GateCT™ surface tracking system for respiratory signal reconstruction in 4DCT imaging

• DOI: 10.1118/1.3671941
• Publication Date: 2011-12-30T18:24:02Z
• Authors: Kevin I. Kauweloa; Dan Ruan; Justin C. Park; Ajay Sandhu; Gwe Ya Kim; Todd Pawlicki; W. Tyler Watkins; Bongyong Song; William Y. Song
• Keywords: Ground truth; Tracking (education); SIGNAL (programming language); Position (finance)

Purpose: To assess the temporal and spatial accuracy of GateCT™ system (VisionRT, London, UK), a recently released respiratory tracking for 4DCT, under both ideal nonideal conditions. Methods: Three experiments were performed by benchmarking comparing its results with ground-truth input data those generated widely used Varian RPM™ (Real-time Position Management, Varian, Palo Alto, CA). The first experiment 10 sinusoidal breathing patterns (constant amplitude frequency using sin6ωt), “consistent” patient patterns, “sporadic” patterns. Motion was simulated quasar™ Programmable Respiratory Platform (MODUS, Canada) as surrogate. systems to track from then analyzed in Fourier domain, evaluate temporal/phase accuracy, Pearson's correlation coefficient (PCC). analysis correlated against results, respectively. second five period 2.0 s, 5.0 varying abdominal amplitudes found clinical cases (peak-to-peak range: 1.67–10 mm) test sensitivity reconstruct various range motion. And, third 12 consecutive patients motion simultaneously systems. baseline two via mean-position-estimate (MPE) calculations. All tracked at least 120 s. Results: In experiment, average PCC values (±SD) all thirty 0.9995 ± 0.00035 0.9994 0.00041 system, These nearly identical demonstrated similar however, revealed pattern which uncertainty mean-position increased decreased. For example, non-negligible drift up 29.3% respect peak-to-peak 1.67-mm observed. On contrary, displayed more consistent recording over time greatest being <7.7%. confirmed these findings setting. Consistent decrease due increase artificial drifts, decreased, found. lowest value 0.7239 1.57-mm amplitude. Conclusions: consistency but had limitations accurately absolute positions, thus suggesting appropriateness phase-sorting 4DCT rather than amplitude-sorting. contrast, stable signal ranges phase amplitude, is robust use amplitude-sorting techniques. impact observed on resulting image quality, amplitude-sorting, needs further investigation.