Haptic Feedback for Soft-Tissue Robotic Surgery: from Training Palpation to Haptic Augmentation

Robotic minimally invasive surgery (RMIS) is a field in full development. Since evidence of the benefits of robotic approaches are surfacing the field, RMIS is receiving an increased interest by both academia and industry. There are nevertheless some major hurdles to this technology. Apart from the cost, a major impediment to its adoption is the long learning curve. As in RMIS the surgeon is decoupled from the patient and only interacts through the robotic system, she/he is to less extent aware of what is going on at the patient side. Critical situations can therefore go unnoticed. In a response robotic systems are increasingly being equipped with all sorts of sensors to measure various aspects of the patient’s internals. The challenge remains how to communicate this information back to the surgeon so that this information is easy to interpret. A seemingly convenient approach consists in decoding the acquired information and presenting it via a visual overlay to the surgeon. However, the visual channel is already heavily in use and there is thus a risk of an increased mental load. Instead, the haptic sense, which is the sense related to touch, remains unused. Indeed while in traditional surgery surgeons rely heavily on haptics, due to the decoupling, in RMIS this feedback channel is not available anymore. Nevertheless the haptic sensorimotor system is fast, natural and elicits reactions that are much faster than the visual ones.

This thesis, develops and introduces novel haptic technology to aid the surgeon in surgical training and enhance her/his awareness in RMIS.

A novel encountered-type haptic display is proposed and developed for palpation training. Palpation is a crucial skill that is important both during diagnosis and surgical interventions. Surgeons rely on it to detect anomalies in organs and delineate tissue margins. Conventional systems fail to replicate natural palpation as they force the user to interact via a stylus or pen-based interface. On the contrary the proposed encountered-type display allows unrestricted freehand palpation. Two prototypes were developed to render three surface haptic features: normal stiffness, 1st order geometry and lateral forces. The realism of the interface is validated with experimental campaigns.

Haptic feedback, which is missing in current robotic solutions for RMIS, has been implemented in a new bimanual surgical teleoperation platform. The validity of haptic feedback has been investigated on a challenging gynaecological intervention, namely endometriosis surgery. ESE, a dedicated training model that replicates endometriosis surgery, has been developed for this purpose. The benefit of force feedback for RMIS during complex surgical interventions has been demonstrated through user experiments on the ESE model. User experiments showed improved awareness reducing interaction forces and lower amounts of errors with force fed back. As far as the author is aware of these were the most sophisticated experiments investigating the role of haptic feedback in RMIS up to date.

Finally, haptic augmentation has been investigated. Beyond feedback of interaction forces the haptic channel can be used to encode various sensory information, force feedback and control can also be used to simplify and automate parts of surgical procedures. Two applications, a first introducing haptic fixture and a second introducing force fixtures, both showing the potential of haptic augmentation, have been developed. First, in cardiovascular surgery, haptic guidance is used to help the interventionalist guide a robotic catheter along the centre-line of the aortic arch. In a haptic shared-control approach, the control is shared with the surgeon. The latter remains responsible for decision-making but does not have to worry too much about intricate catheter steering. Second, automatic tissue tensioning is explored. This may help during gynaecological surgery where surgeons need to remove diseased tissue via laser ablation. The teleoperated robot takes control over the tissue tension task presenting it in an optimal condition for the surgeon who can then ablate it. The surgeon teaches the system to apply a required force level. This becomes possible through the force feedback system which allow the surgeon to feel and thus set a desirable force level. User experiments show improved manipulation when using this new feature.

The haptic sensorimotor system is a sophisticated distributed perceptual system, the result of a long evolutionary process. While the understanding of this perceptual system is improving and already very mature, methods to interface and interact with it are still under development. RMIS is one particularly interesting field that could benefit from such interactive technology. This thesis presents contributions in hardware, software and control, as well as experimental validation, bringing haptic interactive technology for RMIS to a next level.