A study on a method for accelerating bronchoscope tracking based on image registration on GPU

We have developed a method for tracking a bronchoscope by image registration between real and virtual bronchoscopic images. Since the computation cost of image registration is quite expensive, it is difficult to track the bronchoscope in real-time. To solve this problem, we try to accelerate the process of image registration by utilizing GPU (Graphics Processing Unit). Specifically, we accelerate two parts: (1) virtual bronchoscopic image generation by volume rendering, and (2) image similarity calculation between a real and a virtual bronchoscopic images. We applied the proposed method to ten real bronchoscopic videos and corresponding their CT images. The experimental results showed that the proposed method could track the bronchoscope about 16 times faster than the method using only CPU.     «

We have developed a method for tracking a bronchoscope by image registration between real and virtual bronchoscopic images. Since the computation cost of image registration is quite expensive, it is difficult to track the bronchoscope in real-time. To solve this problem, we try to accelerate the process of image registration by utilizing GPU (Graphics Processing Unit). Specifically, we accelerate two parts: (1) virtual bronchoscopic image generation by volume rendering, and (2) image similar...     »