Introduction of a robot patient into dental education


Takeshi Tanzawa
Department of Orthodontics
School of Dentistry
Showa University
Tokyo 145-8515
Tel: +81-3-3787-1151
Fax: +81-3-3784-6641


In recent years, with the increasing social awareness of safety in medical practice, improving clinical skills has become very important, especially for recently graduated dentists. Traditionally, mannequins have been used for clinical skill training, but a mannequin is quite different from a real patient because they have no autonomous movement or conversational ability. This indicates that pre-clinical simulation education is inadequate. We have, therefore, developed a robot patient that can reproduce an authentic clinical situation for dental clinical training. The robot patient, designed as a full-body model with a height of 157 cm, has eight degrees of freedom in the head and the ability to perform various autonomous movements. Moreover, saliva secretion and conversation with the trainee can be reproduced. We have introduced the robot patient into an objective structured clinical examination targeted at fifth-grade students in our dental school to evaluate their skills in cavity preparation, whilst considering the safety of the treatment. As a result, many of the students were able to deal appropriately with a patient’s unexpected movement. Moreover, results of a questionnaire survey showed that almost all the students recognised the educational value of the robot patient especially for ‘risk management’, and they preferred the robot patient to traditional mannequins. Practical application of the robot patient in dental clinical education was evaluated through the experiences of the fifth-grade students, which showed the effectiveness of the robot patient in the dental field.


A problem for dental education is that the use of human subjects in clinical training is ethically difficult. Moreover, the use of human subjects in clinical examinations is now absolutely inappropriate owing to the impossibility of standardising human subjects anatomically, physiologically, pathologically and psychologically (1). On the other hand, the traditional static mannequin’s jaw and phantom head are quite different from those of a real human being. The main disadvantages identified in the use of mannequins are the following: (i) lack of clinical authenticity: they are unrealistic compared with authentic clinical situations in the completion of routine clinical tasks; (ii) lack of communication skill testing: there is no opportunity for the routine communication skills of interviewing such as greetings, and also those associated with clinical tasks such as the explanation of a clinical procedure; and (iii) lack of patient management/behavioural problems: there are certain routine patient management problems in dentistry, such as dealing with apprehension, restlessness and anxiety, or managing an unexpected movement such as a vomiting reflex, coughing and tongue thrusting (2). Based on these points, it is not obvious that dental students receive sufficient simulation education. Therefore, care provided by students may require more time and more visits in general, compared with services provided by dental graduates (3). Moreover, such slow performance may in itself produce emotional and physical distress. This is not, therefore, patient-centred care. Although lower-income patients may be willing to trade time for low-cost care in student clinics, all patients have an equal right to receive the most proficient treatment.

To overcome these deficiencies, in recent years, simulation using advanced technology has become prominent in the dental field (4–7). The role of simulators has been recognised as an important aspect of training in the health field as it supports and improves patient safety (8). In fact, many undergraduate dental students tend to prefer the educational environment of the simulator to the traditional bench and mannequin laboratory (9).

The authors previously developed a mastication robot to investigate the cause of mastication dysfunction and to support recovery (10). Later, the mastication robot was further developed into a robot patient able to reproduce authentic clinical situations precisely (11, 12). The robot patient has since been improved to endure continuous use by many trainees. The reasons for the development of the robot patient are the following: (i) to reproduce the oral maxillofacial anatomy for dental treatment; (ii) to present a full body; (iii) to reproduce autonomous movement through robotics; and (iv) to enable conversation with the trainee.

First, we established the competency of ‘risk management during treatment’ for undergraduate students and implemented ‘cavity preparation using the robot patient’ as the objective structured clinical examination (OSCE).

The aim of this study is to investigate the efficiency of the robot patient and its role in dental clinical education.

Materials and methods

Development of the robot patient

We developed the robot patient jointly with Tmusk Co., Ltd. (Fukuoka, Japan), a company manufacturing robots.

The specifications were as follows:

  • Outer appearance

  • The robot patient was designed as a full-body model with a height of 157 cm. Vinyl chloride elastic material was used to reproduce the epidermis (Kyoto Kagaku, Physiko). The robot patient was assigned a female form to allow for consideration of avoiding contact to her chest.

  • The robot patient was fitted with clothes, a wig and other ornaments not only to improve her appearance but also to consider the possibility of spillage of a drug solution or impression material on the clothing (Fig. 1).

  • Degrees of freedom and movement method

  • The robot patient comprises autonomous moving parts with eight degrees of freedom (eyelids, eyeballs, jaw, tongue and neck) (Fig. 2) and passive moving parts (shoulder, arm, elbow, waist, hip joint, knee and ankle). The autonomous moving parts are driven by a compressor (Jun-Air 6-15-FTD) located externally to the robot patient.

  • Conversation

  • The robot patient is able to talk to a trainee during the medical examination and treatment using conversation software (Tmusk Co., Ltd.).

  • Oral cavity

  • Silicon is used to reproduce the soft tissue of the oral mucosa, tongue and uvula (Kyoto Kagaku). Regarding the tongue, autonomous movement is triggered when contact is made, and deglutition can be reproduced. Because a pressure sensor is built into the uvula, a simulated vomiting reflex occurs when contact is made. A dentition model (Nissin Dental Products D18FE-500E), used in daily practice, has been installed in the robot patient. However, its shape has been greatly modified to fit the oral volume similarly to a human being and to enable the model to be detached easily (Fig. 3).

  • Saliva secretion

  • Saliva is reproduced by water, which is drawn up by a rotating electric motor, and secreted from the parotid region.

  • Control program

  • The autonomous moving parts are controlled by a personal computer (HP ProBook 4510s). The control program, running on Windows XP (Microsoft Corp., Redmond, WA, USA), communicates with the control equipment of the moving parts and the control panel via an Ethernet LAN connection and enables control of the robot patient. Blinking, ocular movement, tongue movement and mouth movement are automatically controlled by the program at all times.

  • The instructor can control the patient’s movement at a touch of the panel interface on the controller. Therefore, unexpected situations during the dental procedure, such as ‘Shaking neck’, ‘Coughing’, ‘Tongue thrusting’ and ‘Mouth fatigue’, can be reproduced. In addition, the robot patient can deny or affirm what the trainee has said and is able to open and close her mouth, or change the direction of her neck on instruction from the trainee.

Figure 1.

 Robot patient.

Figure 2.

 Degrees of freedom. The head of the robot patient has 8 degrees of freedom (eyelids, eyeballs, jaw, tongue and neck).

Figure 3.

 Close-up of the open mouth.

Design and implementation of the OSCE

In advance of the OSCE, we established the competency for risk management taking into account the patient’s unexpected movement under treatment.

We used 88 fifth-grade students in the Department of Dentistry at Showa University as subjects to implement the OSCE. The task of ‘cavity preparation using the robot patient’ is part of the OSCE, which includes 11 stations. This station was implemented in the skill laboratory of Showa University dental hospital and was done using two affiliate straight methods. Therefore, we used two robot patients with the same specification for the station. Two preceptors scored each candidate using the examiner’s score sheet (Table 1). In the examination, the duration of the task inspection is 1 min and that of the task implementation is 10 min.

Table 1.   Examiner’s score sheet
Station: cavity preparation applied to the robot patient
Student’s Name: Student No: 
Instructions to examiner: please check appropriate score
Check list:Score
a. Prior explanation of the clinical procedure 10
b. Appropriate positioning under treatment 10
c. Cavity preparation on occlusal surface 10
d. Preparing marginal bevel 10
e. Correspondence to ‘Shaking neck’210
f. Correspondence to ‘Coughing’210
g. Correspondence to ‘Tongue thrusting’210
h. Correspondence to ‘Mouth fatigue’210
i. Consideration of cleanliness 10
j. Consideration of safety 10
k. Consideration of patient’s feelings 10
Date: Examiner’s Signature: 

The candidates were asked to perform a class I inlay cavity preparation task on the upper left first molar. During the cavity preparation, the robot patient made each of four unexpected movements in turn. The first of these is ‘Shaking neck’ to the left side in response to pain. The second movement is ‘Coughing’ because of the collection of saliva and water from the turbine, and the third is ‘Tongue thrusting’ for deglutition. Finally, ‘Mouth fatigue’ involves closing the mouth gradually owing to fatigue of the jaw. These movements are initiated by the instructor using the interface on the controller.

We mainly evaluated the student’s communication ability with the patient, skill of cavity preparation in a near authentic clinical situation, reaction to the patient’s unexpected movement and consideration of the patient’s entire body and feelings, such as pain or unpleasantness. We provided 11 checkpoints on the examiner’s score sheet. For items a, b, c, d, i, j and k on the checklist, a score of 1 denotes ‘Carried out satisfactorily’, whilst a score of 0 denotes ‘Not attempted’. For items e, f, g and h on the checklist, a score of 2 means ‘Stopped the operation safely and discussed the risk of unexpected movement during the treatment with the patient’; a score of 1 means ‘Stopped the operation safely, but without any discussion or discussed the risks, but without stopping the operation’; and a score of 0 means ‘Attempted neither’.

The grading method calculated the average of the two preceptor’s scores for each student and then converted it to a mark out of 100. The final average score was calculated from the average of 88 students’ scores.

Student questionnaire

To assess student responses to the robot patient, a questionnaire was completed after the examination. Students were asked to provide feedback on the reproducibility of the robot patient in terms of ‘outer appearance’, ‘oral cavity’, ‘movement’ and ‘conversation’; the educational value of the robot patient for ‘cavity preparation’, ‘risk management’ and ‘communication skills’; and the importance of the robot patient. They were asked to rate each item on a scale from one to three (where 1 = poor and 3 = excellent) and also to provide open-ended responses about the educational advantages and disadvantages of using the robot patient and the training they would like to do using the robot patient in the future.


Table 2 shows the average scores for each of the checkpoints. Almost all the candidates gave an appropriate explanation of the clinical procedure to the patient without using technical terms. They were also able to complete the cavity preparation on an occlusal surface from an optimal position at the patient’s head between 2 o’clock and 10 o’clock. Of the four unexpected movements, reactions to ‘Shaking neck’, ‘Coughing’ and ‘Mouth fatigue’ earned 80 points or more. Several candidates stopped the turbine operation before taking it out of the mouth and explained the risk of unexpected movement during treatment to the patient. Moreover, they were considerate of pain or unpleasantness to the robot patient. However, reaction to ‘Tongue thrusting’ was slightly lower than the other three movements, because some candidates either did not notice it or disregarded it. Regarding ‘consideration of cleanness during treatment’, many candidates recognised the distinction between a ‘clean operating zone’ and an ‘unclean operating zone’. Answers to selected questions are shown in Table 3.

Table 2.   Average score for each item on the checklist
Check listAverage scoreSD
  1. SD, standard deviation.

a. Prior explanation of the clinical procedure9813
b. Appropriate positioning under treatment9318
c. Cavity preparation on occlusal surface1000
d. Preparing marginal bevel8634
e. Correspondence to ‘Shaking neck’8621
f. Correspondence to ‘Coughing’8424
g. Correspondence to ‘Tongue thrusting’7231
h. Correspondence to ‘Mouth fatigue’8023
i. Consideration of cleanliness7524
j. Consideration of safety7821
k. Consideration of patient’s feelings7021
Table 3.   Student responses to the robot patient
QuestionsDistribution of questionnaire responses (%)
  n = 88 (100%)
Reproducibility of the robot patient
 Outer appearance264529
 Oral cavity472528
Educational value of the robot patient
 Cavity preparation8992
 Risk management training9550
 Communication skills583012
Importance of the robot patient
 Dental educational effect (compared with mannequin)9514
 Necessity of the robot patient in dental education8884

Based on the responses to the selected questions on the reproducibility of the robot patient, ‘conversation’ received a most favourable review from 53% of the students. Forty-seven per cent of the students responded positively to the reproducibility of ‘oral cavity’ and ‘movement’. However, with regard to the ‘outer appearance’, positive opinion was only 26%.

Regarding the educational value of the robot patient, 95% of the students recognised its usefulness in risk management training, whilst 89% of the students recognised its usefulness in cavity preparation. On the other hand, 58% of the students considered the robot patient to be effective in the training of communication skills. Compared with traditional mannequins, 95% of the students preferred the robot patient. Moreover, 88% of the students considered the introduction of the robot patient into dental education to be necessary.

The open-ended responses indicated that many candidates appreciated the movement of the robot patient as an educational advantage, and they agreed that risk management during treatment is important. Some candidates, however, commented negatively on the movement, because it prevented them from concentrating on the cavity preparation. Practical training they wished to perform on the robot patient included extractions, caries treatment and dealing with medical emergencies such as shock in the dental clinic.


Generally, an occupation that affects a person’s life or health benefits from simulation education (13). The advantage of simulation education is the ability to acquire appropriate skills to deal with unexpected situations, like an accident or emergency, without physical danger. However, a simulation environment with advanced reproducibility has not previously existed in the dental field. The need for a simulated clinical environment is timely to provide the learning platform for dental students to develop and hone their clinical management skills. This is the reason we developed the robot patient, which is able to reproduce many clinical situations.

The robot patient is a full-body model able to be positioned on a dental chair. The use of a dental chair unit is useful for trainees to recognise the distinction between a ‘clean operating zone’ and an ‘unclean operating zone’. Hair and clothes are necessary to ensure that the candidate considers not only the oral cavity but also the whole body. Moreover, the robot patient is able to mimic the movements and conversation of an actual patient through computer control. This sense of providing an actual treatment provides a certain degree of stress for the candidates.

The objective structured clinical examination (OSCE) was initially advocated in 1975 by Harden et al. (14) and has since been introduced throughout the world as a method of objective evaluation of clinical skills. However, although the OSCE is useful in the examination of diagnostic, interpretation and treatment planning skills, it has apparent limitations in the examination of invasive operative procedures (2). Therefore, there is a need to develop and evaluate objective methods for assessment of invasive clinical operative procedures. To address this need, we introduced the robot patient into the OSCE. The major advantages of the robot patient are that the trainees can experience the danger or difficulty of treatment caused by a patient’s movement and they can learn how to handle such situations. Moreover, their communication ability can be evaluated through the conversation dimension. Besides, the objectivity of the examination is not lost, because the same situation can be reproduced repeatedly. Additionally, in this OSCE, we reproduced unexpected movements that generally occur during cavity preparation, and we evaluated whether the candidate could stop the turbine operation before taking it out of the mouth, and whether the risks of the treatment were discussed with the patient.

Based on the results, many of the candidates were able to handle the unexpected movement safely, although some took the turbine out of the mouth whilst it was still rotating without paying attention to the tip of the bur and could not explain to the patient the risk of unexpected movements during the treatment. In particular, attention to ‘Tongue thrusting’, which shows the largest range of values in terms of the standard deviation, needs to be further improved. In dental environments, compared with many other medical environments, sharp injuries are more likely to occur because of the small operating field, frequent patient movement and the variety of sharp dental instruments used in everyday practice (15). Therefore, adequate pre-clinical simulation education is necessary to prevent these accidents.

From the results of the questionnaire, about half the candidates put a high value on ‘oral cavity’, ‘movement’ and ‘conversation’. This seems to point to high educational value of the robot patient especially with respect to ‘cavity preparation’ and ‘risk management’. However, the outer appearance of the robot patient needs to be improved with respect to the skin, and an upgrade in voice recognition is also required for communication education.

Almost all candidates confirmed that the robot patient is more effective than a traditional mannequin because of its movement and conversation. On the other hand, some candidates responded that the movement of the robot patient was not useful, because it prevented them from concentrating on the cavity preparation. The comments from these students suggest that, at this juncture in their clinical training, students may not have appreciated the importance of such movements because cavity preparation is performed with the real possibility of reflex and voluntary movements in the oral cavity. This led to the negative opinions about the importance of the robot patient in the questionnaire. It may be advisable for the assessment of the treatment of the whole person including risk management using the robot patient to be implemented after sufficient basic skill training using a static simulator such as a mannequin. Most candidates would also like to be able to practise invasive treatments and uncommon situations such as a medical emergency using the robot patient, because these cannot be performed with human subjects. For this purpose, we need to improve the robot patient to provide a pulse and to enable bleeding, breathing, raising its hand and so on.

This approach to training in clinical management, not only in terms of technical skills but also in terms of risk management, hopes to achieve greater appreciation by students for proper patient-centred treatment. Based on these observations, the robot patient can be considered an effective alternative approach to clinical simulation training in the area of cavity preparation.


We have developed a robot patient for dental clinical education. Application of the robot patient in an OSCE was carried out in practice, and the effectiveness thereof was shown in dental education.


The authors are grateful to Professor Atsuo Takanishi of Waseda University, Associate Professor Hideaki Takanobu of Kogakuin University and Dr Mutsumi Madokoro of Showa University for their generous assistance. We also thank Mr Yoichi Takamoto, Mr Kenichi Miyamoto and Mr Yusuke Ishii of Tmusk Co. Ltd. for the development of the robot patient. This study was partially supported by a grant from the Ministry of Education, Culture, Science, Sports and Technology, Japan.