Ultrasound-guided robotic assistance for prostate brachytherapy

• Author / Creator

  Carriere, Jay T

• Percutaneous procedures, those involving the insertion of needles into tissue, encompass a wide array of clinical applications and are used in both therapeutic and diagnostic modalities. In either of these modalities, the needle must be steered accurately towards a target location to maximize the therapeutic or diagnostic efficacy. Specific procedures, including biopsy, drug delivery, or radiotherapy delivery, require the use of long flexible needles to reach a target location deep inside the body. During insertion, these long flexible needles will bend, or deflect, away from the desired target location or target insertion axes. Surgical robotic systems can be used to reduce or eliminate this deflection during insertion, thus increasing needle placement accuracy. Of particular clinical and research interest are systems which assist, rather than replace, the clinician to capitalize on the clinician’s intuition, training, and expertise.These assistive surgical systems, commonly referred to as surgeon-in-the-loop systems, can be applied to percutaneous procedures to control some aspect of the needle insertion while allowing the clinician to maintain control over other elements of the insertion. The surgeon-in-the-loop systems focused on in this work take the form of needle steering systems, whereby a steering device compensates for any needle deflection while the clinician manually inserts the needle to the desired target depth. The primary target therapy in this work is prostate brachytherapy, a procedure where long flexible needles are loaded with radioactive seeds, with the needles then being inserted into the patient, to permanently deposit the seeds into the prostate in order to treat cancerous tissue. In current clinical practice, clinicians rely on live ultrasound images to determine the accuracy with which the needles are being placed during insertion. This work aims to use technologies which are already incorporated into clinical practice to minimize the cost and complexity of future clinical integration. The first portion of this thesis thus focuses on needle-tip tracking in ultrasound images. The first work proposes a needle-shape predictive framework for the static case where the needle is fully inserted and thus supported by tissue. The framework uses a mechanical model and incorporates shape information captured in a single ultrasound image near the base of the needle, in order to predict the shape of the entire needle. The second work presents a real-time needle tip-path predictor, incorporating a particle filter and the well known kinematic bicycle model so that needle-tip path prediction can be iteratively updated as each ultrasound image is processed. The second half of this thesis contains the work related to surgeon-in-the-loop needle steering systems. Two control frameworks are developed which both model needle motion using a reduced order 3D kinematic bicycle model. The first controller implements a switching style regulator using ultrasound image deflection feedback and is proven to asymptotically, in the sense of Lyapunov, stabilize the needle deflection to zero. This switching controller uses direct ultrasound image feedback to measure the deflection of the needle and modulates the needle bevel angle, thus steering the needle, to bring the needle tip back to the axis of insertion. The second controller is based on an event triggered control framework which incorporates needle-tip path prediction to determine the future needle tip deflection. The event triggering controller then uses the predicted needle path to optimize the location of various event points, points of particular needle insertion depth, where the control output is changed to steer the needle. The last portion of the thesis contains an image processing algorithm to autonomously detect the contour and location of the prostate within 2D B-mode ultrasound images. This prostate contouring, or segmentation, algorithm can be used to determine the location of the prostate prior to needle insertion to correct for any change in prostate location between the original seed deposition plan and the observed prostate during the insertion procedure.

• Subjects / Keywords

  • Needle Steering
  • Medical Robotics
  • Ultrasound Image Processing

• Graduation date

  Spring 2019

• Type of Item

  Thesis

• Degree

  Doctor of Philosophy

• DOI

  https://doi.org/10.7939/r3-jfks-v334

• License

  Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.