bioRxiv preprint doi: https://doi.org/10.1101/849505; this version posted March 2, 2020. The copyright holder for this preprint (which was not
certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under
aCC-BY-ND 4.0 International license.

E-TAN, A TECHNOLOGY- ENHANCED PLATFORM WITH
TANGIBLE OBJECTS FOR THE ASSESSMENT OF VISUAL NEGLECT:
A MULTIPLE SINGLE - CASE STUDY

Antonio Cerrato∗
University of Naples Federico II
Naples, Italy
antonio.cerrato@unina.it

Daniela Pacella∗
University of Naples Federico II
Naples, Italy
pacelladaniela@gmail.com

Francesco Palumbo
University of Naples Federico II
Naples, Italy
fpalumbo@unina.it
Diane Beauvais
Inserm U 1127, CNRS UMR 7225, Sorbonne Université, Institut du Cerveau et de la Moelle épinière, ICM
Hôpital de la Pitié-Salpêtrière
75013 Paris, France
diabeauvais@gmail.com
Michela Ponticorvo
University of Naples Federico II
Naples, Italy
michela.ponticorvo@unina.it
Orazio Miglino
University of Naples Federico II
Naples, Italy
Institute of Cognitive Sciences and Technologies, National Research Council
orazio.miglino@unina.it
Paolo Bartolomeo
Inserm U 1127, CNRS UMR 7225, Sorbonne Université, Institut du Cerveau et de la Moelle épinière, ICM
Hôpital de la Pitié-Salpêtrière
75013 Paris, France
paolo.bartolomeo@gmail.com

March 2, 2020

∗

these authors contributed equally

bioRxiv preprint doi: https://doi.org/10.1101/849505; this version posted March 2, 2020. The copyright holder for this preprint (which was not
certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under
aCC-BY-ND 4.0 International license.

A BSTRACT
Visual neglect is a frequent and disabling consequence of right brain damage. Traditional paper-andpencil tests of neglect have limitations in sensitivity and ecological validity. The Baking Tray Task
(BTT), instead, approaches real-life situations, because it requires participants to place 16 physical
objects on a board. The number of objects placed on the left and right portions of the board provides
a clinical index of visual neglect. Here we present E-TAN, a technology-enhanced platform which
allows patients to perform an enhanced version of the BTT (E-BTT). This platform automatically
determines the object locations on the board, and also records the sequence and timing of their
placement. We used E-BTT to test 9 patients with right hemisphere damage, and compared their
performance with that obtained by 115 healthy participants. To this end, we developed a new method
of analysis of participants’ performance, based on the use of the convex hull described by the objects
on the board. This measure provides an estimate of the portion of space processed by each participant,
and can effectively discriminate neglect patients from patients without neglect. E-TAN allows
clinicians to assess visuospatial performance by using a convenient, fast, and relatively automatized
procedure, that patients can even perform at home to follow-up the effects of rehabilitation.
Keywords Attention · Visual neglect · Stroke · Rehabilitation · Diagnosis · Follow-up

2

bioRxiv preprint doi: https://doi.org/10.1101/849505; this version posted March 2, 2020. The copyright holder for this preprint (which was not
certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under
aCC-BY-ND 4.0 International license.

1 Introduction
Visual neglect is a disabling pattern of spatial and nonspatial deficits, which frequently results from right hemisphere
damage (Bartolomeo, 2007; Cubelli, 2017). Neglect patients behave as if the left part of the world did not exist anymore,
and have poor functional outcome (Bartolomeo, 2013). Diagnosis of neglect usually relies on the administration of
standard paper-and-pencil tests, such as cancellation and line bisection tasks (Wilson, Cockburn & Halligan, 1987;
Azouvi, Bartolomeo, Beis, Perennou, Pradat-Diehl & Rousseaux, 2006). However, the clinical and ecological validity
of these paper and pencil tests has limitations. For example, it has long being known that some patients can perform in
the normal range on neglect tests, and yet show clinical signs of spatial deficits, which can emerge when questioning
the patient’s carers (Azouvi, Olivier, de Montety, Samuel, Louis-Dreyfus & Tesio, 2003). In still other cases, deficits
can only be unveiled by computerized response time tasks (Bartolomeo, 1997), or by tasks taxing executive attention
(Bartolomeo, 2000; Bonato, 2012), which presumably prevent the use of compensatory mechanisms. These patients are
at risk of being considered as having fully recovered from neglect, while still suffering from spatial deficits affecting
their daily routine. These deficits can engender dangerous situations (e.g., in car driving), or induce substantial levels of
handicap in everyday life (e.g., when cooking or when choosing items at a supermarket).
The Baking Tray Task (BTT) (Tham & Tegnér, 1996) is a promising technique to detect the behavioral consequence
of visual neglect without relying on subjective questionnaires or on computerized response time tasks. Patients are
requested to dispose 16 cubes on a board as equally spaced as possible, "as if they were buns on a baking tray". In the
original study (Tham & Tegnér, 1996), no healthy participant in a group of 28 placed more than two buns in excess on
one side of the board as compared with the other. On the basis of this cutoff score, seven patients with right hemisphere
damage, who performed in the normal range on paper-and-pencil tests, showed neglect on the BTT. Thus, the BTT
proved to be a simple and yet sensitive test of neglect. Additionally, performance seems independent of demographic
variables such as participants’ age (Facchin, Beschin, Pisano & Reverberi, 2016).Given its simple and user-friendly
setup, the BTT can easily be used to follow up the effects of rehabilitation (Facchin, Beschin & Daini, 2017). However,
the analysis of results based on the number of items situated on the left/right halves of the board remains relatively
raw, and is liable to miss subtler signs of spatial deficit, such as a pathological shrinking of the explored space. Other
abnormal behavioral patters that could go undetected are the tendency to start the task from the right side (Gainotti,
D’Erme & Bartolomeo, 1991), whereas healthy participants’ tendency to start from the left side (Bartolomeo, D’Erme
& Gainotti, 1994; Gigliotta, Malkinson, Miglino & Bartolomeo, 2017), a general slowing of performance, or a spatially
disorganized placement of items on the board, without lateralized aspects. Enhancing the BTT with technological
features can enable the clinician to detect these patterns of performance.

2 Methods
2.1 E-TAN: a technologically-enhanced platform to assess visuospatial cognition with tangible interfaces
A tangible user interface system consists of an integrated system of concrete objects that participants can manipulate.
The E-TAN platform is a prototype derived from the enhancement of BTT, a tool devised to evaluate of spatial cognition
(Cerrato & Ponticorvo, 2017; Cerrato, Ponticorvo, Bartolomeo & Miglino, 2018; Cerrato, Ponticorvo, Gigliotta,
Bartolomeo & Miglino, 2019a,b; Gentile, Cerrato & Ponticorvo, 2019). The tangible physical interfaces in E-TAN
are disks 4-cm in diameter. A 30 fps camera placed on top of the board at a fixed distance, and connected to a
laptop computer, can detect the disks when they disposed on a predefined board/surface, thanks to ArUco Markers
(Garrido-Jurado, Muñoz-Salinas, Madrid-Cuevas & Marín-Jiménez, 2014). ArUco Marker tags are often used in
augmented reality systems (Cerrato, Siano & De Marco, 2018). They consist of a pattern of black and white tiles
containing compressed information, similar to QR codes (see Fig. 1).
Thus, E-TAN detects the spatial coordinates of the objects arranged on the board, and offers the opportunity to
standardise the data collection and to store the subjects’ performance in both local and online databases. E-TAN also
allows one to record the order and timing in which the objects are placed on the board, thus permitting the analysis of
the temporal sequence of subjects’ performance.
2.2 Experimental procedure
The procedure was similar to that used for the traditional BTT (Tham & Tegnér, 1996; Bailey, Riddoch & Crome, 2004;
Facchin et al., 2016), the only exceptions being that disks were used instead of cubes, and that the board dimensions
were 60 x 45 cm. Participants were asked to dispose the disks on the board as equally spaced as possible. The
administration of the task lasted about 5 minutes.
Nine right-handed patients (3 females, mean age 61.4 years, SD 13.5, range 43 - 80) were included on the basis of their
having had a first stroke in the right hemisphere. The exploratory nature of the study design required no randomization
3

bioRxiv preprint doi: https://doi.org/10.1101/849505; this version posted March 2, 2020. The copyright holder for this preprint (which was not
certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under
aCC-BY-ND 4.0 International license.

or blinding/masking. Patients performed the enhanced version of the BTT (E-BTT) on the E-TAN platform at the Brain
and Spine Institute (ICM), Pitié-Salpêtrière Hospital, Paris, France, following the procedures approved by the Inserm
(CPP 13-41) and by the local institutional review board. No adverse events occurred. Table 1 reports demographical,
clinical and neuropsychological characteristics of patients. Patients’ performance was compared with that of 115
healthy participants (61 females, mean age 27.2 years, SD 8.6, range 19-75) (Palumbo, Cerrato, Ponticorvo, Gigliotta,
Bartolomeo & Miglino, 2019). Participants were tested either at the University of Naples, Italy, or at the ICM, after
giving their consent according to the Declaration of Helsinki. The procedure was approved by the local institutional
review boards. Performance data were collected anonymously and stored in an online database. After performing
the E-BTT, some of the patients (1, 2, 4, and 6) spontaneously commented that it was an agreeable and quick task to
perform, and a welcome change after so many boring paper-and-pencil tests.
2.3 Paper-and-pencil neglect tests
In addition to E-BTT, brain-damaged patients also performed traditional paper-and-pencil tasks from the GEREN
neglect battery (Azouvi et al., 2006), including the bells cancellation task, the copy of a landscape, the overlapping
figures task, and the line bisection task (see Table 1). No patient made any omissions on the overlapping figures task.
For bells cancellation, we calculated the difference between omission on the left side and on the right side of the
sheet, with a cutoff score of 2 (i.e., two more omissions on the left than on the right half of the sheet), as proposed
by previously published guidelines (Azouvi et al., 2006). The landscape drawing consisted of a central house with
two trees per side (Bartolomeo & Chokron, 1999); the total score was the sum of left-sided omissions, with each
omission of a whole tree or of half of the house receiving a score of 1, and incompletely drawn items receiving 0.5.
For the the line bisection task (Urbanski & Bartolomeo, 2008), patients were asked to bisect five horizontal 200-mm
lines presented individually on A4 sheets of paper in landscape format (Azouvi et al., 2006). The scoring procedure
consisted in measuring the deviation in mm from the true midline, averaged across the five stimuli (Azouvi et al., 2006;
Rousseaux, Beis, Pradat-Diehl, Martin, Bartolomeo, Bernati, Chokron, Leclercq, Louis-Dreyfus, Marchal & others,
2001). Leftward or rightward deviations from the true center were assigned negative or positive scores, respectively.
The cut-off score is -7.3 mm for leftwards deviations and +6.5 mm for rightwards deviations (Rousseaux et al., 2001).

Patient

Gender/
Age/EL

Onset of
Illness
(months)

P1

M/68/3

5

H

P2

F/63/3

12

I

P3

M/50/2

6

I

P4
P5
P6
P7
P8
P9

M/46/2
M/55/3
M/43/3
M/73/1
F/80/2
F/75/1

12
133
28
123
108
10

H
I
I
I
H
I

Aetiology

Visual
Field

Visual
Extinction

Bells cancellation
(left/right hits,
max = 15/15)

Landscape
drawing
(omissions)

200-mm line
bisection
(deviation in mm)

Left inferior
quadrantanopia
Normal

No

12/14

0.5*

-11.8*

No

9/13*

1.5*

+1.6

Left
hemianopia
Normal
Normal
Normal
Normal
Normal
Normal

No

9/13*

2.5*

+1.6

No
No
No
Yes
Yes
Yes

15/15
15/14
14/14
2/11*
3/12*
3/11*

0
0
0
1.5*
1.5*
1*

-3.2
-5.6
-4.8
+22*
+27*
-9.2*

Table 1: Demographical and clinical characteristics of patients, with their performance on the neglect battery. Gender:
M, Male; F, Female. EL, educational level (1, less than 8 years of formal education; 2, 9–12 years; 3, > 12 years).
Aetiology: H, haemorrhagic; I, ischemic. Asterisks indicate pathological scores compared with normative data (Azouvi
et al., 2006; Gainotti et al., 1991; Rousseaux et al., 2001). Scores for landscape drawing indicate the number of omitted
left-sided items. For line bisection, positive values indicate rightward deviations, negative values indicate leftward
deviations.

4

bioRxiv preprint doi: https://doi.org/10.1101/849505; this version posted March 2, 2020. The copyright holder for this preprint (which was not
certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under
aCC-BY-ND 4.0 International license.

Patients can be classified in three groups according to their performance on the paper-and-pencil test battery (Table
1). Patients 1-3 had pathological performance on two tests out of three. Patients 4-6, performed normally on all tests.
Patients 7-9 obtained pathological scores on all the tests.

3 Results
3.1 Patterns of final configuration
Figure 1 displays the final disk configuration obtained by patient 6. Figure 2 shows the sequences of disk placement
used by each of the 9 patients. Five patients out of 9 started to place the disks on the right side of the board, similar to
the initial rightward orienting shown by right-brain damaged patients’ on different sorts of task (Gainotti et al., 1991;
Jalas, Lindell, Brunila, Tenovuo & Hamalainen, 2002; Siéroff, Decaix, Chokron & Bartolomeo, 2007); six out of 9
started the configuration from the bottom side of the surface. Patient 6 started from the down-right side of the surface
and then proceeded to dispose the disks vertically in four columns, occupying the whole board (see Fig. 1 and 2).
The right-sided start might indicate a residual initial rightward bias in a patient who performs normally on paper-andpencil tests (Bartolomeo, 1997, 2000; Bonato, 2012). Patients 2, 3 and 5 showed a similar sequence pattern as patient
6, but they grouped all the disks in a smaller area. In patients 2 and 3 the shrunken area is displaced rightwards, as
typical of left neglect. Also the configurations of patient 8 and 9 clearly indicate severe neglect behaviors and impaired
utilization of the available space. Patient 4 showed instead a generally disorganized placing strategy, without lateralized
aspects.

Figure 1: Example of E-BTT final configuration, provided by patient 6
Patient 7 paced the first 4 disks in the 4 corners of the surface (albeit without actually reaching the left upper corner),
then placed the remaining 12 disks along the border of the surface, using the right-sided board limit as a guide. The
final configuration remained, however, substantially asymmetrical, with only 4 disks placed in the right half of the
board. Finally, patient 1 disposed the disks without any left-right asymmetry, but confined them to the upper part of the
board. The scoring procedure of the traditional BTT would classify this pattern of performance as normal.
E-BTT analysis also offers the possibility of determining the localization of the first disk, which can give a clue to
identify the initial attentional focus during the exploration of the surface. During neglect compensation, some patients
might still initiate their exploration from the right side, and then reach the left part of the board (Bartolomeo, 1997,
2000); other patients, perhaps those who better recognize their neglect, might intentionally start their exploration from
the left (Takamura, Imanishi, Osaka, Ohmatsu, Tominaga, Yamanaka, Morioka & Kawashima, 2016). E-BTT can
record such shifts of exploratory preferences during patient follow-up, and thus characterize distinct behavioral patterns
even in patients whose final performance reaches the normal limits. For example, in our patient sample the left or
right direction of a pathological deviation on line bisection generally corresponded with the hemispace of placement of
the first E-BTT object: patients who deviated leftward on line bisection also chose to place the first object on the left
hemispace, and vice-versa. The the only exception was Patient 9, who deviated leftwards on line bisection but placed
the first E-BTT object on the right extremity of the board. Evaluation of larger patient samples is needed to assess the
generality of this relationship.
5

bioRxiv preprint doi: https://doi.org/10.1101/849505; this version posted March 2, 2020. The copyright holder for this preprint (which was not
certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under
aCC-BY-ND 4.0 International license.

Figure 2: Final E-BTT conguration and sequence of disks’ placing (from 1 to 16) for each patient. Numbers represent
the order of disk placement. The vertical midline of the board is represented in red.

3.2 Advanced analysis of the exploited area
The final spatial configuration of the disks on the board, which is accurately recorded by E-TAN platform, defines
the vertices of a 2D polygon. The polygon area thus represents the total space used by the patients during E-BTT
performance on the left and on the right halves of the board, and can be considered as an index of the surface explored
during task performance. This feature is thus likely to provide a quantitative estimate of neglect patients’ tendency to
under-utilize the left portions of space. The healthy participants’ patterns of distribution of the explored spaces on each
half of the board can be used as an anchor to identify pathological performance in brain-damaged patients.
Let a be the total covered area that is defined by the coordinates of the objects lying on the edge of the convex polygon
containing the 16 disks; the Monte Carlo integration algorithm (Press & Farrar, 1990) provides an accurate estimation
of a. The algorithm estimates the area as follows: N points are randomly generated from uniform distributions, points
are plotted on the board, and the algorithm counts the points falling into the convex hull region. As the points are
uniformly randomly generated, the covered area can be approximated by the expression a = A × n × N −1 , where n
and N indicate the points falling into the convex hull and the total points, respectively; A = 700 × 500 = 350, 000
is the total area of the board measured in pixels. As the random process follows the Binomial random variable, with
1
parameters N and π, with π unknown, p = n × N −1 is the best estimator of π with standard error (SE) of order N − 2 ,
as a consequence a is the best estimator of the covered area, and its SE is of the same order. Analogously, the portion of
a falling in the right/left portion of the board has been estimated and denoted with ar and al . Choosing N = 105 the
SE can be negligible.
A potential problem of the measure of the convex hull area is that participant may place one or a few "outlier" disks far
away from the others, thus leading to an overestimation of the total area explored. To address this issue, we estimated
a second left/right area measure for each participant, by removing the disks lying on the edge of the most external
6

bioRxiv preprint doi: https://doi.org/10.1101/849505; this version posted March 2, 2020. The copyright holder for this preprint (which was not
certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under
aCC-BY-ND 4.0 International license.

convex hull (such a removal technique is known as peeling; Eddy, 1982). Let a(E) be the area determined by the
most external convex hull, and a(I) the inner part of the polygon computed without the external points (i.e. disks)
of the whole area. The number of removed disks varies according to the configuration. Then, the measure assumed
as proxy of the explored area is given by the mean of these two convex hull areas. Defining the following quantities
â = (a(E) + a(I))/2A, âl = (al (E) + al (I))/A and aˆr = (ar (E) + ar (I))/A, it is possible to assume that they are
determination of the logit-normal random variable having support in [0, 1], with unknown parameters µ and σ. The
logit-normal is a very flexible distribution over x ∈ [0, 1], and by definition its logit function follows the Gaussian
distribution. If â ∼ logit − normal then
logit(â) = ln

â
∼ N (µ, σ).
1 − â

(1)

The same holds for âl and âr .
Plots in figure 3 refer to reference population, and they show the approximation of the empirical distribution of â, âl ,
and âr (in blue) with respect to the theoretical logit-normal distribution (in red). It is worth noting that the test based on
the measure of discrepancy cannot reject the null hypothesis H0 : f1 ≡ f2 , where f1 , f2 are the respective empirical
and theoretical density functions.

Figure 3: Empirical densities and theoretical logit-normal approximation with respect to the variables â, âl , and âr ,
from the left to the right.
The plot in Figure 4 jointly considers the standardized logit transformations of the âr and âl variables. The notation lâl
defines the logit transformation for the left covered area while lâr defines the logit transformation for the right one. For
a bivariate Gaussian the membership depends on the distance from the mean. Because the distance in more than one
dimension does not have a predefined sign, a possible solution consists in considering the squared distance. It is worth
to remind that the sum of g > 0 independent squared Gaussian random variables is distributed according to a chi-square
distribution with g degrees of freedom (DoF). In this case the variables are not independent as the correlation index ρ =
.68. The Mahalanobis distance breaks the correlation and the squared distances from the mean are characterised by a
chi-square distribution with two DoF even if the variables are correlated. The color of the points in Figure 4 depends on
membership probability to the chi-square distribution with two DoF.
Dots designate the healthy participants, squares refer to the patients with right hemisphere damage. Along with an
elliptical distribution contour, points are represented according to a color gradient (from blue to red), depending on the
similarity of the E-BTT configuration provided. Points with a green contour line represent those patients (4 and 6) plotted
along the elliptical distribution and those healthy participants (2, 43, 44, 48, 70, 89, 92, 100, 101) that, considering their
E-BTT dispositions, have a probability to belong to the "normal" group lower than .05: P (zl , zr : 0, ρ = .68) < .05. It
is expected that 5% of the reference population is wrongly classified as false positive. Consistent with this prediction,
only 9 of the 115 healthy participants (less than 8%) showed impaired E-BTT performance. On the other hand, the
performance of two of the 9 brain-damaged patients was classified as normal (patients 4 and 6 in Figure 2, who also
performed paper-and-pencil tests in the normal range). Note, however, that patient 4 displayed a spatially disordered,
although nonlateralized, pattern of object placement; as already mentioned, patient 6 started from the low right corner.
Additional methods of analysis, e.g. focusing on spatial entropy (Batty, 1974) and sequence of placement, may increase
the discrimination power of E-BTT.

7

bioRxiv preprint doi: https://doi.org/10.1101/849505; this version posted March 2, 2020. The copyright holder for this preprint (which was not
certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under
aCC-BY-ND 4.0 International license.

Another possibility is to compute the ‘center of mass’ of the explored surface (Rorden & Karnath, 2010; Gainotti, Perri
& Cappa, 2002). Patient 7 is plotted on the uppermost part of distribution, because he placed all the disks along the
borders of the surface, and thus used a wider area than expected for typical neglect-related configurations. Performance
of patients 1 and 5 was plotted in the inferior external part of the distribution, far from the center, because they covered
a relatively small area of the board.

Figure 4: Joint plot of the standardized logit transformations of the âr and âl variables. In black the ID numbers of
patients, in red the ID numbers of healthy subjects
Table 2 compares the results of the present explored area method with the original BTT scoring (Tham & Tegnér,
1996)), and with the BTT bias, a scoring method which determined normative data for clinical use of the traditional
BTT, by using the percentage ratio of the difference between right- and left-sided objects divided by the total number of
placed objects (Facchin et al., 2016).

8

bioRxiv preprint doi: https://doi.org/10.1101/849505; this version posted March 2, 2020. The copyright holder for this preprint (which was not
certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under
aCC-BY-ND 4.0 International license.

Patient

Original
BTT Scoring

BTT Bias

Explored Area

P1
P2
P3
P4
P5
P6
P7
P8
P9

0
8*
8*
2
2
0
4*
10*
16*

12.5
68.75*
66.67*
-12.5
-25*
0
25*
56.25*
100*

pathological
pathological
pathological
non pathological
pathological
non pathological
pathological
pathological
pathological

Table 2: Three types of scoring (BTT, BTT Bias and Explored Area). Asterisks indicate pathological scores. For the
Explored Area, the statistical procedure showed in the present paper, a dichotomic classification pathological/ non
pathological has been used. Original BTT scores greater than 2 indicate left neglect. BTT bias scores less than -12.5%
indicate pathological leftward bias (right neglect); values greater than 18.7% indicate pathological rightward bias (left
neglect).

When applied to our patient sample, the BTT bias (Facchin et al., 2016) led to similar conclusions, except for Patient 1,
whose performance was classified as normal by the BTT bias scoring method, but as pathological by our method.

4 Discussion
Signs of visual neglect are likely to result from the interaction of multiple deficits, with substantial variability from
patient to patient (Bartolomeo, 2007; Siéroff et al., 2007). Thus, it is not surprising that a battery of tests has more
diagnostic sensitivity that any single test alone (Azouvi et al., 2006). However, patients often manage to achieve
normal performance on paper-and-pencil tests, while still suffering from substantial levels of clinical disability. The
administration of E-BTT through the E-TAN platform allows clinicians to assess visuospatial performance by using a
convenient, fast, and relatively automatized procedure. The recorded data can be anonymized and sent to an online
platform for automatic analysis and storage for online or offline reference.
E-BTT keeps the ecological flavor of the traditional BTT, which is close to everyday life use and manipulation of
objects (Miglino & Ponticorvo, 2018; Di Fuccio, Ponticorvo, Ferrara & Miglino, 2016); however, E-BTT also offers
sensitivity to subtle forms of spatial deficits, thanks to a novel, advanced statistical procedure to assess patients’ patterns
of performance.
In addition, the E-BTT board requires patients to reach further in the left space than during typical paper-and-pencil
tests. This characteristic opens the possibility of unveiling instances of reduced leftward hand movements, or directional
hypokinesia (Bartolomeo, D’Erme, Perri & Gainotti, 1998). The E-TAN platform is easily portable and can thus be
used to follow-up patients during home-based rehabilitation (Rossit, Benwell, Szymanek, Learmonth, McKernan-Ward,
Corrigan, Muir, Reeves, Duncan, Birschel, Roberts, Livingstone, Jackson, Castle & Harvey, 2019).
Assessment procedures exploiting the usage of tangible objects are increasingly adopted for both diagnostic and
rehabilitation purposes (Ferrara, Ponticorvo, Di Ferdinando & Miglino, 2016; Cerrato, Siano, De Marco & Ricci, 2019;
Rabuffetti, Meriggi, Pagliari, Bartolomeo & Ferrarin, 2016). Previous examples of technology-enhanced assessment of
spatial deficits include virtual reality mazes (Morganti & Riva, 2014), which however do not require any manipulation
of tangible objects. Other examples include a technology-enhanced line cancellation task (D’Amico, Landucci &
Pezzatini, 2013), and a mixed reality interface equipped with tangible objects and a virtual reality environment to
evaluate and rehabilitate cognitive mistakes in daily tasks (Edmans, Gladman, Walker, Sunderland, Porter & Fraser,
2007). These systems are mainly focused on rehabilitation, and lack an in-depth performance analysis. They either
employ standard evaluation procedures of the traditional line cancellation task (Albert, 1973), or need the expert
evaluation of an occupational therapist. The SIG-Blocks interface (Lee, Jeong, Schindler & Short, 2016) makes use of
tangible objects equipped with motion sensors for cognitive assessment, based on some items of the Wechsler Adult
Intelligent Scale. Another technology-enhanced assessment and training tool, MOTO tiles (Liu, Lund & Wu, 2018),
consists of interactive modular tiles similar to jigsaw pieces. The tiles need to be placed one next to the other to
compose a sort of mat, with audio and visual stimuli. Also this platform, however, lacks the possibility of systematic
data collection and monitoring of performance.
9

bioRxiv preprint doi: https://doi.org/10.1101/849505; this version posted March 2, 2020. The copyright holder for this preprint (which was not
certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under
aCC-BY-ND 4.0 International license.

5 Conclusions and future directions
E-BTT, administered through the E-TAN platform, allows clinicians to assess patients’ patterns of performance not
only on the basis of the raw count of the objects disposed in each half of the board, but also by recording the spatial
coordinates of the objects. In addition, other analysis techniques will be developed to increase E-BTT’s discriminating
power, based on the sequence, timing and order of the patterns arranged by participants, and on the density of the disks
placed in different portions of the board. In particular, the study of the regularity of object placement can provide
information on poorly known clinical conditions, such as nonlateralized spatial deficits after brain damage.
Because of its fast, easy and user-friendly procedure, the E-TAN platform seems also ideally suited for test-retest
procedures in follow-up studies, and as a means to assess the efficacy of rehabilitation interventions. A commercial
version of the E-TAN platform will be made available shortly. Clinicians will be able to easily develop or reproduce
additional neuropsychological tasks to assess other cognitive abilities, such as numerical cognition or language (De Renzi
& Faglioni, 1978).

Disclosure Statement
No potential conflict of interest was reported by the authors.

Funding
This project has been supported by Programma Operativo Nazionale Ricerca e Innovazione 2014-2020, funded by
Italian Ministry for Education, University and Research. P.B. is supported by the Agence Nationale de la Recherche
through ANR-16-CE37-0005 and ANR-10-IAIHU-06.

References
Albert, M. L. (1973). A simple test of visual neglect. Neurology, 23(6), 658–664.
Azouvi, P., Bartolomeo, P., Beis, J.-M., Perennou, D., Pradat-Diehl, P., & Rousseaux, M. (2006). A battery of tests for
the quantitative assessment of unilateral neglect. Restorative neurology and neuroscience, 24(4-6), 273–285.
Azouvi, P., Olivier, S., de Montety, G., Samuel, C., Louis-Dreyfus, A., & Tesio, L. (2003). Behavioral assessment
of unilateral neglect: study of the psychometric properties of the catherine bergego scale. Arch Phys Med Rehabil,
84(1), 51–7.
Bailey, M. J., Riddoch, M. J., & Crome, P. (2004). Test–retest stability of three tests for unilateral visual neglect in
patients with stroke: Star cancellation, line bisection, and the baking tray task. Neuropsychological Rehabilitation,
14(4), 403–419.
Bartolomeo, P. (1997). The novelty effect in recovered hemineglect. Cortex, 33(2), 323–333.
Bartolomeo, P. (2000). Inhibitory processes and spatial bias after right hemisphere damage. Neuropsychological
Rehabilitation, 10(5), 511–526.
Bartolomeo, P. (2007). Visual neglect. Current opinion in neurology, 20(4), 381–386.
Bartolomeo, P. (2013). Attention disorders after right brain damage: Living in halved worlds. Springer.
Bartolomeo, P. & Chokron, S. (1999). Egocentric frame of reference: its role in spatial biasafter right hemisphere
lesions. Neuropsychologia, 37(8), 881–894.
Bartolomeo, P., D’Erme, P., & Gainotti, G. (1994). The relationship between visuospatial and representational neglect.
Neurology, 44(9), 1710–1710.
Bartolomeo, P., D’Erme, P., Perri, R., & Gainotti, G. (1998). Perception and action in hemispatial neglect. Neuropsychologia, 36(3), 227–237.
Batty, M. (1974). Spatial entropy. Geographical analysis, 6(1), 1–31.
Bonato, M. (2012). Neglect and extinction depend greatly on task demands: a review. Frontiers in human neuroscience,
6, 195.
Cerrato, A. & Ponticorvo, M. (2017). Enhancing neuropsychological testing with gamification and tangible interfaces:
The baking tray task. In International Work-Conference on the Interplay Between Natural and Artificial Computation,
(pp. 147–156). Springer.
10

bioRxiv preprint doi: https://doi.org/10.1101/849505; this version posted March 2, 2020. The copyright holder for this preprint (which was not
certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under
aCC-BY-ND 4.0 International license.

Cerrato, A., Ponticorvo, M., Bartolomeo, P., & Miglino, O. (2018). Btt-scan: An ecological and technology enhanced
tool to assess visual neglect. In Cognitive Processing, volume 19, (pp. S36–S36). Springer.
Cerrato, A., Ponticorvo, M., Gigliotta, O., Bartolomeo, P., & Miglino, O. (2019a). The assessment of visuospatial
abilities with tangible interfaces and machine learning. In International Work-Conference on the Interplay Between
Natural and Artificial Computation, (pp. 78–87). Springer.
Cerrato, A., Ponticorvo, M., Gigliotta, O., Bartolomeo, P., & Miglino, O. (2019b). Btt-scan: uno strumento per la
valutazione della negligenza spaziale unilaterale. Sistemi intelligenti, 31(2), 253–270.
Cerrato, A., Siano, G., & De Marco, A. (2018). Augmented reality: from education and training applications to
assessment procedures. Qwerty-Open and Interdisciplinary Journal of Technology, Culture and Education, 13(1).
Cerrato, A., Siano, G., De Marco, A., & Ricci, C. (2019). The importance of spatial abilities in creativity and their
assessment through tangible interfaces. In International Conference in Methodologies and intelligent Systems for
Techhnology Enhanced Learning, (pp. 89–95). Springer.
Cubelli, R. (2017). Definition: Spatial neglect. Cortex: A Journal Devoted to the Study of the Nervous System and
Behavior.
De Renzi, E. & Faglioni, P. (1978). Normative data and screening power of a shortened version of the token test. Cortex,
14(1), 41–49.
Di Fuccio, R., Ponticorvo, M., Ferrara, F., & Miglino, O. (2016). Digital and multisensory storytelling: narration with
smell, taste and touch. In European Conference on Technology Enhanced Learning, (pp. 509–512). Springer.
D’Amico, G., Landucci, L., & Pezzatini, D. (2013). Natural interactive system for hemispatial neglect rehabilitation. In
International Conference on Image Analysis and Processing, (pp. 501–508). Springer.
Eddy, W. F. (1982). Convex hull peeling. In Compstat 1982 5th Symposium held at Toulouse 1982, (pp. 42–47).
Springer.
Edmans, J., Gladman, J., Walker, M., Sunderland, A., Porter, A., & Fraser, D. S. (2007). Mixed reality environments
in stroke rehabilitation: development as rehabilitation tools. International Journal on Disability and Human
Development, 6(1), 39–46.
Facchin, A., Beschin, N., & Daini, R. (2017). Rehabilitation of right (personal) neglect by prism adaptation: A case
report. Annals of physical and rehabilitation medicine, 60(3), 220–222.
Facchin, A., Beschin, N., Pisano, A., & Reverberi, C. (2016). Normative data for distal line bisection and baking tray
task. Neurological Sciences, 37(9), 1531–1536.
Ferrara, F., Ponticorvo, M., Di Ferdinando, A., & Miglino, O. (2016). Tangible interfaces for cognitive assessment and
training in children: Logicart. In Smart Education and e-Learning 2016 (pp. 329–338). Springer.
Gainotti, G., D’Erme, P., & Bartolomeo, P. (1991). Early orientation of attention toward the half space ipsilateral
to the lesion in patients with unilateral brain damage. Journal of Neurology, Neurosurgery & Psychiatry, 54(12),
1082–1089.
Gainotti, G., Perri, R., & Cappa, A. (2002). Left hand movements and right hemisphere activation in unilateral spatial
neglect: a test of the interhemispheric imbalance hypothesis. Neuropsychologia, 40(8), 1350–1355.
Garrido-Jurado, S., Muñoz-Salinas, R., Madrid-Cuevas, F. J., & Marín-Jiménez, M. J. (2014). Automatic generation
and detection of highly reliable fiducial markers under occlusion. Pattern Recognition, 47(6), 2280–2292.
Gentile, C., Cerrato, A., & Ponticorvo, M. (2019). Using technology and tangible interfaces in a visuospatial cognition
task: the case of the baking tray task. In Miglino, O. & Ponticorvo, M. (Eds.), Proceedings of the First Symposium
on Psychology-Based Technologies co-located with XXXII National Congress of Italian Association of Psychology
- Development and Education section (AIP 2019), Naples, Italy, 25-26 September, 2019, volume 2524 of CEUR
Workshop Proceedings. CEUR-WS.org.
Gigliotta, O., Malkinson, T. S., Miglino, O., & Bartolomeo, P. (2017). Pseudoneglect in visual search: Behavioral
evidence and connectional constraints in simulated neural circuitry. eNeuro, 4(6).
Jalas, M. J., Lindell, A. B., Brunila, T., Tenovuo, O., & Hamalainen, H. (2002). Initial rightward orienting bias in
clinical tasks: Normal subjects and right hemispheric stroke patients with and without neglect. Journal of Clinical
and Experimental Neuropsychology, 24(4), 479–490.
Lee, K., Jeong, D., Schindler, R. C., & Short, E. J. (2016). Sig-blocks: Tangible game technology for automated
cognitive assessment. Computers in Human Behavior, 65, 163–175.
11

bioRxiv preprint doi: https://doi.org/10.1101/849505; this version posted March 2, 2020. The copyright holder for this preprint (which was not
certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under
aCC-BY-ND 4.0 International license.

Liu, Y.-X., Lund, H. H., & Wu, L.-L. (2018). Playful cognitive training with physical interactive tiles for elderly. In
2018 International Conference on Information and Communication Technology Robotics (ICT-ROBOT), (pp. 1–4).
IEEE.
Miglino, O. & Ponticorvo, M. (2018). Enhancing multi-sensory and handling-based psycho-pedagogical approaches
through new technologies. Qwerty-Open and Interdisciplinary Journal of Technology, Culture and Education, 13(1).
Morganti, F. & Riva, G. (2014). Virtual reality as allocentric/egocentric technology for the assessment of cognitive
decline in the elderly. In MMVR, (pp. 278–284).
Palumbo, F., Cerrato, A., Ponticorvo, M., Gigliotta, O., Bartolomeo, P., & Miglino, O. (2019). Clustering of behavioral
spatial trajectories in neuropsychological assessment. In SIS 2019-Smart Statistics for Smart Applications, (pp.
463–470). Pearson.
Press, W. H. & Farrar, G. R. (1990). Recursive stratified sampling for multidimensional monte carlo integration.
Computers in Physics, 4(2), 190–195.
Rabuffetti, M., Meriggi, P., Pagliari, C., Bartolomeo, P., & Ferrarin, M. (2016). Differential actigraphy for monitoring
asymmetry in upper limb motor activities. Physiological Measurement, 37(10), 1798–1812.
Rorden, C. & Karnath, H.-O. (2010). A simple measure of neglect severity. Neuropsychologia, 48(9), 2758–2763.
Rossit, S., Benwell, C. S. Y., Szymanek, L., Learmonth, G., McKernan-Ward, L., Corrigan, E., Muir, K., Reeves, I.,
Duncan, G., Birschel, P., Roberts, M., Livingstone, K., Jackson, H., Castle, P., & Harvey, M. (2019). Efficacy of
home-based visuomotor feedback training in stroke patients with chronic hemispatial neglect. Neuropsychol Rehabil,
29(2), 251–272.
Rousseaux, M., Beis, J., Pradat-Diehl, P., Martin, Y., Bartolomeo, P., Bernati, T., Chokron, S., Leclercq, M., LouisDreyfus, A., Marchal, F., et al. (2001). Presenting a battery for assessing spatial neglect. norms and effects of age,
educational level, sex, hand and laterality. Revue neurologique, 157(11 Pt 1), 1385–1400.
Siéroff, E., Decaix, C., Chokron, S., & Bartolomeo, P. (2007). Impaired orienting of attention in left unilateral neglect:
A componential analysis. Neuropsychology, 21(1), 94.
Takamura, Y., Imanishi, M., Osaka, M., Ohmatsu, S., Tominaga, T., Yamanaka, K., Morioka, S., & Kawashima, N.
(2016). Intentional gaze shift to neglected space: a compensatory strategy during recovery after unilateral spatial
neglect. Brain, 139(11), 2970–2982.
Tham, K. & Tegnér, R. (1996). The baking tray task: a test of spatial neglect. Neuropsychological Rehabilitation, 6(1),
19–26.
Urbanski, M. & Bartolomeo, P. (2008). Line bisection in left neglect: the importance of starting right. Cortex, 44(7),
782–793.
Wilson, B., Cockburn, J., & Halligan, P. (1987). Behavioural inattention test thames valley test company: Titchfield.
Hampshire, UK.

12