A navigation system for an ultrathin scanning fiber bronchoscope in the peripheral airways

Transbronchial biopsy of peripheral lung nodules is hindered by inability to access lesions endobronchially due to the large diameter of conventional bronchoscopes. An ultrathin scanning fiber endoscope (SFE) has recently been developed to advance image-guided biopsy several branching generations deeper into the peripheral airways. With this new technology, high-resolution, full-color images can be acquired at video frames rates from within the small peripheral airways that extend out to these peripheral nodules. However, navigating a potentially complex 3D path to the region of interest presents a significant challenge to the bronchoscopist, whose working knowledge of the airway anatomy is limited to the more central lung. To promote minimally invasive and accurate biopsy of peripheral nodules using this new device, a guidance system was developed to track the SFE within the airway anatomy and direct the bronchoscopist to region of interest via a user interface. Assisted navigation is broken into preoperative and intraoperative stages. From a preoperative planning session, the bronchoscopist identifies the lesion location and defines a path to navigate to the desired biopsy site. During bronchoscopy, an electromagnetic sensor tracks the position and pose of the SFE, which is displayed on the preoperative CT image. At each bifurcation, the predefined path directs the bronchoscopist to the region of interest where biopsy is performed. This dissertation outlines the guidance system development and its validation in live animal experiments. First, image analysis software was developed to construct a virtual airway model from CT image data, providing an anatomical map. Assisted navigation was tested using electromagnetic tracking (EMT) within a rigid airway model. In considering future navigation within a live subject, an analysis of airway deformation was performed. Lung motion due to breathing was quantified and modeled using deformable registration of multiple CT scans acquired at various levels of lung inflation. In conjunction with EMT-based localization, image-based tracking (IBT) also permitted localization of the SFE by registration between real and virtual bronchoscopic images. Ultimately, a hybrid tracking strategy was adopted by combining EMT and IBT tracking. At each video frame, the position of the SFE is approximated by the position sensor and then optimized using the video images themselves to reconcile localization errors introduced by EMT system registration and deformation of the anatomy. The hybrid tracking system presented in this dissertation is a novel approach to SFE localization within peripheral airways. As part of this strategy, a means of respiratory motion compensation is integrated to account, for large excursions undergone by peripheral lung regions during breathing. Preliminary in vivo swine studies verify that the SFE can be adequately tracked within peripheral airways, providing guidance that is crucial for navigation and biopsy of peripheral lesions. The greater clinical impact of a trackable SFE may be earlier and more accurate diagnosis of peripheral lesions, resulting in reduced financial cost and compromise to patient health.