Open Collections

A system for fused ultrasound-MR image guidance for robot-assisted radical prostatectomy

University of British Columbia

Robot-assisted laparoscopic radical prostatectomy is a surgical operation where the entire prostate gland is removed. This complex procedure requires the surgeon to establish a fine balance between completely removing the cancer while sparing critical anatomy responsible for continence and potency. Much of the difficulty of this surgery lies in the inability for the surgeon to actively localize the tumours during the surgery and distinguish between cancerous tissue to be removed and healthy tissue to be left behind. This thesis details the design and development of an image guidance system for radical prostatectomy with the goal of improving patient outcomes by providing the surgeon with improved cancer localization. An image guidance system was proposed, consisting of pre-operatively segmented Magnetic Resonance (MR) images registered to intraoperative transrectal ultrasound rendered in a 3D virtual scene. This system builds upon prior work done by previous members of the lab who developed a TRUS robot which is able to perform a da Vinci-TRUS registration as well as a separate system for MRI-TRUS registration. The registered MRI and TRUS images are presented to the surgeon through the da Vinci surgical console's TilePro display system. Improvements to this prior system in this thesis include specifically: rendering the MRI and TRUS in a visual representation of the scene, an imaging pipeline framework for capturing video from multiple sources, improvements to the calibration between da Vinci and TRUS coordinate systems, an augmented reality overlay for the ultrasound image displaying the da Vinci instrument location, prostate and tumour boundaries and a real-time registration algorithm for tracking prostate motion over the course of a surgery was integrated into the main workflow. A series of experiments were conducted in order to validate the system. First, a phantom study was conducted to evaluate the accuracy of the entire registration process. Second, experiments were conducted to measure the runtime performance and latency of the application. Third, the accuracy of the real-time registration algorithm was tested. Lastly, the results of 15 of the ongoing surgical studies using the image guidance system are presented.

Text

2019-12-12

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/72701

Biomedical Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/