Real-time volumetric image reconstruction and 3D tumor localization based on a single x-ray projection image for lung cancer radiotherapy

Ruijiang Li, Xun Jia, John H. Lewis, Xuejun Gu, Michael Folkerts, Chunhua Men, Steve B. Jiang

Research output: Contribution to journalArticle

65 Citations (Scopus)

Abstract

Purpose: To develop an algorithm for real-time volumetric image reconstruction and 3D tumor localization based on a single x-ray projection image for lung cancer radiotherapy. Methods: Given a set of volumetric images of a patient at N breathing phases as the training data, deformable image registration was performed between a reference phase and the other N-1 phases, resulting in N-1 deformation vector fields (DVFs). These DVFs can be represented efficiently by a few eigenvectors and coefficients obtained from principal component analysis (PCA). By varying the PCA coefficients, new DVFs can be generated, which, when applied on the reference image, lead to new volumetric images. A volumetric image can then be reconstructed from a single projection image by optimizing the PCA coefficients such that its computed projection matches the measured one. The 3D location of the tumor can be derived by applying the inverted DVF on its position in the reference image. The algorithm was implemented on graphics processing units (GPUs) to achieve real-time efficiency. The training data were generated using a realistic and dynamic mathematical phantom with ten breathing phases. The testing data were 360 cone beam projections corresponding to one gantry rotation, simulated using the same phantom with a 50% increase in breathing amplitude. Results: The average relative image intensity error of the reconstructed volumetric images is 6.9%±2.4%. The average 3D tumor localization error is 0.8±0.5 mm. On an NVIDIA Tesla C1060 GPU card, the average computation time for reconstructing a volumetric image from each projection is 0.24 s (range: 0.17 and 0.35 s). Conclusions: The authors have shown the feasibility of reconstructing volumetric images and localizing tumor positions in 3D in near real-time from a single x-ray image.

Original languageEnglish (US)
Pages (from-to)2822-2826
Number of pages5
JournalMedical Physics
Volume37
Issue number6
DOIs
StatePublished - Jun 2010

Fingerprint

Computer-Assisted Image Processing
Lung Neoplasms
Radiotherapy
Principal Component Analysis
X-Rays
Respiration
Neoplasms

Keywords

  • GPU
  • Localization
  • Lung cancer radiotherapy
  • Reconstruction

ASJC Scopus subject areas

  • Biophysics
  • Radiology Nuclear Medicine and imaging

Cite this

Real-time volumetric image reconstruction and 3D tumor localization based on a single x-ray projection image for lung cancer radiotherapy. / Li, Ruijiang; Jia, Xun; Lewis, John H.; Gu, Xuejun; Folkerts, Michael; Men, Chunhua; Jiang, Steve B.

In: Medical Physics, Vol. 37, No. 6, 06.2010, p. 2822-2826.

Research output: Contribution to journalArticle

@article{8bd1cd4a05d04b919311793ad20da2b0,
title = "Real-time volumetric image reconstruction and 3D tumor localization based on a single x-ray projection image for lung cancer radiotherapy",
abstract = "Purpose: To develop an algorithm for real-time volumetric image reconstruction and 3D tumor localization based on a single x-ray projection image for lung cancer radiotherapy. Methods: Given a set of volumetric images of a patient at N breathing phases as the training data, deformable image registration was performed between a reference phase and the other N-1 phases, resulting in N-1 deformation vector fields (DVFs). These DVFs can be represented efficiently by a few eigenvectors and coefficients obtained from principal component analysis (PCA). By varying the PCA coefficients, new DVFs can be generated, which, when applied on the reference image, lead to new volumetric images. A volumetric image can then be reconstructed from a single projection image by optimizing the PCA coefficients such that its computed projection matches the measured one. The 3D location of the tumor can be derived by applying the inverted DVF on its position in the reference image. The algorithm was implemented on graphics processing units (GPUs) to achieve real-time efficiency. The training data were generated using a realistic and dynamic mathematical phantom with ten breathing phases. The testing data were 360 cone beam projections corresponding to one gantry rotation, simulated using the same phantom with a 50{\%} increase in breathing amplitude. Results: The average relative image intensity error of the reconstructed volumetric images is 6.9{\%}±2.4{\%}. The average 3D tumor localization error is 0.8±0.5 mm. On an NVIDIA Tesla C1060 GPU card, the average computation time for reconstructing a volumetric image from each projection is 0.24 s (range: 0.17 and 0.35 s). Conclusions: The authors have shown the feasibility of reconstructing volumetric images and localizing tumor positions in 3D in near real-time from a single x-ray image.",
keywords = "GPU, Localization, Lung cancer radiotherapy, Reconstruction",
author = "Ruijiang Li and Xun Jia and Lewis, {John H.} and Xuejun Gu and Michael Folkerts and Chunhua Men and Jiang, {Steve B.}",
year = "2010",
month = "6",
doi = "10.1118/1.3426002",
language = "English (US)",
volume = "37",
pages = "2822--2826",
journal = "Medical Physics",
issn = "0094-2405",
publisher = "AAPM - American Association of Physicists in Medicine",
number = "6",

}

TY - JOUR

T1 - Real-time volumetric image reconstruction and 3D tumor localization based on a single x-ray projection image for lung cancer radiotherapy

AU - Li, Ruijiang

AU - Jia, Xun

AU - Lewis, John H.

AU - Gu, Xuejun

AU - Folkerts, Michael

AU - Men, Chunhua

AU - Jiang, Steve B.

PY - 2010/6

Y1 - 2010/6

N2 - Purpose: To develop an algorithm for real-time volumetric image reconstruction and 3D tumor localization based on a single x-ray projection image for lung cancer radiotherapy. Methods: Given a set of volumetric images of a patient at N breathing phases as the training data, deformable image registration was performed between a reference phase and the other N-1 phases, resulting in N-1 deformation vector fields (DVFs). These DVFs can be represented efficiently by a few eigenvectors and coefficients obtained from principal component analysis (PCA). By varying the PCA coefficients, new DVFs can be generated, which, when applied on the reference image, lead to new volumetric images. A volumetric image can then be reconstructed from a single projection image by optimizing the PCA coefficients such that its computed projection matches the measured one. The 3D location of the tumor can be derived by applying the inverted DVF on its position in the reference image. The algorithm was implemented on graphics processing units (GPUs) to achieve real-time efficiency. The training data were generated using a realistic and dynamic mathematical phantom with ten breathing phases. The testing data were 360 cone beam projections corresponding to one gantry rotation, simulated using the same phantom with a 50% increase in breathing amplitude. Results: The average relative image intensity error of the reconstructed volumetric images is 6.9%±2.4%. The average 3D tumor localization error is 0.8±0.5 mm. On an NVIDIA Tesla C1060 GPU card, the average computation time for reconstructing a volumetric image from each projection is 0.24 s (range: 0.17 and 0.35 s). Conclusions: The authors have shown the feasibility of reconstructing volumetric images and localizing tumor positions in 3D in near real-time from a single x-ray image.

AB - Purpose: To develop an algorithm for real-time volumetric image reconstruction and 3D tumor localization based on a single x-ray projection image for lung cancer radiotherapy. Methods: Given a set of volumetric images of a patient at N breathing phases as the training data, deformable image registration was performed between a reference phase and the other N-1 phases, resulting in N-1 deformation vector fields (DVFs). These DVFs can be represented efficiently by a few eigenvectors and coefficients obtained from principal component analysis (PCA). By varying the PCA coefficients, new DVFs can be generated, which, when applied on the reference image, lead to new volumetric images. A volumetric image can then be reconstructed from a single projection image by optimizing the PCA coefficients such that its computed projection matches the measured one. The 3D location of the tumor can be derived by applying the inverted DVF on its position in the reference image. The algorithm was implemented on graphics processing units (GPUs) to achieve real-time efficiency. The training data were generated using a realistic and dynamic mathematical phantom with ten breathing phases. The testing data were 360 cone beam projections corresponding to one gantry rotation, simulated using the same phantom with a 50% increase in breathing amplitude. Results: The average relative image intensity error of the reconstructed volumetric images is 6.9%±2.4%. The average 3D tumor localization error is 0.8±0.5 mm. On an NVIDIA Tesla C1060 GPU card, the average computation time for reconstructing a volumetric image from each projection is 0.24 s (range: 0.17 and 0.35 s). Conclusions: The authors have shown the feasibility of reconstructing volumetric images and localizing tumor positions in 3D in near real-time from a single x-ray image.

KW - GPU

KW - Localization

KW - Lung cancer radiotherapy

KW - Reconstruction

UR - http://www.scopus.com/inward/record.url?scp=77953527375&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77953527375&partnerID=8YFLogxK

U2 - 10.1118/1.3426002

DO - 10.1118/1.3426002

M3 - Article

C2 - 20632593

AN - SCOPUS:77953527375

VL - 37

SP - 2822

EP - 2826

JO - Medical Physics

JF - Medical Physics

SN - 0094-2405

IS - 6

ER -