Incipient slip detection and grasping automation for robotic surgery

Robotic minimally invasive surgery provides multiple improvements over traditional laparoscopic procedures, but one significant issue still encountered is their limited force control during the grasping and retraction of tissue, as the surgeon is separated from the instrument, and therefore denuded of their sense of touch and the applied forces. Prior solutions have largely looked towards haptic feedback to resolve this issue, but an alternative approach is to detect and monitor the occurrence of tissue slip events. This would allow the force to be automatically adjusted to prevent slip, minimising the clamp force used to maintain control, thus reducing the probability of tissue trauma. The aim of this work is to develop a method for the early detection and mitigation of tissue slip during robotic surgical manipulation tasks, helping to reduce tissue trauma and minimise tissue slip events. Initial investigations into literature, and evaluation of the slip mechanics when grasping soft, lubricated, deformable materials, indicated that small localised slips occur before the onset of macro slip. Two phenomena were identified in the slip mechanics investigation that could be employed to induce these slip in a measurable and repeatable manner. Firstly through using the tissue's deformable properties to create slip differentials between the front and rear of the grasper face, and secondly through using a curved surface to create a variation in the normal force, and thus frictional force, across the surface. Two instrumented grasper faces were developed, based on each of these phenomena, that were capable of monitoring the occurrence of localised tissue slip through monitoring the displacement of a series of independent movable islands that made up the grasper face. These were then demonstrated to be capable of automatically detecting slip events for a range of test conditions with tissue simulants, before being utilised to automatically control the grasping forces during a tissue retraction task. Both sensor systems provided similar levels of tissue control to one which utilised the maximum clamp force throughout the task, whilst applying lower forces during the early stages of retraction, reducing the probability of tissue damage. In addition the normal force based method, with the curved grasper face, was demonstrated to be effective for the early detection of slip when grasping porcine liver tissue, successfully detecting incipient slip in 77% of cases. This work provides a strong basis for further development of incipient slip sensing for surgical applications. It provides novel contributions in the understanding of slip mechanics of soft tissues, as well as presenting two separate novel sensing approaches for the automatic detection and mitigation of slip events, offering an opportunity for reducing the occurrence of tissue slip events whilst minimising tissue trauma, as well as surgeon fatigue.