In this work, we develop an approach for guiding robots to automatically localize and find the shapes of tumors and other stiff inclusions present in the anatomy. Our approach uses Gaussian processes to model the stiffness distribution and active learning to direct the palpation path of the robot. The palpation paths are chosen such that they maximize an acquisition function provided by an active learning algorithm. Our approach provides the flexibility to avoid obstacles in the robot's path, incorporate uncertainties in robot position and sensor measurements, include prior information about location of stiff inclusions while respecting the robot-kinematics. To the best of our knowledge this is the first work in literature that considers all the above conditions while localizing tumors. The proposed framework is evaluated via simulation and experimentation on three different robot platforms: 6-DoF industrial arm, da Vinci Research Kit (dVRK), and the Insertable Robotic Effector Platform (IREP). Results show that our approach can accurately estimate the locations and boundaries of the stiff inclusions while reducing exploration time.

In search and surveillance applications in robotics, it is intuitive to spatially distribute robot trajectories with respect to the probability of locating targets in the domain. Ergodic coverage is one such approach to trajectory planning in which a robot is directed such that the percentage of time spent in a region is in proportion to the probability of locating targets in that region. In this work, we extend the ergodic coverage algorithm to robots operating in constrained environments and present a formulation that can capture sensor footprint and avoid obstacles and restricted areas in the domain. We demonstrate that our formulation easily extends to coordination of multiple robots equipped with different sensing capabilities to perform ergodic coverage of a domain.

One of the goals of computer-aided surgery is to register intraoperative data to preoperative model of the anatomy, and hence add complementary information that can facilitate the task of surgical navigation. In this context, mechanical palpation can reveal critical anatomical features such as arteries and cancerous lumps which are stiffer than the surrounding tissue. This work uses position and force measurements obtained during mechanical palpation for registration and stiffness mapping. Prior approaches, including our own, exhaustively palpated the entire organ to achieve this goal. To overcome the costly palpation of the entire organ, a Bayesian optimization framework is introduced to guide the end effector to palpate stiff regions while simultaneously updating the registration of the end effector to an a priori geometric model of the organ, hence enabling the fusion of intraoperative data into the a priori model obtained through imaging. This new framework uses Gaussian processes to model the stiffness distribution and Bayesian optimization to direct where to sample next for maximum information gain. The proposed method was evaluated with experimental data obtained using a Cartesian robot interacting with a silicone organ model and an ex vivo porcine liver.

Registering a surgical tool to an a priori model of the environment is an important first step in computer-aided surgery. In this paper we present an approach for simultaneous registration and stiffness mapping using blind exploration of flexible environments. During contact-based exploration of flexible environments, the physical interaction with the environment can induce local deformation, leading to erroneous registration if not accounted for. To overcome this issue, a new registration method called complementary model update (CMU), is introduced. By incorporating measurements of the contact force, and contact location, we minimize a unique objective function to cancel out the effect of local deformation. We are thus able to acquire the necessary registration parameters using both geometry and stiffness information. The proposed CMU method is evaluated in simulation and using experimental data obtained by probing silicone models and an ex vivo organ.

Discretely actuated steerable cannula is a multi-degree-of-freedom hollow needle (cannula) that could potentially be used in needle-based procedures to deliver therapeutic and diagnostic tools to a target region through its hollow inner core. Needle-based procedures are commonly performed using intra-operative image guidance. 2D ultrasound is one of the most commonly used imaging modalities in clinics. It is portable, inexpensive and free of ionizing radiation. The success of the needle-based procedures depends on accurate detection of the needle. The accuracy of the out-of-plane detection, where the ultrasound transducer is placed at a right angle to the long-axis of the needle, depends on the ultrasound beam width. Finite width of the ultrasound beam and uniform cross-section of the needle introduce errors in tracking. The accuracy of in-plane tracking depends on the tracking algorithm used and the spatial resolution of the ultrasound transducer. This work presents a method to quantify the finite ultrasound beam width and the spatial accuracy of the transducer. An out-of-plane detection method was implemented to locate the tip of the discretely actuated steerable cannula. An in-plane tracking algorithm based on optical flow was also developed to obtain the shape of a planar cannula. The algorithms and the methods developed in this work are general and they can be extended to needles having straight, curved and arbitrary shapes.

Several percutaneous needle-based and intravascular procedures require guidance of the diagnostic or therapeutic tool to the target location by maneuvering the needle or catheter to correct for the error in reaching the target location. Hence, in this paper we present our work towards developing a discretely actuated steerable needle/cannula with multiple degrees-of-freedom to `steer' the cannula by discrete actuation along the cannula length. We are interested in using the cannula to introduce both diagnostic and therapeutic tools, which may otherwise be difficult to deliver to the appropriate location. We use two antagonistic SMA wires as actuators to generate the required bending forces at each joint. SMA wires were annealed through a customized training process to an arc shape and mounted in the machined grooves on the outer surface of the cannula to generate local bending upon thermal actuation. We propose to use temperature feedback to control the position of the SMA actuators. To use temperature feedback as the feedback signal for enabling individual joint actuation, we had to fully characterize the SMA actuator. Classical uniaxial testing devices and experimental setup used in characterizing straight annealed SMA wires are not applicable in this work since the SMA wire is annealed in an arbitrary shape. Hence, we also present an experimental setup and a procedure for characterizing an SMA actuator that transforms into an arc shape upon thermal actuation.