Four-dimensional deformable image registration using trajectory modeling

Abstract

A four-dimensional deformable image registration (4D DIR) algorithm, referred to as 4D local trajectory modeling (4DLTM), is presented and applied to thoracic 4D computed tomography (4DCT) image sets. The theoretical framework on which this algorithm is built exploits the incremental continuity present in 4DCT component images to calculate a dense set of parameterized voxel trajectories through space as functions of time. The spatial accuracy of the 4DLTM algorithm is compared with an alternative registration approach in which component phase to phase (CPP) DIR is utilized to determine the full displacement between maximum inhale and exhale images. A publically available DIR reference database (http://www.dir-lab.com) is utilized for the spatial accuracy assessment. The database consists of ten 4DCT image sets and corresponding manually identified landmark points between the maximum phases. A subset of points are propagated through the expiratory 4DCT component images. Cubic polynomials were found to provide sufficient flexibility and spatial accuracy for describing the point trajectories through the expiratory phases. The resulting average spatial error between the maximum phases was 1.25 mm for the 4DLTM and 1.44 mm for the CPP. The 4DLTM method captures the long-range motion between 4DCT extremes with high spatial accuracy.

Patient-Specific Biomechanical Modeling of the Lung Tumor for Radiation Therapy

“…The proposed patient-specific biomechanical model has been implemented and evaluated on two real patients P1 and P6 from DIR-Lab Dataset (Castillo et al 2009), with small and large breathing amplitudes. First, we have evaluated the motion estimation accuracy by comparing the simulation results with ground truth (CT images) on 75 landmarks, where the mechanical properties are given in (Fuerst et al 2015, Ladjal et al2015.…”

Statistical Motion Mask and Sliding Registration

“…These errors are based on manually identified correspondences and thus are subject to inter-and intra-observer errors which range between 0.70mm and 1.13mm [2]. This poses a lower limit on the measurable accuracy.…”

“…The first ten image sets were taken from the publicly available DIR-Lab dataset [2,3]. The remaining datasets were acquired as part of standard clinical practice from patients with either earlystage (14 cases) or locally-advanced (8 cases) non-small cell lung cancer.…”

“…The sliding framework is presented in section 2.2. The results are quantitatively evaluated on the publicly available DIR-Lab dataset [2,3].…”

A comparison between adaptive kernel density estimation and Gaussian Mixture Regression for real-time tumour motion prediction from external surface motion

“…Datasets 1-5 were acquired using a helical 4D CT system located at the Lon Brard Cancer Center [8]. Datasets 6-8 were acquired on a cine 4D CT system located in the MD Anderson Cancer Center [9]. The 9th dataset was a cine 4D CT from Royal Surrey Country Hospital.…”