Iterative inversion of deformation vector fields with feedback control

Dubey, Abhishek; Iliopoulos, Alexandros-Stavros; Sun, Xiaobai; Yin, Fang‐Fang; Ren, Lei

doi:10.1002/mp.12962

Cited by 5 publications

(7 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Accurate inversion of the DVFs generated by DIR is also a requirement. We used a straightforward fixed‐point iterative method, but other more advanced methods could be explored in future work . The type of motion model used, and its ability to correctly represent anatomical motion, is another critical component.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Reconstruction of a high‐quality volumetric image and a respiratory motion model from patient CBCT projections

et al. 2019

View full text Add to dashboard Cite

Purpose To develop and evaluate a method of reconstructing a patient‐ and treatment day‐ specific volumetric image and motion model from free‐breathing cone‐beam projections and respiratory surrogate measurements. This Motion‐Compensated Simultaneous Algebraic Reconstruction Technique (MC‐SART) generates and uses a motion model derived directly from the cone‐beam projections, without requiring prior motion measurements from 4DCT, and can compensate for both inter‐ and intrabin deformations. The motion model can be used to generate images at arbitrary breathing points, which can be used for estimating volumetric images during treatment delivery. Methods The MC‐SART was formulated using simultaneous image reconstruction and motion model estimation. For image reconstruction, projections were first binned according to external surrogate measurements. Projections in each bin were used to reconstruct a set of volumetric images using MC‐SART. The motion model was estimated based on deformable image registration between the reconstructed bins, and least squares fitting to model parameters. The model was used to compensate for motion in both projection and backprojection operations in the subsequent image reconstruction iterations. These updated images were then used to update the motion model, and the two steps were alternated between. The final output is a volumetric reference image and a motion model that can be used to generate images at any other time point from surrogate measurements. Results A retrospective patient dataset consisting of eight lung cancer patients was used to evaluate the method. The absolute intensity differences in the lung regions compared to ground truth were 50.8 ± 43.9 HU in peak exhale phases (reference) and 80.8 ± 74.0 in peak inhale phases (generated). The 50th percentile of voxel registration error of all voxels in the lung regions with >5 mm amplitude was 1.3 mm. The MC‐SART was also applied to measured patient cone‐beam projections acquired with a linac‐mounted CBCT system. Results from this patient data demonstrate the feasibility of MC‐SART and showed qualitative image quality improvements compared to other state‐of‐the‐art algorithms. Conclusion We have developed a simultaneous image reconstruction and motion model estimation method that uses Cone‐beam computed tomography (CBCT) projections and respiratory surrogate measurements to reconstruct a high‐quality reference image and motion model of a patient in treatment position. The method provided superior performance in both HU accuracy and positional accuracy compared to other existing methods. The resultant reference image and motion model can be combined with respiratory surrogate measurements to generate volumetric images representing patient anatomy at arbitrary time points.

show abstract

Section: Discussionmentioning

confidence: 99%

“…We used a straightforward fixed-point iterative method, but other more advanced methods could be explored in future work. 27,47,48 The type of motion model used, and its ability to correctly represent anatomical motion, is another critical component. Previous studies on the 5D model have reported uncertainties of~2 mm.…”

Section: Discussionmentioning

confidence: 99%

Reconstruction of a high‐quality volumetric image and a respiratory motion model from patient CBCT projections

et al. 2019

View full text Add to dashboard Cite

show abstract

“…We use an iterative inversion procedure and employ adaptive bi-residual feedback control to achieve global convergence and local acceleration (Dubey, 2018;Dubey, Iliopoulos, Sun, Yin, & Ren, 2018). The inverse DVF estimate and the forward DVF must meet the inverse consistency (IC) condition.…”

mentioning

confidence: 99%

“…We use them as feedback in a two-phase iteration. In the first phase, we modulate the study-domain IC residual with adaptive feedback control for guaranteed global convergence, under certain mild conditions (Dubey et al, 2018). Once the error is made sufficiently small, we switch to phase two where we use the reference-domain IC residual and achieve locally quadratic convergence rate.…”

mentioning

confidence: 99%

See 1 more Smart Citation

idvf: Iterative Inversion of Deformation Vector Field with Adaptive Bi-residual Feedback Control

Iliopoulos¹,

Dubey²,

Sun³

2019

JOSS

Self Cite

View full text Add to dashboard Cite

We provide a package for fast and accurate inversion of a deformation vector field (DVF). A DVF, also known as flow or dense motion field, describes a non-linear mapping between a study image and a reference image. DVFs are fundamental to several medical image analysis and computer vision applications. The inverse DVF is often needed together with the forward DVF to enable back-and-forth mappings, and composite mappings among multiple images. Relevant applications include 4D image reconstruction, anatomical atlas generation, dose accumulation estimation in adaptive radiotherapy, simultaneous deformable registration, and symmetric registration completion. To our knowledge, no other package is available for DVF inversion with guaranteed convergence.

show abstract

Ab initio nonrigid X-ray nanotomography

et al. 2019

View full text Add to dashboard Cite

Reaching the full potential of X-ray nanotomography, in particular for biological samples, is limited by many factors, of which one of the most serious is radiation damage. Although sample deformation caused by radiation damage can be partly mitigated by cryogenic protection, it is still present in these conditions and, as we exemplify here using a specimen extracted from scales of the Cyphochilus beetle, it will pose a limit to the achievable imaging resolution. We demonstrate a generalized tomographic model, which optimally follows the sample morphological changes and attempts to recover the original sample structure close to the ideal, damage-free reconstruction. Whereas our demonstration was performed using ptychographic X-ray tomography, the method can be adopted for any tomographic imaging modality. Our application demonstrates improved reconstruction quality of radiation-sensitive samples, which will be of increasing relevance with the higher brightness of 4th generation synchrotron sources.

show abstract

Iterative inversion of deformation vector fields with feedback control

Cited by 5 publications

References 17 publications

Reconstruction of a high‐quality volumetric image and a respiratory motion model from patient CBCT projections

Reconstruction of a high‐quality volumetric image and a respiratory motion model from patient CBCT projections

idvf: Iterative Inversion of Deformation Vector Field with Adaptive Bi-residual Feedback Control

Ab initio nonrigid X-ray nanotomography

Contact Info

Product

Resources

About