Prediction of high-dimensional states subject to respiratory motion: A manifold learning approach

Wenyang Liu, Amit Sawant, Dan Ruan

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

The development of high-dimensional imaging systems in image-guided radiotherapy provides important pathways to the ultimate goal of real-time full volumetric motion monitoring. Effective motion management during radiation treatment usually requires prediction to account for system latency and extra signal/image processing time. It is challenging to predict high-dimensional respiratory motion due to the complexity of the motion pattern combined with the curse of dimensionality. Linear dimension reduction methods such as PCA have been used to construct a linear subspace from the high-dimensional data, followed by efficient predictions on the lower-dimensional subspace. In this study, we extend such rationale to a more general manifold and propose a framework for high-dimensional motion prediction with manifold learning, which allows one to learn more descriptive features compared to linear methods with comparable dimensions. Specifically, a kernel PCA is used to construct a proper low-dimensional feature manifold, where accurate and efficient prediction can be performed. A fixed-point iterative pre-image estimation method is used to recover the predicted value in the original state space. We evaluated and compared the proposed method with a PCA-based approach on level-set surfaces reconstructed from point clouds captured by a 3D photogrammetry system. The prediction accuracy was evaluated in terms of root-mean-squared-error. Our proposed method achieved consistent higher prediction accuracy (sub-millimeter) for both 200 ms and 600 ms lookahead lengths compared to the PCA-based approach, and the performance gain was statistically significant.

Original languageEnglish (US)
Pages (from-to)4989-4999
Number of pages11
JournalPhysics in Medicine and Biology
Volume61
Issue number13
DOIs
StatePublished - Jun 14 2016

Fingerprint

Passive Cutaneous Anaphylaxis
Learning
Photogrammetry
Image-Guided Radiotherapy
Radiation
Therapeutics

Keywords

  • high-dimensional
  • manifold learning
  • prediction
  • respiratory motion

ASJC Scopus subject areas

  • Radiological and Ultrasound Technology
  • Radiology Nuclear Medicine and imaging

Cite this

Prediction of high-dimensional states subject to respiratory motion : A manifold learning approach. / Liu, Wenyang; Sawant, Amit; Ruan, Dan.

In: Physics in Medicine and Biology, Vol. 61, No. 13, 14.06.2016, p. 4989-4999.

Research output: Contribution to journalArticle

Liu, Wenyang ; Sawant, Amit ; Ruan, Dan. / Prediction of high-dimensional states subject to respiratory motion : A manifold learning approach. In: Physics in Medicine and Biology. 2016 ; Vol. 61, No. 13. pp. 4989-4999.
@article{95d716e79e6b4111af5378a25e996436,
title = "Prediction of high-dimensional states subject to respiratory motion: A manifold learning approach",
abstract = "The development of high-dimensional imaging systems in image-guided radiotherapy provides important pathways to the ultimate goal of real-time full volumetric motion monitoring. Effective motion management during radiation treatment usually requires prediction to account for system latency and extra signal/image processing time. It is challenging to predict high-dimensional respiratory motion due to the complexity of the motion pattern combined with the curse of dimensionality. Linear dimension reduction methods such as PCA have been used to construct a linear subspace from the high-dimensional data, followed by efficient predictions on the lower-dimensional subspace. In this study, we extend such rationale to a more general manifold and propose a framework for high-dimensional motion prediction with manifold learning, which allows one to learn more descriptive features compared to linear methods with comparable dimensions. Specifically, a kernel PCA is used to construct a proper low-dimensional feature manifold, where accurate and efficient prediction can be performed. A fixed-point iterative pre-image estimation method is used to recover the predicted value in the original state space. We evaluated and compared the proposed method with a PCA-based approach on level-set surfaces reconstructed from point clouds captured by a 3D photogrammetry system. The prediction accuracy was evaluated in terms of root-mean-squared-error. Our proposed method achieved consistent higher prediction accuracy (sub-millimeter) for both 200 ms and 600 ms lookahead lengths compared to the PCA-based approach, and the performance gain was statistically significant.",
keywords = "high-dimensional, manifold learning, prediction, respiratory motion",
author = "Wenyang Liu and Amit Sawant and Dan Ruan",
year = "2016",
month = "6",
day = "14",
doi = "10.1088/0031-9155/61/13/4989",
language = "English (US)",
volume = "61",
pages = "4989--4999",
journal = "Physics in Medicine and Biology",
issn = "0031-9155",
publisher = "IOP Publishing Ltd.",
number = "13",

}

TY - JOUR

T1 - Prediction of high-dimensional states subject to respiratory motion

T2 - A manifold learning approach

AU - Liu, Wenyang

AU - Sawant, Amit

AU - Ruan, Dan

PY - 2016/6/14

Y1 - 2016/6/14

N2 - The development of high-dimensional imaging systems in image-guided radiotherapy provides important pathways to the ultimate goal of real-time full volumetric motion monitoring. Effective motion management during radiation treatment usually requires prediction to account for system latency and extra signal/image processing time. It is challenging to predict high-dimensional respiratory motion due to the complexity of the motion pattern combined with the curse of dimensionality. Linear dimension reduction methods such as PCA have been used to construct a linear subspace from the high-dimensional data, followed by efficient predictions on the lower-dimensional subspace. In this study, we extend such rationale to a more general manifold and propose a framework for high-dimensional motion prediction with manifold learning, which allows one to learn more descriptive features compared to linear methods with comparable dimensions. Specifically, a kernel PCA is used to construct a proper low-dimensional feature manifold, where accurate and efficient prediction can be performed. A fixed-point iterative pre-image estimation method is used to recover the predicted value in the original state space. We evaluated and compared the proposed method with a PCA-based approach on level-set surfaces reconstructed from point clouds captured by a 3D photogrammetry system. The prediction accuracy was evaluated in terms of root-mean-squared-error. Our proposed method achieved consistent higher prediction accuracy (sub-millimeter) for both 200 ms and 600 ms lookahead lengths compared to the PCA-based approach, and the performance gain was statistically significant.

AB - The development of high-dimensional imaging systems in image-guided radiotherapy provides important pathways to the ultimate goal of real-time full volumetric motion monitoring. Effective motion management during radiation treatment usually requires prediction to account for system latency and extra signal/image processing time. It is challenging to predict high-dimensional respiratory motion due to the complexity of the motion pattern combined with the curse of dimensionality. Linear dimension reduction methods such as PCA have been used to construct a linear subspace from the high-dimensional data, followed by efficient predictions on the lower-dimensional subspace. In this study, we extend such rationale to a more general manifold and propose a framework for high-dimensional motion prediction with manifold learning, which allows one to learn more descriptive features compared to linear methods with comparable dimensions. Specifically, a kernel PCA is used to construct a proper low-dimensional feature manifold, where accurate and efficient prediction can be performed. A fixed-point iterative pre-image estimation method is used to recover the predicted value in the original state space. We evaluated and compared the proposed method with a PCA-based approach on level-set surfaces reconstructed from point clouds captured by a 3D photogrammetry system. The prediction accuracy was evaluated in terms of root-mean-squared-error. Our proposed method achieved consistent higher prediction accuracy (sub-millimeter) for both 200 ms and 600 ms lookahead lengths compared to the PCA-based approach, and the performance gain was statistically significant.

KW - high-dimensional

KW - manifold learning

KW - prediction

KW - respiratory motion

UR - http://www.scopus.com/inward/record.url?scp=84976417404&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84976417404&partnerID=8YFLogxK

U2 - 10.1088/0031-9155/61/13/4989

DO - 10.1088/0031-9155/61/13/4989

M3 - Article

C2 - 27299958

AN - SCOPUS:84976417404

VL - 61

SP - 4989

EP - 4999

JO - Physics in Medicine and Biology

JF - Physics in Medicine and Biology

SN - 0031-9155

IS - 13

ER -