Screening programs for lung cancer have led to a significant increase in the number of biopsies of suspicious lesions. Transbronchial needle biopsy (TBNbx) is preferred over other options because it is minimally invasive, and it has been shown to be more comfortable and much safer than both CT guided percutaneous needle biopsy and surgical biopsies. One safe, dose reducing and time-saving technique to guide TBNbx is electromagnetic navigation bronchoscopy (ENB), which guides biopsy tools to predetermined locations within the bronchial tree using manual registration between a prior CT and an electromagnetic navigation board and sensor probe. The accuracy of ENB degrades for lesion biopsies farther down the bronchial tree, mainly due to the CT- to-body divergence (a miss-registration between the prior CT and the body at the time of intervention) due to breathing motion, body motion, etc. Therefore, current literature suggests a need for a multimodality guidance strategy that combines imaging techniques with ENB to provide real-time image guidance (IG). We propose a new clinically viable 3D real-time imaging system that : 1) permits real-time simultaneous visualization of the target nodule, bronchial tree, and the needle/biopsy instrument, 2) provides accurate, high resolution 3D volume information to provide localization of the target and needle, and 3) maintains patient and operator dose as low as possible. We base our approach on the 'inverse geometry'scanned beam digital x-ray (SBDX) system - a large-area scanned anode x-ray source and a small- area very fast digital detector. Using this system, tomosynthesis images of an object (in-plane resolution 0.2mm, slice thickness 5mm) can be acquired in as little as 1/30 of a second with radiation dose levels comparable to conventional fluoroscopy. The specific goals of this proposal are: 1) to optimize the SBDX hardware for lung nodule detection;2) to develop real-time, concurrent reconstruction hardware and software for SBDX data reconstruction;and 3) to demonstrate improved accuracy and yield of TBNbx using SBDX and ENB in an ex vivo breathing lung model. While we have focused on guidance of TBNbx, the potential applications for our system go far beyond this specific (though important) application. Our optimized system could also be used to obtain first-pass perfusion images of iodinated contrast through nodules and lung tissue, and could facilitate the development of minimally invasive procedures for 'on the spot'resection or therapy of small, cancerous lesions.
We propose a new 3D real-time tomosynthesis imaging system to be used in conjunction with electromagnetic navigation bronchoscopy for image guidance of transbronchial needle biopsy procedures. The proposed system will provide 3D real-time visualization of the target nodule, bronchial tree, and the biopsy instrument with radiation levels comparable to conventional fluoroscopy. Our proposed system has the potential to directly contribute to a more effective and potentially earlier diagnosis of lung cancer.
