Concurrent nonparametric estimation of organ geometry and tissue stiffness using continuous adaptive palpation

Abstract

Surgeons often manually palpate tissue or organs in order to find tumors or other anatomical structures. Information about organ geometry and tissue stiffness gained from palpation can also be extremely useful in robotic surgery for diagnosis, surgical guidance, and registration to other preoperative information. However, it is not always easy to obtain, even if the robot is equipped with force sensors. This paper reports our approach for concurrent estimation of stiffness and surface geometry, using a continuous motion similar to a sweeping palpation motion used by surgeons. Our method relies on force data captured by a tactile sensor rigidly attached to an end-effector probe. We use Gaussian processes to simultaneously estimate geometry and stiffness. The method is not tied to any specific robotic platform and is consistent with a variety of palpation strategies. For simplicity, we discuss the results based on two different palpation primitives. This is our first step towards developing an adaptive high fidelity model reconstruction and path optimization technique.

Autonomy in Surgical Robotics

“…In these procedures, it is also crucial to spare as much healthy tissue as possible to prevent organ failure and subsequent need of a transplant, should this be an available option. A possible approach is to perform mechanical palpation to create a local stiffness map for guiding ablation [97,98], as discussed in Section 4.4. Alternative imaging methods such as ultrasound [99] and optical coherence tomography [100] may be adopted to guide cardiac ablation.…”

Feasibility of Remote Landmark Identification for Cricothyrotomy Using Robotic Palpation

“…Related works to this investigation include works on mechanical imaging of tumors [8][9][10], mechanical probing [11], mechanical excitation for tissue impedance characterization [12], and robot-assisted palpation [13]. These works predominantly focused on the identification of tumors and have not been applied to palpation of the airways.…”

Medical Technologies and Challenges of Robot-Assisted Minimally Invasive Intervention and Diagnostics

“…Finally, many of the surgical paradigms still fail due to lack of perception and invivo sensory information allowing the surgeons to correlate the surgical scene with pre-operative plans. To overcome the fact that organs shift and swell during surgery, recent works such as [145,146,147,148] have started to explore the use of in-vivo model update based on adapting a pre-operative to a model created using sensory data including force, contact location and stiffness. These approaches complement prior works on using geometric scanning and registration of organs (e.g.…”