one route in each condition: the semantic route for
diagnosis in research. A patient whose score on imitation of known gestures and the direct route for new gestures [5, 12,
known gestures is below 5th percentile is likely to have a 16–18].
damaged semantic route, while a patient failing at imitating Other advantages of STIMA over other tests are that it is
new gestures is likely to have an impaired direct route. We quick to administer (which makes it usable in the bed-side
also provided equivalent scores and percentiles for distal screening phase) and it includes differential evaluation of
and proximal movements, as a large literature showed their body segments, distal vs. proximal.
sensitivity to different anatomical lesions [8, 25, 28, 30, 31, Longitudinal studies (e.g., [22]) show that IMA reha-
35–37]. bilitation is necessary, since spontaneous recovery rate is
Since STIMA presents known and new gestures in only 50 %. An accurate diagnosis of the specific aspects
separate blocks, it should be generally a more sensitive underlying IMA is critical to choose appropriate reha-
detector of dissociations between the two types of gestures bilitation programs. The correct identification of the

123
Neurol Sci

Table 3 Best quadratic models, residual correlations of corrected scores with education and gender, and further corrections are reported for all
STIMA subscales
Subscale Range Best quadratic model Residual Correlations Further correction model
Cage30/education Cage30/gender
(Spearman) (Mann–Whitney)

Known 0–36 Cage30 = [Raw score ? 0.0391 (Age - q = .125, z = .229,

30)2]/[1 ? (Age - 30)2/931] p = .191 p = .819
New 0–36 Cage30 = [Raw score ? 0.00852 (Age - q = .077, z = 1.325,
30)2]/[1 ? (Age - 30)2/5100] p = .419 p = .185
Proximal 0–36 Cage30 = [Raw score ? 0.02047 (Age - q = .113, z = 1.42,
30)2]/[1 ? (Age - 30)2/1803] p = .236 p = .156
Distal 0–36 Cage30 = [Raw score ? 0.01683 (Age - q = .095, z = .829,
30)2]/[1 ? (Age - 30)2/2297] p = .322 p = .407
Known 0–18 Cage30 = [Raw score ? 0.02002 (Age - q = .142, z = .344,
proximal 30)2]/[1 ? (Age - 30)2/908] p = .136 p = .731
Known 0–18 Cage30 = [Raw score ? 0.01337 (Age - q = .202, z = .445, Cage30/ed20 = Cage30 ? (18 -
distal 30)2]/[1 ? (Age - 30)2/1403] p = .033 p = .657 Cage30) (20 - Educ)/(26 - Educ)
New 0–18 Cage30 = [Raw score ? 0.01406 (Age - q = .096, z = 1.512,
proximal 30)2]/[1 ? (Age - 30)2/1280] p = .317 p = .13
New distal 0–18 Cage30 = [Raw score ? 0.00235 (Age - q = .004, z = .474,
30)2]/[1 ? (Age - 30)2/18,250] p = .967 p = .635
When a corrected score, Cage30 or Cage30/ed20 was outside the range of the original scale, it was brought to the closer limit (e.g., if the corrected
score of the known-gestures subscale, Cage30, was [36, it was brought to 36)

damaged imitation process provided by STIMA makes the general apraxic deficit or dissociations between known and
different stakeholders (psychologists, physiotherapists, new gestures. The present version will allow even subtler
speech therapists, doctors) able to tailor the rehabilitation distinctions, given that it provides nine different scales, and
procedure to the individual patient. For example, if damage for each of them, age- and education-corrected scores as
mainly lies in the direct route, the patient cannot learn new well as equivalent scores; the correction for demographic
gestures by imitation: the rehabilitator may exploit the variables will increase sensitivity, while the use of
(relatively intact) repertoire of gestures that are already equivalent scores will allow an estimation of deficit
known by the patient. Here, the ‘‘substitutive’’ method, in severity (which single cut offs do not provide).
which spared capacities can stand in for the compromised
function by alternative strategies of compensation (e.g., Conflict of interest All authors declare that they do not have any
real or perceived conflicts of interest pertaining to the present study.
[38]), is appropriate. If, on the contrary, the semantic route
is more damaged, the patient is unable to access, retrieve or Ethics statement This research was conducted in accordance with
implement semantic information about known gestures in the Helsinki Declaration and approved by SISSA’s ethics committee.
an appropriate motor program. Here, the rehabilitator may Participants gave their informed consent to participate.
take advantage of the ability to learn by imitation, through
the direct route, and try to create a new trace in episodic
memory [39] using the ‘‘substitutive’’ [38] or the
References
‘‘restorative’’ method, in which the lost function is trained
to bring its effectiveness as close to pre-morbid levels as 1. De Renzi E, Faglioni P (1999) Apraxia. In: Denes G, Pizzamiglio
possible (e.g., [40]; [21] for a review). L (eds) Handbook of clinical and experimental neuropsychology.
A previous version of this test has already been used in Psychology Press, Hove, UK
two large group studies with brain damaged patients 2. Donkervoort M, Dekker J, van den Ende E et al (2000) Preva-
lence of apraxia among patients with a first left hemisphere stroke
Mengotti et al. [41] and patients with Parkinson Disease in rehabilitation centres and nursing homes. Clin Rehabil
Bonivento et al. [42]. Stimuli and procedure were exactly 14(2):130–136
the same, with the difference being in the size of the 3. De Renzi E, Motti F, Nichelli P (1980) Imitating gestures. A
control sample and especially, in the grain of statistical quantitative approach to ideomotor apraxia. Arch Neurol
37(1):6–10
analysis, which only reported cut offs for the total score 4. Rothi LJG, Ochipa C, Heilman KM (1991) A cognitive neu-
and for the known and new gestures subscales. In those ropsychological model of limb praxis. Cogn Neuropsychol
studies, the test proved sensitive in detecting either a 8:443–458

123
 Neurol Sci

5. Cubelli R, Marchetti C, Boscolo G et al (2000) Cognition in 25. Sirigu A, Duhamel JR, Poncet M (1991) The role of sensorimotor
action: testing a model of limb apraxia. Brain Cogn experience in object recognition. Brain 105:301–312
44(2):144–165 26. Poizner H, Mack L, Verfaellie M et al (1991) Three dimensional
6. Buxbaum LJ (2001) Ideomotor apraxia: a call to action. Neuro- computer graphic analysis of apraxia. Brain 113:85–101
case 7(6):445–458 27. Shallice T, Venable N, Rumiati RI (2005) Dissociable distal and
7. Rumiati RI, Tessari A (2002) Imitation of novel and well-known proximal motor components: evidence from perseverative errors
actions: the role of short-term memory. Exp Brain Res in three apraxic patients. Cogn Neuropsychol 22:625–639
142(3):425–433 28. Goldenberg G (1999) Matching and imitation of hand and finger
8. Goldenberg G, Hagmann S (1997) The meaning of meaningless postures in patients with damage in the left or right hemisphere.
gestures: a study of visuo-imitative apraxia. Neuropsychologia Neuropsychologia 37:559–566
35(3):333–341 29. Goldenberg G (2001) Imitation and matching of hand and finger
9. Hanna-Pladdy B, Heilman KM, Foundas AL (2001) Cortical and postures. Neuroimage 14:S132–S136
subcortical contributions to ideomotor apraxia: analysis of task 30. Goldenberg G, Karnath HO (2006) The neural basis of imitation
demands and error types. Brain 124:2513–2527 is body-part specific. J Neurosci 26:6282–6287
10. Papagno C, Della Sala S, Basso A (1993) Ideomotor apraxia 31. Tanaka S, Inui T (2002) Cortical involvement for action imitation
without aphasia and aphasia without apraxia: the anatomical of hand/arm postures versus finger configuration: an fMRI study.
support for a double dissociation. J Neurol Neurosurg Psychiatry Neuroreport 13:1599–1602
56:286–289 32. Bickerton WL, Riddoch MJ, Samson D et al (2012) Systematic
11. Pramstaller PP, Marsden CD (1996) The basal ganglia and assessment of apraxia and functional predictions from the birm-
apraxia. Brain 119:319–340 ingham cognitive screen. J Neurol Neurosurg Psychiatry
12. Tessari A, Canessa N, Ukmar M et al (2007) Neuropsychological 83:513–521
evidence for a strategic control of multiple routes in imitation. 33. Oldfield RC (1971) The assessment and analysis of handedness:
Brain 130:1111–1126 the Edinburgh inventory. Neuropsychologia 9:97–113
13. Leiguarda R (2001) Limb apraxia: cortical or subcortical. Neu- 34. Capitani E (1997) Normative data and neuropsychological
roimage 14:S137–S141 assessment. Common problems in clinical practice and research.
14. Bartolo A, Cubelli R, Della Sala S et al (2001) Double disso- Neuropsychol Rehabil 7:295–310
ciation between meaningful and meaningless gesture production 35. Goldenberg G, Strauss S (2002) Hemisphere asymmetries for
in apraxia. Cortex 37:696–699 imitation of novel gestures. Neurology 59:893–897
15. Peigneux P, Van der Linden M, Garraux G et al (2004) Imaging a 36. Sirigu A, Cohen L, Duhamel JR et al (1995) A selective im-
cognitive model of apraxia: the neural substrate of gesture- pairment of hand posture for object utilization in apraxia. Cortex
specific cognitive processes. Hum Brain Mapp 21(3):119–142 31:41–55
16. Rumiati RI, Weiss PH, Tessari A et al (2005) Common and 37. Tanaka S, Inui T, Iwaki S et al (2001) Neural substrates involved
differential neural mechanisms supporting imitation of mean- in imitating finger configurations: an fMRI study. Neuroreport
ingful and meaningless actions. J Cogn Neurosci 17:1420–1431 12:1171–1174
17. Cubelli R, Bartolo A, Nichelli P et al (2005) List effect in apraxia 38. van Heugten CM, Dekker J, Deelman BG et al (2000) Reha-
assessment. Neurosci Lett 407:118–120 bilitation of stroke patients with apraxia: the role of additional
18. Tessari A, Rumiati RI (2004) The strategic control of multiple cognitive and motor impairments. Disabil Rehabil
routes in imitation of actions. J Exp Psychol Hum Percept Per- 22(12):547–554
form 30(6):1107–1116 39. Tessari A, Bosanac D, Rumiati RI (2006) Effect of learning on
19. Bartolo A, Cubelli R, Della Sala S (2008) Cognitive approach to imitation of new actions: implications for a memory model. Exp
the assessment of limb apraxia. Clin Neuropsychol 22(1):27–45 Brain Res 173(3):507–513
20. Vanbellingen T, Kersten B, Van Hemelrijk B et al (2010) 40. Smania N, Girardi F, Domenicali C, Lora E, Aglioti S (2000) The
Comprehensive assessment of gesture production: a new test of rehabilitation of limb apraxia: a study in left-brain-damaged pa-
upper limb apraxia (TULIA). Eur J Neurol 17(1):59–66 tients. Arch Phys Med Rehabil 81(4):379–388
21. Cantagallo A, Maini M, Rumiati RI (2012) The cognitive reha- 41. Mengotti P, Corradi-Dell’Acqua C, Negri GA, Ukmar M, Pe-
bilitation of limb apraxia in patients with stroke. Neuropsychol savento V, Rumiati RI (2013) Selective imitation impairments
Rehabil 22(3):473–488 differentially interact with language processing. Brain
22. Basso A, Capitani E, Della Sala S et al (1987) Recovery from 136(8):2602–2618
ideomotor apraxia. A study on acute stroke patients. Brain 42. Bonivento C, Rumiati RI, Biasutti E, Humphreys GW (2013) The
110:747–760 role of the basal ganglia in action imitation: neuropsychological
23. Vanbellingen T, Kersten B, Van de Winckel A et al (2011) A new evidence from Parkinson’s disease patients. Exp Brain Res
bedside test of gestures in stroke: the apraxia screen of TULIA 224(2):211–220
(AST). J Neurol Neurosurg Psychiatry 82(4):389–392
24. Goodale MA, Meenan JP, Bülthoff HH et al (1994) Separate
neural pathways for the visual analysis of object shape in per-
ception and prehension. Curr Biol 4(7):604–610

123