Abstract
Robustness to obstacles is the most important factor necessary to achieve accurate tumor tracking without fiducial markers. Some high-density structures, such as bone, are enhanced on X-ray fluoroscopic images, which cause tumor mistracking. Tumor tracking should be performed by controlling “importance recognition”: the understanding that soft-tissue is an important tracking feature and bone structure is unimportant. We propose a new real-time tumor-contouring method that uses deep learning with importance recognition control. The novelty of the proposed method is the combination of the devised random overlay method and supervised deep learning to induce the recognition of structures in tumor contouring as important or unimportant. This method can be used for tumor contouring because it uses deep learning to perform image segmentation. Our results from a simulated fluoroscopy model showed accurate tracking of a low-visibility tumor with an error of approximately 1 mm, even if enhanced bone structure acted as an obstacle. A high similarity of approximately 0.95 on the Jaccard index was observed between the segmented and ground truth tumor regions. A short processing time of 25 ms was achieved. The results of this simulated fluoroscopy model support the feasibility of robust real-time tumor contouring with fluoroscopy. Further studies using clinical fluoroscopy are highly anticipated.
Highlights
Several motion-management techniques have been developed to irradiate targets that move due to respiratory motion, such as lung and liver tumors
The novelty of the proposed method is the combination of the devised random overlay method and supervised deep learning to induce the recognition of structures in tumor contouring as important or unimportant
We believe that the bone-suppression method is unnecessary for tumor tracking if other methods can directly recognize important and unimportant features for tracking, as that occurs in human image recognition
Summary
Several motion-management techniques have been developed to irradiate targets that move due to respiratory motion, such as lung and liver tumors. The second breakthrough was the development of a real-time tumor-tracking method that uses X-ray fluoroscopy [2, 3]. Tracking methods that use bone-suppressed fluoroscopic images have been reported recently [15,16,17] These bone-suppressed images were generated using a dual-energy fluoroscopy system [15, 16] or by a special software using an artificial neural network (ANN) [17]. Since this method suppressed the obstacle features in images, improved tracking accuracy could be expected.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have