Planning and Guidance of Bronchoscopy for Comprehensive Lymph Node Staging

Kuhlengel, Trevor Keith

Planning and Guidance of Bronchoscopy for Comprehensive Lymph Node Staging

Restricted

Author:: Kuhlengel, Trevor Keith
Graduate Program:: Computer Science and Engineering (PHD)
Degree:: Doctor of Philosophy
Document Type:: Dissertation
Date of Defense:: January 28, 2022
Committee Members:: Ali Borhan, Outside Unit & Field Member
Robert Collins, Major Field Member
William Higgins, Chair & Dissertation Advisor
Vishal Monga, Major Field Member
Chitaranjan Das, Program Head/Chair
Keywords:: lung cancer
lymph node staging
multi-destination procedures
image-guided procedures
bronchoscopy
optimal procedure planning
computer-assisted procedure planning
Abstract:: The International Association for the Study of Lung Cancer's (IASLC) TNM lung cancer staging system divides the chest into fourteen anatomical lymph node stations and identifies lymph node N-stages from N0-N3. A patient's lymph node stage determines the available treatment options and prognosis, so accurate staging is crucial for lung cancer management. Best-practice clinical guidelines recommend comprehensively sampling all visible lymph nodes in each central-chest lymph node station using a bronchoscopic lymph node staging procedure. Despite these guidelines, bronchoscopic procedures are rarely comprehensive. On another front, computer-based image-guided bronchoscopy (IGB) systems draw upon the patient's CT images to assist the physician with planning and guidance strategies. However, existing IGB systems focus only on optimizing guidance to a single node at a time, while ignoring optimizations to the overall procedure. The goal of this dissertation is to develop a complete computational system to efficiently plan and guide comprehensive lymph node staging procedures. To realize this goal, we have developed several automatic and robust methods. First, we devised a method to create three-dimensional (3D) models of the IASLC stations and combine them into a model of the nodal staging zones. The method uses the station and zone models to label lymph nodes with station and N-stage descriptors automatically. Subsequently, we devised a multi-destination planning method to optimize whole-procedure plans. This method computes a sequence of paths connecting smoothly from one node to the next, creating an optimized-length ``tour'' to visit all the pre-defined lymph nodes in a case, following best-practice guidelines. Next, we developed a complete IGB system incorporating these comprehensive multi-destination staging procedure plans. This on-line guidance system consists of a novel guidance strategy and supplemental visualizations, which provide optimized live guidance through all the nodes in a procedure. We performed a series of retrospective, phantom, and live studies to evaluate our systems for efficacy, safety, and clinical utility. The retrospective studies tested our proposed methods on a database of 31 human case studies. Next, we completed a multi-physician study testing the on-line guidance system's efficacy using phantom (plastic) airway models. We then performed two live-guided pig studies to evaluate the system's safety in a clinical workflow. Finally, we demonstrated the system's clinical safety and functionality in a live-human clinical case study.

Tools