An in silico performance characterization of respiratory motion guided 4DCT for high-quality low-dose lung cancer imaging. (27th July 2018)

• Record Type: Journal Article
• Authors: Martin, Spencer; O' Brien, Ricky; Hofmann, Christian; Keall, Paul; Kipriditis, John
• Is Part Of: Physics in medicine & biology. Volume 63:Number 15(2018:Aug.)
• Publication Date: 2018-07-27
• Subjects: 4DCT -- respiratory motion -- respiratory-gated imaging -- XCAT phantom -- lung cancer imaging/radiotherapy; Biophysics -- Periodicals; Medical physics -- Periodicals; 610.153
• DOI: 10.1088/1361-6560/aaceca
• Languages: English
• ISSNs: 0031-9155

Abstract

Abstract: This work aims to characterize the performance of an improved 4DCT technique aiming to overcome irregular breathing-related image artifacts. To address this, we have developed respiratory motion guided (RMG) 4DCT, which uses real-time breathing motion analysis to prospectively gate scans based on detection of irregular breathing. This is the first investigation of RMG-4DCT using a real-time software prototype, testing the hypothesis that it can reduce breathing irregularities during imaging, reduce image oversampling and improve image quality compared to a 'conventional' 4DCT protocol without breathing guidance. RMG-4DCT scans were simulated based on 100+ hours of breathing motion acquired for 20 lung cancer patients. Scan performance was quantified in terms of the beam on time (a surrogate for imaging dose), total scan time and the breathing irregularity during imaging (via RMSE of the breathing motion during acquisition). A conventional 4DCT protocol was also implemented using the same software prototype for a direct comparator to the RMG-4DCT results. We investigated the impact of key RMG-4DCT parameters such as gating tolerance, gantry rotation time and the use of baseline drift correction. Using a representative set of algorithm parameters, RMG-4DCT achieved significant mean reductions in estimated imaging dose (−17.8%, p < 0.001) and breathing RMSE during imaging (−12.6%, p < 0.001) compared to conventional 4DCT. These improvements came with increased scan times, roughly doubled on average (104%, p < 0.001). Image quality simulations were performed using the deformable digital XCAT phantom, with image quality quantified based on the normalized cross correlation (NCC) between axial slices. RMG-4DCT demonstrated qualitative image quality improvements for three out of 10 phase bins, however the improvement was not significant across all 10 phases ( p = 0.08) at a population level. In choosing RMG-4DCT scan parameters, the trade-off between gating sensitivity and scan time may be optimized, demonstrating potential for RMG-4DCT as a viable pathway to improve clinical 4DCT imaging.