Human-Robot Interaction and Control Strategies in Minimally-Invasive Surgery

Minimally invasive surgery (MIS) aims to reduce the invasiveness of the current surgical procedure. Decreasing invasiveness means reducing patient discomfort and pain, associated with costs of the procedures, related complications, hospitalization duration, and cosmetic effects. Although the advantages of MIS for the patients are undeniable, technical difficulties are introduced because of the confined operating workspace. These limitations can be overcome by exploiting the benefits introduced by soft robotics. Indeed, the flexible access surgery requires specialized instrumentation, engineered with novel materials and actuation design. New materials ensure enhanced flexibility and dexterity with respect to rigid materials, but they introduce some challenges, such as adding sensors that are compliant with the soft structure. Chapter 2 presents a promising integration of a commercial flex bend sensor into a soft actuator. Sensor measurements are used to implement a closed-loop control of the tip actuator position, achieving a 1 mm accuracy that can be considered a relevant achievement in the soft robotics field. Hereafter, a soft manipulator (composed of two actuators) equipped with a miniaturized camera was integrated with the endoscopic camera manipulator arm of the da Vinci Research Kit (Chapter 3). The experimental tests with non-expert users demonstrate increased workspace and dexterity compared to traditional endoscopes. In order to further reduce the invasiveness of the surgical procedure, focused ultrasound surgery (FUS) was investigated. FUS ablates tumor tissue through ultrasound energy delivered by a transducer positioned outside the patient. A robotic platform, composed of two anthropomorphic industrial manipulators, was developed to increase the accuracy and repeatability of the focused ultrasound therapy. In this framework, a novel method for organ motion compensation was developed (Chapter 4). This method delivers the ultrasound energy through the rotation of the transducer around a virtual pivot point, preserving the acoustic window during the therapy and controlling the force between transducer and patient. However, rigid link manipulator introduces safety problems for both personnel and patient. Indeed, undesired collisions between humans and manipulators need to be avoided or at least managed using hardware and software-based solutions able to reduce. Most of the collision avoidance strategies rely on distances between human and robots; thus, reliably computing these distances is critical. Chapter 5 proposes a method for computing these distances starting from depth measurement. This strategy is intrinsically modular, allowing an easy integration of multiple depth sensors, thus drastically increasing the robustness of the safety system. However, unexpected collisions can still occur. For this reason, a multi-touch sensorized skin was developed to handle these collisions (Chapter 6). The sensorized skin is able to lower the pressure exerted by the robots under the painless threshold. Additionally, the multi-touch feature of the skin enables the development of physically interaction strategies with robotic manipulators. These strategies permit humans (physicians and operators in case of medical procedures) to arrange the manipulator configuration according to the surgical task requirements simply touching the manipulator surfaces; this feature enables innate and intuitive interaction with conventional manipulators in unstructured environments.