Spectrally Constrained L1-Norm Improves Quantitative Accuracy of Diffuse Optical Tomography

Wenqi Lu; Iain Styles

doi:10.1117/12.2285996

Spectrally Constrained L1-Norm Improves Quantitative Accuracy of Diffuse Optical Tomography

Wenqi Lu^*, Iain Styles

^*Corresponding author for this work

Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

We study L1-regularisation of spectrally constrained DOT using three popular
algorithms: IRLS, ADMMM, and FISTA. In 3D simulations, we find that L1-regularised solutions are significantly more highly correlated with the ground truth than Tikhonov-regularised solutions.

Original language	English
Title of host publication	Diffuse Optical Spectroscopy and Imaging VI
Publisher	Optica Publishing Group (formerly OSA)
ISBN (Electronic)	9781510612822
DOIs	https://doi.org/10.1117/12.2285996
Publication status	Published - Jun 2017
Event	2017 European Conferences on Biomedical Optics (ECBO) - Munich, Germany Duration: 25 Jun 2017 → 29 Jun 2017

Publication series

Name	SPIE Proceedings
Publisher	Optica Publishing Group
Volume	10412

Conference

Conference	2017 European Conferences on Biomedical Optics (ECBO)
Country/Territory	Germany
City	Munich
Period	25/06/17 → 29/06/17

Access to Document

10.1117/12.2285996Licence: None: All rights reserved

Cite this

@inproceedings{01ba704a795d46ab8ce666677641ef9d,

title = "Spectrally Constrained L1-Norm Improves Quantitative Accuracy of Diffuse Optical Tomography",

abstract = "We study L1-regularisation of spectrally constrained DOT using three popularalgorithms: IRLS, ADMMM, and FISTA. In 3D simulations, we find that L1-regularised solutions are significantly more highly correlated with the ground truth than Tikhonov-regularised solutions.",

author = "Wenqi Lu and Iain Styles",

year = "2017",

month = jun,

doi = "10.1117/12.2285996",

language = "English",

series = "SPIE Proceedings",

publisher = "Optica Publishing Group (formerly OSA)",

booktitle = "Diffuse Optical Spectroscopy and Imaging VI",

note = "2017 European Conferences on Biomedical Optics (ECBO) ; Conference date: 25-06-2017 Through 29-06-2017",

}

TY - GEN

T1 - Spectrally Constrained L1-Norm Improves Quantitative Accuracy of Diffuse Optical Tomography

AU - Lu, Wenqi

AU - Styles, Iain

PY - 2017/6

Y1 - 2017/6

N2 - We study L1-regularisation of spectrally constrained DOT using three popularalgorithms: IRLS, ADMMM, and FISTA. In 3D simulations, we find that L1-regularised solutions are significantly more highly correlated with the ground truth than Tikhonov-regularised solutions.

AB - We study L1-regularisation of spectrally constrained DOT using three popularalgorithms: IRLS, ADMMM, and FISTA. In 3D simulations, we find that L1-regularised solutions are significantly more highly correlated with the ground truth than Tikhonov-regularised solutions.

UR - https://www.osapublishing.org/conference.cfm?meetingid=62

U2 - 10.1117/12.2285996

DO - 10.1117/12.2285996

M3 - Conference contribution

T3 - SPIE Proceedings

BT - Diffuse Optical Spectroscopy and Imaging VI

PB - Optica Publishing Group (formerly OSA)

T2 - 2017 European Conferences on Biomedical Optics (ECBO)

Y2 - 25 June 2017 through 29 June 2017

ER -

Spectrally Constrained L1-Norm Improves Quantitative Accuracy of Diffuse Optical Tomography

Abstract

Publication series

Conference

Access to Document

Fingerprint

Cite this