Effects of communication delay in the dual cockpit remote robotic surgery system

Robotic surgery has spread to facilities in various parts of Japan since the 2018 revision of medical reimbursement regulations began allowing insurance coverage of robotic surgery in various fields, including gastrointestinal and respiratory surgery. In addition, new surgical robots are entering the market, and new technologies are being developed [1]. One of the advantages of robotic surgery is that it enables us to make various innovations based on digitized information. Surgical education methods using swapping and annotation functions with a dual console go beyond what was possible with conventional approaches and are quickly becoming standard protocols among the more useful methods of surgical education. In recent years, research has been conducted to implement these more evolved system tools in society through telesurgery technology, using robots and telecommunication modalities.

We have been conducting experimental studies on telesurgery since February 2021 and have verified the effects of communication delays and image degradation on surgical operations [2‐6]. We have demonstrated the feasibility of remote surgery using Japanese commercial lines and domestic robots and developed guidelines for clinical applications. However, we have not studied the construction of a system in which a physician at an institutional hospital in a remote location can provide surgical support through a dual cockpit to a physician performing surgery locally.

One of the main problems with telesurgery systems is communication delays. To implement remote surgical support in society, it is essential to develop tools that provide guidance to local surgeons from remote locations. Therefore, it is necessary to verify the acceptable delay in the surgical support system.

The present study evaluated the acceptable range of communication delays for teaching using a dual cockpit in telesurgery.

Similar to our previous studies [10], we concluded that it is difficult for a proctor to operate in a remote environment with a delay of ≥ 150 ms. Therefore, when an unexpected situation occurs during telesurgery resulting in a delay of ≥ 150 ms, surgical support from the proctor is considered difficult. However, based on the hypothesis that the delay time limit perceived by the recipients of guidance is similar, we conducted this experiment and concluded that it is possible to receive guidance even at 200 ms, which is more than expected. This was thought to be because a certain amount of delay may be allowed for education using the annotation and swapping functions, as long as the surgery is performed smoothly.

The dual cockpit concept was introduced in the Intuitive Surgical da Vinci Si in 2009 with dual console functionality, allowing two surgeons to operate the surgical robot simultaneously [11]. The dual cockpit system is considered to be educationally useful, contributing to the improvement of surgical skills and the learning curve; this is not limited to the field of gastrointestinal surgery [12]. In addition, the dual console allows the primary surgeon and supporting physician to perform the same surgery simultaneously, making it possible to conduct surgery safely and with a high degree of reproducibility [13]. However, few reports have focused on dual cockpits in remote surgical environments. Oki et al. reported the usefulness of a dual cockpit in a remote area 140 km away, where a non-specialist surgeon performed emergency hemostasis, cholecystectomy, and renal vein ligation in surgery using pigs with remote assistance from a specialist surgeon [6].

We believe that the maximum delay time that occurs in telesurgery is limited to < 100 ms, and in our own example, the delay time was 29 ms at a distance of 150 km [3]. There are several reports of acceptable delay times, with reports of increased task times and error rates above 300 ms [14]. Similarly, surgical performance was notable worse at ≥ 300 ms than with less of a delay and showed a significantly increased error rate at ≥ 500 ms, suggesting that such delays lead to surgical risk [15]. In our report, we examined the acceptable durations of delay and found that a delay time of ≥ 100 ms affects surgical outcomes [10]. As in previous reports, we discovered that both annotation and swapping functions (≥ 150 ms) were insufficient for robotic surgical guidance from a remote location. However, it is notable that the surgeons on the receiving end of the instruction did not notice the effects of the delay as much as was felt on the instructor side. Therefore, when focusing on providing instructions from a remote location, there is a high possibility that even a delayed environment is sufficient for providing annotation and voice instruction, even if such a delay would be unacceptable when it came to operating and moving the robot.

No notable effect of delay was observed with respect to the switching time. When surgical support from a remote location is provided, it is thought that the quality of instruction is increased, provided the switching time is kept as short as possible. In addition, in an emergency situation, such as bleeding, it is desirable for the instructor to be able to instantly shift control to the proctor in order to stop any bleeding.

The significance of this experiment lies in the fact that the robot operability and environment were evaluated by not only the proctor but also the operator. The intervention of the remote instructor did not adversely affect the surgical operation of the primary surgeon, and both parties expressed their expectations concerning the usefulness of the dual cockpit remote instruction in actual clinical practice.

In this verification study, the experiments were conducted by constructing a simulated remote environment indoors, but communication modalities need to be evaluated using the commercial lines that will actually be used. As telesurgery has become more widespread, variations in the modalities used are expected to increase. There are reports that dual-console surgery using 5G and wired networks is possible [16], and it will be necessary to study and evaluate various such modalities in the future.

In the United States, the use of robots in general surgical procedures, especially for managing hernias, has become prominent, more so than in cancer surgeries, and robotic approaches are rapidly replacing most conventional approaches. In Japan, although surgical-assisting robots are becoming more widespread throughout the country, the number of surgical-assisting robots and the number of surgical slots are still insufficient to cover the total number of surgeries. Consequently, evidence of treatment outcomes is available only for a limited number of advanced centers. While robotic surgery in Japan is expected to continue to focus on cancer treatment, it is also expected to eventually follow the lead of the United States and expand its application to various general surgical procedures. With this in mind, the issue for the future is to determine how to provide surgical education to young surgeons who are just starting to use robotic surgery. There is no doubt that robotic surgery is a technology that young surgeons aim for higher quality, and more precise surgical modalities should be explored. Young surgeons working in rural areas in particular will benefit greatly from remote surgical training from a medical advisor at a core hospital in a more central location.

In this verification study, telesurgery with a dual console using hinotori^™ was practical, and the acceptable delay between the proctor and operator was clarified. When considering the characteristics of each hospital and region, it is safe to assume that environmental conditions will vary widely, such as what kind of communication modality will be used to connect the core hospital and the remote hospital and which surgical robot will be used. Therefore, further studies under diverse conditions will be necessary in the future.

Remote robotic surgical guidance using a dual cockpit with a hinotori^™ device is practical. The surgeon receiving the instruction was able to follow the instruction even with a delay of ≥ 150 ms; however, the greatest acceptable communication delay for the surgeon giving the instruction was < 150 ms.