Premium
SU‐D‐213CD‐03: Live Video‐Guided Volumetric Tracking of Respiration Motion
Author(s) -
Li S,
Haiti B,
Serratore D,
Neicu T,
Chan P,
Valakh V,
Bizhan M,
Miyamoto C,
Geng J
Publication year - 2012
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.4734688
Subject(s) - tracking (education) , volume (thermodynamics) , breathing , computer vision , motion (physics) , computer science , physics , synchronization (alternating current) , nuclear medicine , artificial intelligence , medicine , anatomy , psychology , pedagogy , quantum mechanics , computer network , channel (broadcasting)
Purpose: To introduce video surface imaging guidance in synchronization with 4D cone‐beam CT (CBCT) scans, and in combination with respiration‐ gated or target‐tracked dose delivery to treat mobile tumors, without collaterally damaging nearby critical structures. Methods: The approach uses the concept that the integral of balanced forces over the moving surfaces is directly proportional to the lung volume changes. The respiratory motions, representing the lung volume variations, were measured with the dynamic volume under the moving surfaces of the thorax and abdomen. Sequential surface images on several patients and volunteers were acquired for the feasibility study. Respiratory motions were repeatedly measured on volunteers undertaking a quiet (normal) or a forced (deep) breath. The dynamic volume under the moving surfaces were robustly fitted with a linear trend and a trigonometric wave function that was compared with the fitted curves for target moving trajectories derived from forty 4D‐CBCT scans. Results: A large chest wall superior‐outward movement was the unique characteristic of a forced breath that had doubled the volume variations and elongated the respiration period from quiet breath of ∼4 seconds to >6 seconds. Under a quiet breath, target motion trajectories could be easily described by single sine functions that were consistent with dynamic surface volume modeling except for having different motion amplitudes. The accuracy in synchronization of the real‐time surface motion with respiration motion was within the measurement uncertainty of ∼2 mm. Conclusions: The analytical results with a hypothetical single sine platform allow us to accurately predict internal target motion with use of real‐time video images. Synchronization of dynamic volume with respiratory motion appears applicable for association of 4D medical imaging with 4D videoimaging.