This article was downloaded by: [188.229.29.

180]
On: 13 February 2013, At: 01:14
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Advanced Robotics
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/tadr20

A novel method to develop an animal model of

depression using a small mobile robot
Hiroyuki Ishii

a b

, Qing Shi , Shogo Fumino , Shinichiro Konno , Shinichi Kinoshita ,

d

Satoshi Okabayashi , Naritoshi Iida , Hiroshi Kimura , Yu Tahara , Shigenobu Shibata

& Atsuo Takanishi

a c

a b c e

Faculty of Science and Engineering, Waseda University, Tokyo, Japan

Research Institute for Science and Engineering, Waseda University, Tokyo, Japan

Graduate School of Science and Engineering, Waseda University, Tokyo, Japan

Faculty of Letters, Art and Science, Waseda University, Tokyo, Japan

HRI, Waseda University, Tokyo, Japan

Version of record first published: 29 Jan 2013.

To cite this article: Hiroyuki Ishii , Qing Shi , Shogo Fumino , Shinichiro Konno , Shinichi Kinoshita , Satoshi Okabayashi ,
Naritoshi Iida , Hiroshi Kimura , Yu Tahara , Shigenobu Shibata & Atsuo Takanishi (2013): A novel method to develop an animal
model of depression using a small mobile robot, Advanced Robotics, 27:1, 61-69
To link to this article: http://dx.doi.org/10.1080/01691864.2013.752319

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.

Advanced Robotics, 2013

Vol. 27, No. 1, 6169, http://dx.doi.org/10.1080/01691864.2013.752319

FULL PAPER
A novel method to develop an animal model of depression using a small mobile robot
Hiroyuki Ishiia,b*, Qing Shia, Shogo Fuminoc, Shinichiro Konnoc, Shinichi Kinoshitac, Satoshi Okabayashid, Naritoshi
Iidad, Hiroshi Kimurad, Yu Taharac, Shigenobu Shibataa,c and Atsuo Takanishia,b,c,e
a
Faculty of Science and Engineering, Waseda University, Tokyo, Japan; bResearch Institute for Science and Engineering, Waseda
University, Tokyo, Japan; cGraduate School of Science and Engineering, Waseda University, Tokyo, Japan; dFaculty of Letters, Art
and Science, Waseda University, Tokyo, Japan; eHRI, Waseda University, Tokyo, Japan

Downloaded by [188.229.29.180] at 01:14 13 February 2013

(Received 10 April 2012; accepted 15 August 2012)

Robotics is contributing to studies on animal behavior. Mobile robots are actually used as devices to give external stimulus to animals in several experiments. We consider that this approach can be applied to studies in psychic medicine. In
psychic medicine, all new drugs are evaluated in experiments using animal models of mental disorder before using it in
clinical practice. However, conventional animal models have some problems in the construct validity. The animal models
should be developed through the method which is consistently associated with the theory of the mental disorders while
many of conventional models had been developed by genetic manipulations or surgical operations on nerve system. We
considered that a novel animal model could be developed by stress exposure using a small mobile robot. We then
implemented this method to the experimental system which had been developed in our past study. An experiment was
conducted using the system, and the method was then veried. Therefore, we conclude that the animal model of depression developed by proposed method, exposing continuous attack by the robot in immature period and interactive attack
in mature period, can be a novel animal model of depression.
Keywords: mobile robot; bio-inspired design; animal behavior and mental disorder

1. Introduction
Recently, robotics is contributing to studies on animal
behavior in ethology, animal psychology, and behavioral
biology [16]. Several mobile robots with mimic feature
and motion of animals are developed. Use of these
robots offers novel methodologies to study response of
an animal to the stimulus form other individuals of same
species or other species. For instance, small mobile
robots are used to understand decision making of an
individual or a group of animals such as insects [1,2],
birds [3], shes [4], and rodents [5,6]. Some applications
to use mobile robots in animal breeding are also
proposed [3].
On the other hands, experiments on animal behavior
have been playing a very important role in psychic medicine [711]. All psychotropic drugs have been evaluated
in experiments with animal models of mental disorder
such as mice and rats before used in clinical practice [7].
These experiments are called drug screening test. The
animal models of mental disorder are living animals that
represent phenotypes of human patients with mental
disorders such as depression, schizophrenia, or anxiety
disorder. They are currently produced by genetic manipulation [12], surgical operation on the nerve system [13],
*Corresponding author. Email: hiroyuki@aoni.waseda.jp
2013 Taylor & Francis and The Robotics Society of Japan

or stressful environment [14]. In the drug screening tests,

a drug is administered into a model animal, and its effect
is then evaluated through behavior observation of the
animal. Rodents such as rats and mice are commonly
selected as experimental subjects in these experiments.
Recently, several new psychotropic drugs are developed
through these tests [10,11]. However, some researchers
have recently mentioned limitations of the conventional
animal models of mental disorders because of lack of the
validity. Three sets of criteria are proposed as the validity for assessing animal models of mental disorder:
predictive validity (performance in the test predicts performance in the condition being modeled), face validity
(phenomenological similarity), and construct validity
(theoretical rationale) [15,16]. However, few of the conventional models have all the three sets of criteria
together. Especially, there are few models that have the
construct validity [15,17,18]. In terms of the construct
validity, an animal model should be developed through
the procedure which is consistently associated with the
theory of the mental disorder. The current leading
hypothesis for mental disorders is stress-vulnerability
hypothesis proposed by Zubin [19,20]. He proposed that
an individual had unique vulnerability (strengths) for

Downloaded by [188.229.29.180] at 01:14 13 February 2013

62

H. Ishii et al.

dealing with stress, which was included in its biological

and psychological elements. Vulnerability of each individual is inherent in genetic level and also developed
through experiences during immature period such as
abuse or lost of parents. An individual is affected by the
mental disorder when he/she experiences much stress on
the vulnerability through events in his/her life such as
accidents or lost someone very special. According to this
hypothesis, animal models of mental disorder should
have stress-vulnerability and be affected by the disorder
when they receive much stress on the vulnerability.
Use of robot technology can break through these
limitations in the drug screening tests. Thus, the purpose
of this study is to build a novel method to develop an
animal model of mental disorder with the construct
validity using a small mobile robot. A small mobile
robot and an experimental setup were then developed,
and some experiments were conducted using the robot
and setup [21]. Through these experiments, we found
that a rat which had been exposed stress by the robot in
immature period exhibited lower activity than the normal
rat [22]. We considered it could be an animal model of
depression. However, this model still has some problems
in terms of the construct validity when it is compared
with the stress-vulnerability hypothesis. This model does
not consistent with the hypothesis because the disorder is
not triggered by environment process in mature period
while stress-vulnerability was developed in a rat by
stress exposure in immature period.
Therefore, we considered that it is possible to
develop an animal model of depression with the construct validity by exposing stress not only in immature
period but also in mature period as the trigger for the
disorder. We then built a method based on this concept
and implemented it in the experimental system that we
had developed in our past studies. An experiment is
performed to verify this method. In the experiment, two
different ways of stress exposure were prepared to
nd the one to induce much stress in a rat. In this paper,
a small mobile robot is shortly described in Chapter 2
and control system for it is described in Chapter 3. The
experiment is described in Chapter 4 and discussion for it
is described in Chapter 5. Chapter 6 is the conclusion.
2. Small mobile robot WR-3
2.1. Mechanism
We developed a small mobile robot WR-3 as shown in
Figure 1 [21]. It was designed to interact with a rat in
the manner of interactions between rats. Therefore, its
size and locomotion performance are almost equal to a
mature rat as shown in Table 1.
WR-3 has 14 active degrees of freedom (DOFs) as
shown in Figure 2. Twelve of the DOFs are used to
mimic body motions of a rat such as rearing (pitch

Figure 1. WR-3 (front) and a mature rat (back).

Table 1. Specications of WR-3.

Size (mm)
Weight (g)
Max speed (m/s)
Operation time (min)

70  240  90
1000
1.0
30

Figure 2. DOF arrangement of WR-3. Total 14 DOFs: roll and

pitch in the neck, two pitch in each leg, pitch and yaw in the
waist and two wheels.

motion of the waist) or grooming (yaw motion of the

waist and neck, with pitch motion of the fore legs).
DOFs in a pitch and yaw in the neck are driven by
Shape Memory Alloy (SMA) wires. Each DOF in the
fore leg and hind leg is driven by a DC servo motor.
Each DOF in a pitch and yaw in the waist joint is also
driven by a DC servo motor. In addition to these 12
DOFs, two active wheels for locomotion in x-y plain are
implemented at the hip. Each of these two DOFs is driven by a DC motor, and its velocity is servo-controlled.
Therefore, locomotion of WR-3 receives the non-holonomic constraint.

Advanced Robotics
2.2. Control circuit and power supply

Downloaded by [188.229.29.180] at 01:14 13 February 2013

A control circuit board which is originally designed for

WR-3 is implemented in the robot. This circuit board
consists of a microcontroller (STM32F103, STMicro
electronics Ltd.), a Bluetooth wireless communication
module, H-bridge motor drivers, and drivers for the
SMA wires. Only low-level control is handled by the
circuit board while high-level control such as the pattern
generation is handled by the PC. The circuit board
receives instructions from the PC via the wireless communication module, and controls angle of each joint and
velocities of driving wheels according to the instructions.
A Li-ion battery is implemented in WR-3. WR-3
keeps its operation for more than 30 min with fully
charged battery.
3. Control system of WR-3
3.1. Locomotion control by visual feedback
We developed behavior control software of WR-3 which
enabled the robot to locomote autonomously and to
interact with a rat in the open eld as shown in Figure 3.
This software consists of the position calculator (image
processing), visual feedback controller, and behavior
generator as shown in Figure 4. Pictures of WR-3 and a

Figure 3. Picture of a rat and WR-3 in the open eld. Pwr, the
position of WR-3 and Prat, the position of the rat are calculated
using image processing technique. dr-wr, distance between the
rat and WR-3 is also calculated.

rat in the open eld are taken by a charge coupled

device (CCD) camera which is placed above the open
eld. These pictures are sent to the software, and the
position calculator then calculates positions of WR-3 and
a rat by image processing. The position of WR-3 Pwr is
calculated using color marker (yellow and blue) put on
the body, and that of the rat Pr is calculated using its
body color (white). Movement distance of the rat and
distance from WR-3 can be calculated from the position
data.
The target position of WR-3 Pwr tg is generated by the
behavior generator based on the algorithm implemented
in it as described in the next section. The visual
feedback controller then calculates direction and distance
between the target position and current position of
WR-3. According to the distance and direction, the
visual feedback controller generates instructions of
locomotion for WR-3.

3.2. Behavior algorithms

Three different behavior generation algorithms, chasing,
continuous attack, and interactive attack are prepared
as shown in Figure 5. In the experiment, the experimenter can select one from these three algorithms
according to the experimental design.
When WR-3 is controlled according to the algorithm
of chasing, WR-3 keeps distance from the rat less than
Dwr-r th c . Dwr-r th c is 350 mm in the experiment
described in Chapter 4, while length of WR-3 is
240 mm. Therefore, WR-3 rarely hits the rat on the
body. If the rat does not move, WR-3 turns right and
left to keep the distance from the rat. When WR-3 is
controlled according to the algorithm of continuous
attack, WR-3 keeps attacking to the rat. To provide
attack to the rat, WR-3 keeps distance from the rat less
than Dwr-r th . Dwr-r th is 150 mm in the experiment.
Therefore, WR-3 keeps hitting the rat on the body.
When WR-3 is controlled according to the algorithm of
interactive attack, WR-3 starts an attack sequence to
the rat immediately after the rat moves more than
Dr mov th . Dr mov th is 50 mm in the experiment. WR-3
keeps the attack sequence for 5 s and stops movement
after these 5 s. During the attack sequence, WR-3 is

Figure 4. Control system of WR-3. Prat is the position of the rat and Pwr is the position of WR-3. Pwr

of WR-3.

63

tg

is the target position

H. Ishii et al.

Downloaded by [188.229.29.180] at 01:14 13 February 2013

64

Figure 5. Behavior algorithms prepared for the experiment.

controlled according to the same algorithm of continuous attack.

4. Experiment
4.1. Subject
F344 rats are selected as experimental subjects. Twentyfour immature rats (three weeks old) are prepared. The
rats are divided into two groups as shown in Table 2.

4.2. Procedure
(i) Stress exposure in immature period.
Each rat in both groups A and B receives continuous attack (see Chapter 3) by WR-3 as the stress
exposure. Attacks by WR-3 are exposed for 10 min
a day for ve days from the day when the rat
becomes three weeks old. After these ve days,
each rat is bred in a small cage individually.

Advanced Robotics

65

Table 2. Experimental condition of rats in group A and B.

Group A
Number of rats
Stress exposure in immature
period (3 weeks old)
Behavior assessment before
stress exposure in mature
period (8 weeks old)
Stress exposure in mature
period (9 weeks old)

Downloaded by [188.229.29.180] at 01:14 13 February 2013

Behavior assessment after

stress exposure in mature
period (10 weeks old)

Group B

12
12
Continuous attack by WR-3
Open-eld test, robot chasing
test
Continuous
Interactive
attack by
attack by
WR-3
WR-3
Open-eld test, robot chasing
test

(ii) Behavior assessment before stress exposure in

mature period
Two different behavior tests, the open-eld test
and robot chasing test, are performed when the
rat becomes eight weeks old. In the open-eld
test, a rat is released into the open eld, and its
total movement distance in 10 min is measured as
its activity. In the robot chasing test, each rat is
also released into the open eld with WR-3
which is controlled according to the algorithm for
chasing (see Chapter 3). Total movement distance of the rat in 10 min is measured as its
activity too.
(iii) Stress exposure in mature period
Each rat in group A receives continuous attack by
WR-3. Each rat in group B receives interactive
attack (see Chapter 3) by WR-3. Attacks by WR-3
are exposed for 10 min a day for ve days from the
day when the rat becomes nine weeks old.
(iv) Behavior assessment after stress exposure in mature
period
Two behavior tests which have been performed in
eight weeks old are performed again when the rat
becomes 10 weeks old.

4.3. Result
Activities in the behavior tests in both before and after
the stress exposure in mature period are shown in
Figure 6. Activities during stress exposure in mature
period are shown in Figure 7. Statistical analyses (t-test)
are conducted to nd signicant differences of activities
in robot chasing test between before and after second
stress exposure. Another statistical analysis (t-test) is
conducted to nd a signicant difference of activities in
robot-chasing test after second stress exposure between
group A and B. We consider that there is a signicant
difference if p value is less than 0.05 between two data
sets.

Figure 6. Experimental result in behavior tests.

Figure 7. Activities during stress exposure in mature period.

5. Discussion
Construct validity of the depression model rats can be
conrmed by comparing the experimental procedure,
which is described in Chapter 4, with the stress-vulnerability hypothesis. The basic conguration of this hypothesis
is that an individual is affected by the mental disorder
when he/she experiences much stress on the vulnerability
through events in his/her life. In the experiment, the
vulnerability was developed in rats through attacks by

Downloaded by [188.229.29.180] at 01:14 13 February 2013

66

H. Ishii et al.

WR-3 during immature period. As conrmed in the past

experiment [22], attacks by a mobile robot induce severe
stress in immature rats. These stresses are caused by
psychological fear and physical pain. The stress on the
vulnerability was induced in rats through attacks by WR-3
during mature period. In human cases, levels of stress
induced by life events are different depending on the type
of the event [23]. Therefore, three different behavior
algorithms of WR-3 were prepared to induce different
level of stress in the experiment. One is chasing by WR-3,
and the others are interactive attack and continuous attack
by WR-3.
Chasing by WR-3 induces no pain but fear in rats.
We then considered that the level of stress induced
through chasing by WR-3 is less than those by other two
algorithms, and that robot chasing test can be a method
to assess depression level like forced swimming test [24]
or fear-conditioning test [25]. The forced swimming test
is conducted in a pool, and depression level of a rat is
assessed by immobility time after it is released into the
pool. The fear conditioning test is conducted in a box,
and a rat is exposed strong aversive stimulus such as
electric shock in the box. The rat is then released into
the box next day, and depression level is assessed by
immobility time. The basic idea in these two tests is that
a rat becomes immobile when it falls into depression
under the stressful environment. Therefore, activity of
the rat with experience of robot attack in the robot
chasing test represents depression level.
Both interactive attack and continuous attack by
WR-3 induce pain and fear in rats. The levels of stress
might be different from each other. Signicance and
effect of each way of attack by WR-3 can be discussed
through an analysis on experimental result. Before comparing the effects of the attacks by WR-3, spontaneous
activity should be compared between groups. Open-eld
test is a well-known method to assess spontaneous activity of a rat [26]. No large difference is found in activities
between before and after the stress exposure in mature
period both in groups A and B. Therefore, spontaneous
activities of rats are not different between before and
after the stress exposure, and those are not different
between groups A and B either. On the other hands, we
found several signicant differences in activities in the
result of robot chasing test. Based on this consideration,
depression level are not different between groups A and
B before the stress exposure in mature period, while a
signicant difference is conrmed between them after
the stress exposure. A signicant difference is also
conrmed between before and after the stress exposure
in group A. Therefore, the interactive attack can be an
impact on the stress vulnerability while continuous attack
cannot be that.
It is very interesting that a rat in group B receives
much stress than a rat in group A, while a rat in group

A receives much attack by WR-3 than a rat in group B.

The rat in group A receives attack just after it starts
moving, while the rat in group B receives attack continuously. It can be explained by the idea that behavior
inhibition induces large stress in the rat. The reason why
the rats in group A exhibited high activities which were
representation of low depression level could be explained
by the theory of exposure therapy [27]. The basic
concept of the exposure therapy is that exposing stressor
continuously under safe condition develops tolerance for
it in the individual. As shown in Figure 7, activities of
rats in group A increased day by day during the stress
exposure in mature period. This result agrees with the
process of the exposure therapy. Therefore, attack
itself cannot be a stressor for mature rats, while it can
be a strong stressor to develop vulnerability for
immature rats.
Therefore, we conclude that a depression model
animal can be developed by exposing continuous attack
by the robot in immature period and interactive attack in
mature period. The process to develop this model agrees
with the theory of the stress-vulnerability hypothesis.
Behavioral phenotype also agrees with depression. Thus,
we consider that the rat treated by this method can be an
animal model of mental disorder with the construct
validity and face validity. There might be some
objections to this consideration. For instance, validity of
the robot chasing test as a stress evaluation test was not
experimentally conrmed. Additional evidences to
answer these objections can be obtained by evaluating
depression level of the animal models through conventional behavior tests such as forced swimming test and
fear conditioning test.
6. Conclusion
We proposed a method to develop a novel animal model
of depression and veried it through an experiment. In
the experiment, the face validity and construct validity
are conrmed. Thus, we conclude that the animal model
of depression developed by proposed method, exposing
continuous attack by the robot in immature period and
interactive attack in mature period, can be a novel
depression model which has some advantages over the
conventional models. Its predictive validity should be
veried in next experiment. After that, it can be used in
the drug screening. In addition, this study suggests large
potential use of a mobile robot in experiments on animal
behavior for understanding mechanism of mental disorder. The experimental result suggests that the interactive
attack and the continuous attack have different effects on
rats. Using this methodology, it is possible to make a
theory of how external stimulus induces stress in individuals. It can be a new research paradigm in psychic medicine.

Advanced Robotics

67

Acknowlegments
The experiment with animals has been approved by ethical
committee for animal experiments in Waseda University. This
work was supported by JSPS KAKENHI Grant Number
21760207 and High-Tech Research Center Project for Private
Universities: matching fund subsidy from MEXT (Ministry of
Education, Culture, Sports, Science, and Technology). This
works was also supported by GCOE; Global Robot Academia,
ASMeW, and SolidWorks, Japan.

Downloaded by [188.229.29.180] at 01:14 13 February 2013

Notes on contributors
Hiroyuki Ishii received his PhD degree in
Biomedical
Engineering
at
Waseda
University, Japan, in 2007. He is currently
an assistant professor at Waseda Research
Institute for Science and Engineering,
Waseda University. His research interests
are focused on applications of robot
technology in studies on animal behavior.
He is currently interested in the
development of mental disorder model animals using a small
mobile robot.
Qing Shi received his PhD degree in
Biomedical
Engineering
at
Waseda
University, Japan, in 2012. He is currently a
research associate at GCOE Global Robot
Academia of Waseda University. His
research interests are focused on the
development of rat-inspired autonomous
robots for behavior analysis of rats,
especially involving system integration,
behavior recognition, and motion planning.
Shogo Fumino received his BS degree in
Mechanical Engineering from Waseda
University in Tokyo, Japan in 2008. Since
2009, he became a masters student at
Waseda University in Tokyo, Japan,
focusing
on
the
development
of
experimental systems for rats to expose
stress using robot. His research interests are
embedded system, mechanical design,
robotrat interaction, and mechanism of mental disorder.
Shinichiro Konno received his BS degree
in Mechanical Engineering from Waseda
University, Tokyo, Japan in 2010. Since
2010, he became a masters student at
Waseda University in Tokyo, Japan,
focusing on the development of the ratinspired robot for interaction experiment
with rats. His research interests are
mechanical design, animal experiment, and
robotrat interaction.

Shinichi Kinoshita received his BS degree

in Mechanical Engineering from Waseda
University, Tokyo, Japan in 2011. Since
2011, he became a masters student at
Waseda University in Tokyo, Japan,
focusing on the development of the ratinspired robot for interaction experiment
with rats. His research interests are
mechanical design, robot control, image
processing, animal experiment, and robotrat interaction.
Satoshi Okabayashi received his BA and
MA degrees in Psychology from Waseda
University in Tokyo, Japan in 2005 and
2007, respectively. Since 2007, he became a
PhD student at Waseda University focusing
on the learning behavior in rodents for
psychological studies. Since 2010, he
became a research associate at Waseda
University Faculty of Letters, Arts and
Sciences. His research interests are concerned with animal
psychology, ethology, and behavior analysis.
Naritoshi Iida received his BA and MA
degrees in Psychology from Waseda
University in Tokyo, Japan in 1994 and
1997, respectively. He serves as a university
lecturer at Waseda University in Tokyo,
Japan since 2004. His research interests are
concerned
with
animal
psychology,
behavior analysis, and psychology of
learning. Especially, he is currently focusing
on the response suppression by punishment contingencies.
Hiroshi Kimura received his MA degree in
Psychology from Waseda University,
Tokyo, Japan in 1967. He specialized in the
psychology of learning and behavior
analysis. He focused on the research on the
process of behavior modication by using
rats as the experimental subjects. He was
appointed as a professor emeritus in 2012
after having served as a lecturer, associate
professor, and professor of psychology at Waseda University.
Yu Tahara received his BS and MS
degrees in school of science and
engineering, Waseda university, Japan. In
2010, he became a PhD student at Waseda
University, focusing on circadian rhythms in
mice by physiological and pharmacological
methods.

68

H. Ishii et al.

Downloaded by [188.229.29.180] at 01:14 13 February 2013

Shigenobu Shibata received his BS, MS,

and PhD degrees all in Pharmaceutical
Sciences from Kyushu University, Japan, in
1976, 1978, and 1981, respectively. He is
currently a professor at the School of
Advanced Science and Engineering, Waseda
University.
He
has
specialized
in
physiology, especially circadian rhythms in
animals.
Atsuo Takanishi received his BS, MS, and
PhD degrees all in Mechanical Engineering
from Waseda University, Japan, in 1980,
1982, and 1988, respectively. He is
currently a professor at the Department of
Modern Mechanical Engineering, Waseda
University, and the director of HRI (The
Humanoid Robotics Institute), Waseda
University. He is a member of Robotics
Society of Japan (a board member in 1992 and 1993), Japanese
Society of Biomechanisms, Japanese Society of Mechanical
Engineers, Japanese Society of Instrument and Control
Engineers and Society of Mastication Systems (a major board
member from 1996 to current), IEEE, and other medicine- and
dentistry-related societies in Japan.

References
[1] Halloy J, Sempo G, Caprari G, Rivault C, Asadpour M,
Tche F, Sad I, Durier V, Canonge S, Am JM, Detrain
C, Correll N, Martinoli A, Mondada F, Siegwart R,
Deneubourg JL. Social integration of robots into groups
of cockroaches to control self-organized choices. Science.
2007;318:11551158.
[2] Colot A, Caprari G, Siegwart R. InsBot: design of an
autonomous mini mobile robot able to interact with
cockroaches. Proceedings of the 2004 IEEE International
Conference on Robotics and Automations; 2004.
[3] Gribovskiy A, Halloy J, Deneubourg J-L, Bleuler H,
Mondada F. Towards mixed societies of chickens and
robots. Proceedings of the 2010 IEEE international Conference on Intelligent Systems and Automations; 2010.
[4] Swain DT, Couzin ID, Leonard NE. Real-time feedbackcontrolled robotic sh for behavioral experiments with
sh schools. Proc. IEEE. 2012;100:150163.
[5] Ishii H, Ogura M, Kurisu S, Komura A, Takanishi A, Iida
N, Kimura H. Experimental study on task teaching to real
rats through interaction with a robotic rat. In: Lecture
notes in articial intelligence, Vol. 4095. Springer; 2006.
p. 643654.
[6] Laschi C, Mazzolai B, Patane F, Mattoli V, Dario P, Ishii
H, Ogura M, Kurisu S, Komura A, Takanishi A. Design
and development of a legged rat robot for studying animalrobot interaction. Proceedings of the First IEEE/
RAS-EMBS International Conference on Biomedical
Robotics and Biomechatronics; 2006.

[7] Van der Staay FJ. Animal models of behavioral dysfunctions: basic concepts and classications, and an evaluation
strategy. Brain Res. Rev. 2006;52:131159.
[8] Lopez-Munoz F, Alamo C. Monoaminergic neurotransmission: the history of the discovery of antidepressants from,
1950, until today. Curr. Pharm. Des. 1950;24:15631586.
[9] Griebel G, Moindrot N, Aliaga C, Simiand J, Soubri P.
Characterization of the prole of neurokinin-2 and
neurotensin receptor antagonists in the mouse defense test
battery. Neurosci. Biobehav. Rev. 2001;25:619626.
[10] Louis C, Stemmelin J, Boulay D, Bergis O, Cohen C,
Grieble G. Additional evidence for anxiolytic- and antidepressant-like activities of saredutant (SR48968), an antagonist at neurokinin-2 recepter in various rodent-models.
Pharmacol. Biochem. Behav. 2008;89:3645.
[11] Griebel G, Perrault G, Soubri P. Effects of SR48968, a
selective non-peptide NK2 receptor antagonist on
emotional processes in rodents. Psychopharmacology.
2001;158:241251.
[12] Hiramoto T, Kang G, Suzuki G, Satoh Y, Kucherlapati R,
Watanabe Y, Hiroi N. Tbx1: identication of a 22q11.2
gene as a risk factor for autism spectrum disorder in a
mouse model. Hum. Mol. Genet. 2011;20:47754785.
[13] Kelly JP, Wrynn AS, Leonard BE. The olfactory bulbectomized rat as a model of depression: an update. Pharmacol.
Ther. 1997;74:299316.
[14] Ayensu WK, Pucilowski O, Mason GA, Overstreet DH,
Rezvani AH, Janowsky DS. Effects of chronic mild stress
on serum complement activity, saccharin preference, and
corticosterone levels in Flinders lines of rats. Physiol.
Behav. 1995;57:165169.
[15] Willner P. Validation criteria for animal models of human mental disorders: learned helplessness as a paradigm case. Prog.
Neuropsychopharmacol. Biol. Psychiatry. 1986;10:677690.
[16] Willner P. Behavioral models in psychopharmacology
theoretical, industrial and clinical perspectives. Cambrige:
Cambrige University Press; 1991.
[17] Weiss JM. Dose decreased sucrose intake indicate loss of
preference in CMS model? Psychopharmacology.
1997;134:368370.
[18] Nestler EJ, Hyman SE. Dose decreased sucrose intake
indicate loss of preference in CMS model? Nat. Neurosci.
2010;13:11611169.
[19] Zubin J, Spring B. Vulnerability: a new view on
schizophrenia. J. Abnorm. Psychol. 1977;86:103126.
[20] Gispen-de Wied1 CC, Jansen1 LMC. The stress-vulnerability hypothesis in psychotic disorders: focus on the stress
response systems. Curr. Psychiatry Rep. 2002;4:166170.
[21] Shi Q, Ishii H, Miyagishima S, Konno S, Fumino S, Takanishi
A, Okabayashi S, Iida N, Kimura H. Development of a
hybrid wheel-legged mobile robot WR-3 designed for the
behavior analysis of rats. Adv. Robotics. 2011;25:22552272.
[22] Ishii H, Qing S, Masuda Y, Miyagishima S, Fumino S,
Takanishi A, Okabayashi S, Iida N, Kimura H, Tahara Y,
Hirao A, Shibata S. Development of experimental setup to

Advanced Robotics

Downloaded by [188.229.29.180] at 01:14 13 February 2013

create novel mental disorder model rats using small

mobile robot. Proceedings of 2010 IEEE/RSJ International
Conference on Intelligent Robots and Systems; 2010.
[23] Holmes TH, Rahem RH. The social readjustment rating
scale. J. Psychosom. Res. 1967;11:213218.
[24] David DJP, Renard CE, Jolliet P, Hascot M, Bourin M. Antidepressant-like effects in various mice strains in the forced
swimming test. Psychopharmacology. 2003; 166:373382.
[25] Cuppini R, Bucherelli C, Ambrogini P, Ciuffoli S, Orsini
L, Ferri P, Baldi E. Age-related naturally occurring
depression of hippocampal neurogenesis does not affect
trace fear conditioning. Hippocampus. 2006;16:141148.

69

[26] Foa EB, Dancu CV, Hembree EA, Jaycox LH, Meadows
EA, Street GP. A comparison of exposure therapy, stress
inoculation training, and their combination for reducing
posttraumatic stress disorder in female assault victims. J.
Consult. Clin. Psychol. 1999;67:194200.
[27] Choleris E, Thomas AW, Kavaliers M, Prato FS. A
detailed ethological analysis of the mouse open-eld test:
effects of diazepam, chlordiazepoxide and an extremely
low frequency pulsed magnetic eld. Neurosci. Biobehav.
Rev. 2001;25:235260.