The Multidimensional Image Processing Lab (MIPL) at The Pennsylvania State University has
been developing software to help physicians stage and diagnose lung cancer for over twenty years.
In particular, MIPL has developed the Multimodal Virtual Navigation System (MVNS) that can
be used to help physicians navigate to regions of interest (ROIs) for lung biopsies. Since the biopsy
sites are usually located outside of the airways, convex-probe endobronchial ultrasound (CP-EBUS)
for central chest is used to help physicians \see" the target behind the wall of the airway. Our
MVNS can be used for guiding EBUS procedures. Unfortunately, the guidance plan that it uses to
guide the physician to a desired ROI location for EBUS and biopsy is created via a highly tedious
error-prone approach by a technician. In this thesis, we derive an efficient automated approach
for creating guidance plans for EBUS-based biopsies in the central chest and also integrate this
methodology into the MVNS for technician-free image-guided EBUS.