Journal of Medical Systems

, Volume 32, Issue 3, pp 235–242

Field Testing of a Remote Controlled Robotic Tele-echo System in an Ambulance Using Broadband Mobile Communication Technology


    • Department of Orthopedic SurgeryYokohama City University School of Medicine
  • Hiroshi Harada
    • National Institute of Information and Communications Technology
  • Kohji Masuda
    • Graduate School of Bio-Application and Systems EngineeringTokyo University of Agriculture and Technology
  • Gen-ichiro Ota
    • Global Information & Telecommunication InstituteGraduate School of Waseda University
  • Masaki Yokoi
    • Nomura Research Institute
  • Nobuyasu Teramura
    • Yokosuka Research Park R&D Promotion Committee
  • Tomoyuki Saito
    • Department of Orthopedic SurgeryYokohama City University School of Medicine
Original Paper

DOI: 10.1007/s10916-008-9128-x

Cite this article as:
Takeuchi, R., Harada, H., Masuda, K. et al. J Med Syst (2008) 32: 235. doi:10.1007/s10916-008-9128-x


We report the testing of a mobile Robotic Tele-echo system that was placed in an ambulance and successfully transmitted clear real time echo imaging of a patient's abdomen to the destination hospital from where this device was being remotely operated. Two-way communication between the paramedics in this vehicle and a doctor standing by at the hospital was undertaken. The robot was equipped with an ultrasound probe which was remotely controlled by the clinician at the hospital and ultrasound images of the patient were transmitted wirelessly. The quality of the ultrasound images that were transmitted over the public mobile telephone networks and those transmitted over the Multimedia Wireless Access Network (a private networks) were compared. The transmission rate over the public networks and the private networks was approximately 256 Kbps, 3 Mbps respectively. Our results indicate that ultrasound images of far higher definition could be obtained through the private networks.


Robotic tele-echoBroadband mobile communication technologyAmbulanceSeamlessUltrasound


Despite the rapid progress that has been made in the area of in information technology in Japan, information regarding a patient being transported by ambulance is still transmitted to the receiving hospital using a standard telephone line known as “the hot line”. Although the range of medical practices that paramedics are permitted to perform has expanded greatly in recent year, there is still a shortfall in the treatments that they can provide for serious cases. When a seriously ill or injured patient is transported to hospital, the condition of the whole body is first examined. If organ injuries are then suspected, a range of procedures such including ultrasound, X-ray, CT scan, and MRI can be performed [1, 2]. In particular, ultrasound is an important diagnostic method which is noninvasive and can easily be performed by an experienced doctor [3, 4]. However, if any of the above examinations to assess possible organ trauma could be performed in the ambulance, the condition of the patient could be determined far more quickly, and potentially critical time could be saved upon arrival at the hospital [5]. We have in our current study conducted an experiment to examine whether an ultrasound diagnosis could be made by a doctor using seamless wireless technology to remotely control a robot in an ambulance traveling on public roads.


The area in which we conducted our field trial consisted of 24 km of public road between a simulated accident site in Yokosuka City and the Yokohama City University Hospital. We established a wireless communication network on this route in order to transmit information regarding the patient's condition to the hospital whilst ambulance was in transit.

Use of a multi-router that can freely connect to different wireless communication systems

We examined the feasibility of transmitting images from an ambulance in transit based on the following consideration:
  1. 1.

    Whether it was possible to link base stations for different wireless communication systems and continue communication.

  2. 2.

    Whether it was possible to prioritize different wireless communication systems and seamlessly connect them.

The communication environments we used in our current field experiment are described below and the basic network configuration is shown in Fig. 1.
Fig. 1

Network configuration. A patient was transported for a distance of 24 km by ambulance from Yokosuka Research Park (TRP), Yokosuka City to Yokohama City University Hospital. Four wireless LAN-based broadband wireless base stations (wireless base stations compliant with the IEEE 802.11a, IEEE 802.11g and HiSWANa standards) were placed in the YRP area and six were situated on the road running alongside Seaside Park, Yokohama, which was on the route being tested. These communication systems were connected to a multi-router to enable the linking of different wireless communication systems via a local wired IP network. They were used in concert with a mobile router installed in the ambulance so that communication was not disconnected even if different wireless communication systems were used. Information transmission in areas other than the above was performed using third generation mobile telephone services

Public networks

CDMA networks

We tested the third generation mobile telephone networks, NTT DoCoMo's FOMA and KDDI/au CDMA 1X WIN in our current field trial. To ensure an adequate and reliable transmission rate, we bundled four lines of a third generation mobile telephone system (W-CDMA) [6]. Through this method, a transmission rate of approximately 64 × 4 Kbps became possible over the public networks. The public networks were used for communication in areas outside of those covered by the private networks (linking base stations for wireless communication systems and continuing communication) described below.

Private networks

Multimedia Wireless Access System Network

One of the private networks that we tested is the Multimedia Wireless Access System Network developed by NICT (National Institute of Information and Communications Technology) which handles high-speed mobile communications in the 5 GHz range [7, 8]. To provide a sufficient transmission rate in an rapidly moving vehicle, the function of the receiving circuit of this network has already been enhanced [79]. This network also conforms to wireless LAN (Local Area Network) standard IEEE 802.11a, as prescribed by the 802 Committee of the Institute of Electrical and Electronic Engineers (IEEE), which sets the standards for LAN technologies. For the purpose of this study, we placed four base stations on the road near to the accident site (the departure point for the ambulance).

HiSWANa stands for High Speed Wireless Access Network System type a (HiSWANa) Network

HiSWANa is a type of LAN standard (STD T-70) which uses the 5 GHz bandwidth established by the Association of Radio Industries and Businesses (ARIB) and is the same type of system as HIPERLAN-2, which was established by the BRAN (Broadband Radio Access Networks) Committee of the European Telecommunications Standards Institute (ETSI) [10]. Both IEEE 802.11a and HiSWANa utilize the orthogonal frequency-division multiplexing (OFDM) method for modulation, and can transmit data at rates of up to 54 Mbps in the 5 GHz bandwidth. HiSWANa is however expected to provide a more stable level of communication than IEEE 802.11a because it can offer a bandwidth guarantee for data transmission.

IEEE 802.11g

IEEE 802.11g is one of the wireless LAN standards prescribed by the 802 Committee of the IEEE. As with 802.11a, it uses the OFDM for modulation, and can transmit data at rates up to 54 Mbps in the 2.4-GHz range.

Five access points in total were placed for each of the above-mentioned HiSWANa and IEEE 802.11g networks along a public road in Seaside Park, Yokohama (Fig. 2). When the ambulance enters these private networks, higher-definition pictures and ultrasound images of the patient are transmitted to the receiving hospital in real time. The privately operated communication systems that we utilized were also connected to a multi-router for linking different wireless communication systems via a local wired IP network (NTT East B-FLETS). Furthermore, they were configured to work in cooperation with the mobile router installed in the ambulance to realize the following functions: terminal connection/disconnection, data transmission and reception confirmation, hand-over, and communication media switching. The mobile router in the ambulance was used to continuously confirm the communication link with an intersect (IS) adapter in the multi-router. Communication is thus handed over and automatically connected to a different, higher-priority communication media if the link is momentarily disconnected and another IS adapter is available. Hence, even when different wireless communication systems are used, the communication stream is seamless.
Fig. 2

Route taken by the ambulance and positioning of the communication access points. The access points used in our system were situated in five locations, as indicated on the map, and accessed the W-LAN “HiSWANa,” in the 5-GHz frequency

Robotic tele-echo

We equipped our ambulance with a small lightweight robot fitted with an ultrasound probe (Robotic Tele-echo) on its arm (Fig. 3a) [11]. A doctor at the receiving hospital was then able to control this device remotely through signals sent by the three-dimensional manipulation of a stick connected to a personal computer installed in the control tower (Fig. 3b). In this way, the treating clinician can control the robot in the ambulance in transit, and the ultrasound images obtained are then transmitted to a monitor at the hospital (Fig. 3c). As part of our evaluation in this trial, we judged whether or not these images were useful for actual diagnostic purposes.
Fig. 3

Robotic tele-echo system. a Echo probe and robotic arms. A distortion gauge embedded in the robotic arm ensures patient safety by preventing the robot from applying excessive force to the subject. Also, the depth and brightness of the ultrasound device can be remotely adjusted. b Controller. The control tower is shown, and consists of a controller connected to several wires. When the doctor moves this controller in three dimensions, the change in the wire lengths change are directly transmitted to the robot in the ambulance. c Installation of equipment in the ambulance. The ultrasound probe is attached to the end of a pantograph-shaped robot arm. While viewing the image of the patient taken by the video camera, the operator can remotely move the arm in any direction while the probe remains in contact with the patient's body

The robot that was used was developed in the Masuda's laboratory at the Tokyo University of Agriculture and Technology [3, 12]. The ultrasound probe mounted on the robotic arm can respond to the rotary motion produced by the joint movements of the operator's hand and parallel the movements of the operator's arm. Additionally, the robotic arm is further controlled by strain gauges to prevent excessive force from being applied to the subject. The treating doctor can thus remotely operate the controller and apply the probe to the affected part of the patient, whilst at the same time view images of the patient taken with a web camera installed in the ambulance. In addition, the ultrasound device (the prototyping for which is underway at Panasonic Healthcare) is small and lightweight, and its depth and brightness can also be adjusted remotely.


Transmission rate

We found that the transition between the public networks and the Multimedia Wireless Access System Network (private networks) occurred smoothly in almost all cases. As for the prioritization of the connection between these different wireless communication systems, the wireless system with the highest receiving power was automatically selected by the computer and communication with the control tower could thus continue seamlessly. Since we adopted a method that involves the bundling of four lines of a third generation mobile telephone system (W-CDMA) in this experiment, a transmission rate of approximately 64 × 4 kbps was secured even over public networks, thus allowing the transmission of relatively high quality ultrasound and camera images of the patients. However, on some occasions disconnections did occur in the area covered by the public system.

On the other hand, no disconnections occurred when using the private networks. Furthermore, measurement of the transmission characteristics was possible when accessing these networks. We determined that it was difficult to use the Robotic Tele-echo over public networks due to problems with the speed and robustness of the transmission rate. However, with the private networks, a transmission rate of approximately 3 Mbps was achieved (Fig. 4), and the images on the receiving side provided sufficient amount of information to perform a diagnosis.
Fig. 4

Transmission rate. The transmission rates for access points (AP) 1 and 2 are indicated and 3 Mbps at least was obtained in both cases. a Transmission rate of AP 1. Each bar indicates the transmission rate for one half second period. As the vertical axis is defined as 1 bps, the height of the bar is twice of the measured value. The graph shows the transmission rates of AP 1, the position of which is shown in Fig. 2. The flat part of the chart at around 20 s was due to a red signal at an intersection in the road. The reason for the low transmission rate in this instance was caused by the fading effects of the road surface reflection. b Transmission rate of AP 2. Transmission rates for AP 2, the position of which is also shown in Fig. 2. In this area there are no intersection in the road, and the envelope of the data rates is thus almost symmetric at the center

Robotic tele-echo

As described earlier, the transmission rate over the public networks was approximately 64 × 4 Kbps as four lines were bundled together. In addition, the ultrasound images were transmitted over the public network at a frame rate of about 5 frames per second (fps), and lacked definition. This rate was about only 2 fps when a single standard line was used. These images were also characteristic of still images (Fig. 5a). In contrast, the images transmitted via the private networks were far clearer and had superior definition. The image size on the receiving monitor in these cases was 640 × 480 pixels for the ultrasound images, and 800 × 600 pixels for camera images (Fig. 5b).
Fig. 5

Transmitted abdominal ultrasound images. a Image transmitted over the public network. The transmission rate over the public networks was approximately 64 × 4 Kbps. The ultrasound images were transmitted at a frame rate of about 5 fps, and lacked definition. A frame rate of about 2 fps was obtained when the usual single line was used. b Image transmitted over the private network. When accessing the private networks, we could transmit ultrasound images at a bit rate of 5 Mbps and 30 fps. This is equivalent to the ordinary television transmission rate and the images obtained were thus far clearer than those transmitted over the public networks

The transmission rate achieved when accessing the private networks was 30 fps, with an average bit rate of 3 Mbps for both camera images, and ultrasound images. The pictures of the patient taken using the Web camera were also of a high definition and good quality color. In addition, although there was a slight delay noted between the dispatch of commands from the control tower in the receiving hospital and the movement of the ultrasound probe, the device moved in accordance with operator instructions in almost all cases. Moreover, ultrasound images of the patient's chest and abdomen was transmitted clearly in real time. We thus conclude that a rate of 30 fps was more than adequate to achieve the desired results and that the quality of the ultrasound images was sufficient for useful screening in the ambulance. It would thus be of great benefit to emergency medical practice if this method could be adopted in the near future and further improvements to the equipment could be made.


Communication environment

An assessment of three principal systems for the transmission of patient images and other medical information from an ambulance in transit to the receiving hospital has been undertaken by our laboratory. These are: (1) Image transmission via base stations for Multimedia Wireless Access System Networks. Our group has conducted two public field trials of this method. (2) A method using third generation mobile telephone services (NTT DoCoMo's FOMA and KDDI/au CDMA 1X1 WIN) that allow the transmission of moving images and data. (3) A method that uses communications satellites.

All these methods have both advantages and disadvantages. In method 1, wireless base stations are placed on the road at intervals of 500 m, and are connected to each other via an existing optical fiber network. Therefore, the required number of base stations can be placed where necessary and in accordance with budgetary constraints. As demonstrated in our present experiment, the transmission of several dozen Mbps of information between the mobile station installed in the ambulance and base stations is now possible even when the ambulance is traveling at high speeds. An average frame rate of 30 fps was possible when sending patient images via this system and this is at the same level as ordinary television images, and therefore provides sufficient accuracy for moving images. Moreover, if this system allows the connection of mobile stations and various kinds of medical measuring equipment via a generic Ethernet cable and if it also permits IP communication, then establishing a connection will be relatively easy. In addition, if the base stations have to be located in an area where the B-FLETS service is not available, an alternative communication route can be established by connecting to this service for an adjacent base station via a WIPASS (Wireless Internet Protocol Access System).

As we have shown in our current experiments, this system, in combination with image transmission via mobile telephones, will likely lead to further improvements in the emergency life-saving rate. Each method should be used as appropriate and seamless communication technology should also be used to realize continuous data transmission.

In method 2, the data transmission capacity is low and the communication speed is only about 64 Kbps. Using this method also, patient images are transmitted at a frame rate of only about 2 fps, which is sufficient for still images only. However, one advantage of such a system is that there is almost no need to invest in new equipment. This method may also provide a sufficient degree of information to handle less serious trauma cases, and its utility would obviously then depend on the severity of each individual case.

Method 3 has the noteworthy disadvantage of requiring significant expenditure on equipment. However, this method would very useful for less populated regions where, the equipment investment would need to be very large if method 1 was adopted. In the case of method 3 also, communication in tunnels is impossible in many cases, which can be overcome using method 1 by positioning access point at the entrances of any tunnels. The mobile telephone service area can also be extended using leaky coaxial cables in method 2.

Although the selection of the method used is ultimately dependent on the condition of the patient being transported and the judgment of the doctor at the receiving hospital, we believe that in many cases there would be no problem with using still images to make medial determinations. The requirements for image definition, color reproducibility and other facets of image quality will need to be evaluated for these methods going forward but a wide range of factors will also need to be assessed including the response from the site to the hospital, the frame rate, monitor size, information continuity and the interlinking of these remotely accessed data with electronic medical records.

Robotic tele-echo

Recently, “the necessity of image transmission and image diagnosis from emergency medical sites, especially from ambulances” has gradually become a topic of discussion, and the importance of the availability of medical images, particularly image reproduction for diagnosis, is now being increasingly recognized [1317]. The JATEC (Japan Advanced Trauma Evaluation and Care) Guidelines for the Initial Treatment of Trauma published in 2002 describe how an emergency ultrasonic examination, known as FAST (Focused Assessment with Sonography for Trauma), plays an important role in the primary care of trauma [18]. The purpose of FAST is to search for pericardial effusion or intra-thoracic and intra-abdominal hemorrhaging in an emergency. This kind of diagnosis performed at an emergency site would most likely reduce the time between hospital arrival and actual treatment. Also, it would become possible for doctors at the hospital to give appropriate instructions to paramedics in the ambulance in real time.

In this regard, ultrasound examination (echo) also has a number of great advantages including the compactness of the equipment itself and thus its portability as demonstrated in our current trial. It is also a non-invasive procedure, allowing repeated examinations in a limited time, and it has excellent reproducibility. Therefore, the use of ultrasound is expected to become more prominent at emergency sites. It is noteworthy in this regard that a view has been expressed that the inclusion of ultrasound examinations in ambulances would only place an additional burden on paramedics who are trying to perform various first-aid treatments in a restricted space and under severe time constraints. We have thus considered the issue of a realistic and feasible operation of an ultrasound device at emergency sites, and developed the idea of remote control of this instrument by a doctor. Using this method the doctor would essentially perform the examination, and the burden on busy paramedics would be reduced.


The results of our present experimental trial show that the operability of the Robotic Tele-echo in a traveling vehicle from a remote location was relatively good, and the transmitted ultrasound images were clear. Therefore, we fully anticipate that this method will be put into practical use sometime in the future.


We would like to extend our deep thanks to the many people who have made this field experiment possible through their extraordinary efforts.

Conflict of interest

This study was supported by a 2004 Yokohama City Research Encouragement Grant (funds for significant research). None of the other authors has any conflict interest with this research project.

Copyright information

© Springer Science+Business Media, LLC 2008