Next-generation Teleoperated Semi-Autonomous Interventional Surgical Robots: soft robots, soft sensors, nonlinear modeling, and learning-based control

目前的手术机器人，即第一代，具有铰接臂，连接到长而薄的刚性仪器。这些器械通过皮肤上的小切口插入患者体内。刚性器械的一个问题是，它们都需要在皮肤上有一个专门的切口，它们的运动是相互独立的。而对于精细的手术，例如心脏手术，则需要手臂之间的舞蹈动作。此外，对于机器人手术来说，需要3- 4个关节臂，这给医疗保健系统带来了一大笔资本投资（3-5百万美元），用于运行一个手术机器人。目前，外科医生坐在控制台（类似于驾驶舱）并操纵操纵杆，而机械臂则精确地复制外科医生在患者体内的动作。现在，机器人仅仅遵循来自外科医生的命令，因此，例如，通信中断发生，机器人处于空闲状态，直到下一个命令通过通信通道到达。因此，目前的技术可能无法满足临床对远程手术的需求。这项研究是开发下一代软体半自主手术机器人的一步。更具体地说，这项建议有四个主要目标：1)开发软性触觉传感器，帮助了解患者体内机器人仪器的形状、姿势、力度和相互作用；2)开发用于支气管镜检查和心脏血管内手术的软性手术机器人，这些机器人体积小，2-3个可以装在一个小直径的导管中，在患者的气管或心脏内进行手术；3)开发工具组织模型，帮助理解软手术机器人与患者身体之间的交互状态；4)开发基于人工智能（AI-）的控制方法，为软手术机器人增加一层自我意识和半自主性，以帮助外科医生快速准确地完成手术任务，并消除通信中断时空投的风险。这项研究的成果将有助于加拿大手术机器人知识的进步，使加拿大的研究人员处于受益于最先进的下一代手术机器人的最前沿，加拿大境内的知识产权概念可以在国内进一步商业化或跨越美国/欧盟合作伙伴的边界。此外，本文提出的半自主手术机器人将使早期职业外科医生能够在机器人控制器的嵌入式辅助下进行远程手术。这最终将有助于增加加拿大农村和郊区社区获得最先进治疗的机会。这也是向受益于人工智能在医学领域的力量迈出的一步，加拿大将保持其在世界上的全球领导地位。

The current generation of surgical robots, i.e., first-generation, have articulated arms that are connected to long thin rigid instruments. These instruments are inserted to the patient's body through small incisions on the skin. On of the problems with the rigid instruments is that they each require a dedicated incision on the skin and their motion is independent from each other. While for a delicate surgery, e.g., on heart, choreographic motion among the arms is required. Moreover, for a robotic surgery 3 - 4 articulated arms are required which imposes a large lump of capital investment ($3-5 M) to the healthcare system for having one surgical robot running. Currently, surgeon sits at a control console (similar to a cockpit) and manipulates joysticks while the robotic arms precisely replicate the surgeon's motions inside the patient's body. Now, the robots merely follow the commands coming from the surgeons thus should, for example, a communication disruption occurs, the robot stays idle until the next commands arrive through the communication channel. Therefore, the current technology may fall short of clinical requirement for telesurgery. This research is a step toward developing next-generation surgical robots that are soft and semi-autonomous. More specifically, this proposal has four main objectives: 1) developing soft tactile sensors that help in understanding of the shape, pose, force, and interaction of the robotic instruments inside the patient's body, 2) developing soft surgical robots for bronchoscopy and cardiac endovascular surgery that are small and 2-3 of those can be packed into a small-diameter catheter to perform surgery inside the patient's trachea or heart, 3) developing tool-tissue models that assist in understanding the state of interaction between the soft surgical robots and patient's body, and 4) to develop artificial intelligence (AI-) based control methods to add a layer of self-awareness and semi-autonomy to the soft robots to assist the surgeons in rapid and precise performance of his surgical tasks and to eliminate the risk of idling while the communication is disrupted. The outputs of this research will contribute to the advancement of surgical robotics knowledge in Canada putting Canadian researchers at the forefront of benefiting from state-of-the-art next-generation surgical robots, conception of intellectual property within the Canadian territory that can be further commercialized domestically or across the border with the US/EU partners. Moreover, the semi-autonomous surgical robots proposed here will enable the early-career surgeons to adopt tele-operated surgeries with the embedded assistance from the robot controller. This will eventually contribute to increasing the access to the state-of-the-art treatments in the rural Canada and suburban communities. It is also a step forward toward benefiting from the power of AI in medicine for which Canada shall maintain its global leadership in the world.