Robot-assisted minimally invasive surgery has become a preferred approach to augment surgeons’ skills and improve surgery quality. For example, Using a laparoscope-holding robot has been shown to improve surgeons’ efficiency and operation safety since it gives the surgeon direct control over the laparoscope and eliminates the suboptimal surgeon-assistant communication process. However, the application of new technologies challenges the surgeon both cognitively and physically. Controlling the robotic laparoscope holder can increase the surgeon’s workload and distract them from the primary hands-on operations. Our objective is to help the surgeon focus on the main operation and reduce the laparoscope control workload while maintaining their control authority over it.

This project aims to develop an intelligent laparoscope control system composed of a novel foot interface and a shared autonomy framework. The novel foot interface has been shown to provide natural and intuitive control and reduce the user’s physical and mental fatigue. The shared control framework gives the surgeon direct control over the laparoscope but automatically adjusts the zoom factor. The case study results show that the proposed intelligent control system is able to lower the surgeons’ workload and helps them focus on their hands-on tasks for better performance.

This project develops a novel foot interface that can manipulate 4 Degrees of Freedom of the laparoscope’s pose at two speeds using decoupled control. The novel foot interface is compared with a voice interface using three 4 Degree of Freedom laparoscope control tasks. The results show that the foot interface’s completion time, error rate, and mental burden are significantly lower than those of the voice system, which is typically considered a more natural and intuitive solution. The foot interface benefits from an ergonomic design, natural mapping and clear haptic feedback. Based on the findings, we infer that this better performance will be maintained in surgery-like tasks.

Then the two most common control approaches for foot interfaces:​ decoupled and hybrid control are compared. Although it is a fundamental and critical topic to understand the best way to coordinate and send commands for a laparoscope control task, it has not been studied before. The experiment uses a pure laparoscope control task and recruited eleven surgeons (Six experienced surgeons and five surgeons-in-train). The results show that decoupled control is more suitable because of its better predictability of the final laparoscope pointing position and similar efficiency as hybrid control. In addition, the foot interface usage pattern is analysed, and a guideline for foot interface control approach selection is provided.

In addition, a novel shared autonomy framework is developed. It consists of a perception and a control section. Unlike conventional methods that use sensors or markers to estimate the distance between the laparoscope and the tool, we define a new index based on the tool’s geometric feature and use it for zoom factor control. The performance of the perception block is tested on the Cholec80 dataset and shows a satisfactory result on the tool segmentation and index calculation. Moreover, the common optimal range of the lens zoom factor and the tooltip’s position heatmap are identified and presented quantitatively.

Finally, the proposed intelligent laparoscope control system is evaluated using two surgery-like tasks in a case study with an experienced surgeon and a trainee surgeon. The results show that using shared control can reduce the surgeon’s workload during the operation, improving their hands-on tasks performance and efficiency.

This thesis provides initial insights into human-robot interaction for the laparoscope control task in robot-assisted minimally invasive surgery. A novel foot interface and shared autonomy framework are completed during the project. The evaluation results show the proposed control system effectively meets the design aims. Based on this project, future development on shared autonomy or interface design can further improve the surgeon’s performance and contribute to the surgical robot research field. Moreover, the project is informative for other robot-assisted surgical tasks, such as controlling a surgical tool using a robotic arm.