Robust motion tracking in liver from 2D ultrasound images using supporters

Abstract

Purpose: Effectiveness of image-guided radiation therapy with precise dose delivery depends highly on accurate target localization, which may involve motion during treatment due to, e.g., breathing and drift. Therefore, it is important to track the motion and adjust the radiation delivery accordingly. Tracking generally requires reliable target appearance and image features, whereas in ultrasound imaging acoustic shadowing and other artifacts may degrade the visibility of a target, leading to substantial tracking errors. To minimize such errors, we propose a method based on so-called supporters, a computer vision tracking technique. This allows us to leverage information from surrounding motion for improving robustness of motion tracking on 2D ultrasound image sequences of the liver. Methods: Image features, potentially useful for predicting the target positions, are individually tracked and a supporter model capturing the coupling of motion between these features and the target is learned on-line. This model is then applied to predict the target position, when the target cannot be otherwise tracked reliably. Results: The proposed method was evaluated using the Challenge on Liver Ultrasound Tracking (CLUST)-2015 dataset. Leave-one-out cross validation was performed on the training set of 24 2D image sequences of each 1-5 minutes. The method was then applied on the test set (24 2D sequences), where the results were evaluated by the challenge organizers, yielding 1.04 mm mean and 2.26 mm 95%ile tracking error for all targets. We also devised a simulation framework to emulate acoustic shadowing artifacts from the ribs, which showed effective tracking despite the shadows. Conclusions: Results support the feasibility and demonstrate the advantages of using supporters. The proposed method improves its baseline tracker, which uses optic-flow and elliptic vessel models, and yields the state-of-the-art real-time tracking solution for the CLUST challenge.

ELPHA: Dynamically deformable liver phantom for real‐time motion‐adaptive radiotherapy treatments

“…With a simple template‐tracking algorithm, the motion of the vessel structure was quantified postacquisition and compared to the EM position signal. Several real‐time tracking algorithms have been developed, but these are optimized for in vivo liver ultrasound data and thus are suboptimal for silicone imaging. Nevertheless, such methods yield average tracking accuracies below 1 mm in under 50 ms lag and therefore are essential in future in vivo applications on patients.…”

Compressed Sensing Image Reconstruction of Ultrasound Image for Treatment of Early Traumatic Myositis Ossificans of Elbow Joint by Electroacupuncture

“…PDI cannot display the direction and speed of blood flow, as well as direct quantitative values such as resistance index and maximum systolic peak, so it can only be used as a supplement to CDFI [16][17][18]. However, with the in-depth study of CDFI and PDI in musculoskeletal neurological diseases, it has shown good prospects [19,20]. It is believed that high-frequency ultrasound will be an indispensable imaging method for the diagnosis of musculoskeletal neurological diseases in the near future.…”

Deep learning‐based motion tracking using ultrasound images

“…In addition to these registration-based methods, optical flow-based approaches rely on the locational correlation of a set of points between the tracking and reference frames. [35][36][37] Kondo et al 38 proposed a method using kernelized correlation filter-based algorithm with extensions of adaptive window size selection and motion vector refinement with template matching. Shepard et al 39 investigated a similarity measurement-based block matching algorithm incorporating training methods and multiple simultaneous templates to derive an affine transformation between two frames by using normalized cross-correlation as the similarity measurement metric, which achieves one of the best results on CLUST 2015 2D point-landmark tracking currently.…”