Gaussian process classification of superparamagnetic relaxometry data: Phantom study

Javad Sovizi, Kelsey B. Mathieu, Sara L. Thrower, Wolfgang Stefan, John D. Hazle, David Fuentes

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Motivation Superparamagnetic relaxometry (SPMR) is an emerging technology that holds potential for use in early cancer detection. Measurement of the magnetic field after the excitation of cancer-bound superparamagnetic iron oxide nanoparticles (SPIONs) enables the reconstruction of SPIONs spatial distribution and hence tumor detection. However, image reconstruction often requires solving an ill-posed inverse problem that is computationally challenging and sensitive to measurement uncertainty. Moreover, an additional image processing module is required to automatically detect and localize the tumor in the reconstructed image. Objective Our goal is to examine the use of data-driven machine learning technique to detect a weak signal induced by a small cluster of SPIONs (surrogate tumor) in presence of background signal and measurement uncertainty. We aim to investigate the performance of both data-driven and image reconstruction models to characterize situations that one can replace the computationally-challenging reconstruction technique by the data-driven model. Methods We utilize Gaussian process (GP) classification model and a physics-based image reconstruction method, tailored to SPMR datasets that are obtained from (i) in silico simulations designed based on mouse cancer models and (ii) phantom experiments using MagSense system (Imagion Biosystems, Inc.). We investigate the performance of the GP classifier against the reconstruction technique, for different levels of measurement noise, different scenarios of SPIONs distribution, and different concentrations of SPIONs at the surrogate tumor. Results In our in silico source detection analysis, we were able to achieve high sensitivity results using GP model that outperformed the image reconstruction model for various choices of SPIONs concentration at the surrogate tumor and measurement noise levels. Moreover, in our phantom studies we were able to detect the surrogate tumor phantoms with 5% and 7.3% of the total used SPIONs, surrounded by 9 low-concentration phantoms with accuracies of 87.5% and 96.4%, respectively. Conclusions The GP framework provides acceptable classification accuracies when dealing with in silico and phantom SPMR datasets and can outperform an image reconstruction method for binary classification of SPMR data.

Original languageEnglish (US)
Pages (from-to)47-59
Number of pages13
JournalArtificial Intelligence in Medicine
Volume82
DOIs
StatePublished - Oct 2017

Fingerprint

Iron oxides
Nanoparticles
Tumors
Computer-Assisted Image Processing
Image reconstruction
Neoplasms
Computer Simulation
Uncertainty
Noise
Inverse problems
Physics
Spatial distribution
Magnetic Fields
ferric oxide
Learning systems
Early Detection of Cancer
Image processing
Classifiers
Magnetic fields
Technology

Keywords

  • Gaussian process
  • Superparamagnetic relaxometry
  • Weak source detection

ASJC Scopus subject areas

  • Medicine (miscellaneous)
  • Artificial Intelligence

Cite this

Gaussian process classification of superparamagnetic relaxometry data : Phantom study. / Sovizi, Javad; Mathieu, Kelsey B.; Thrower, Sara L.; Stefan, Wolfgang; Hazle, John D.; Fuentes, David.

In: Artificial Intelligence in Medicine, Vol. 82, 10.2017, p. 47-59.

Research output: Contribution to journalArticle

@article{7a10953562c447b0b4f9344ecd6f010f,
title = "Gaussian process classification of superparamagnetic relaxometry data: Phantom study",
abstract = "Motivation Superparamagnetic relaxometry (SPMR) is an emerging technology that holds potential for use in early cancer detection. Measurement of the magnetic field after the excitation of cancer-bound superparamagnetic iron oxide nanoparticles (SPIONs) enables the reconstruction of SPIONs spatial distribution and hence tumor detection. However, image reconstruction often requires solving an ill-posed inverse problem that is computationally challenging and sensitive to measurement uncertainty. Moreover, an additional image processing module is required to automatically detect and localize the tumor in the reconstructed image. Objective Our goal is to examine the use of data-driven machine learning technique to detect a weak signal induced by a small cluster of SPIONs (surrogate tumor) in presence of background signal and measurement uncertainty. We aim to investigate the performance of both data-driven and image reconstruction models to characterize situations that one can replace the computationally-challenging reconstruction technique by the data-driven model. Methods We utilize Gaussian process (GP) classification model and a physics-based image reconstruction method, tailored to SPMR datasets that are obtained from (i) in silico simulations designed based on mouse cancer models and (ii) phantom experiments using MagSense system (Imagion Biosystems, Inc.). We investigate the performance of the GP classifier against the reconstruction technique, for different levels of measurement noise, different scenarios of SPIONs distribution, and different concentrations of SPIONs at the surrogate tumor. Results In our in silico source detection analysis, we were able to achieve high sensitivity results using GP model that outperformed the image reconstruction model for various choices of SPIONs concentration at the surrogate tumor and measurement noise levels. Moreover, in our phantom studies we were able to detect the surrogate tumor phantoms with 5{\%} and 7.3{\%} of the total used SPIONs, surrounded by 9 low-concentration phantoms with accuracies of 87.5{\%} and 96.4{\%}, respectively. Conclusions The GP framework provides acceptable classification accuracies when dealing with in silico and phantom SPMR datasets and can outperform an image reconstruction method for binary classification of SPMR data.",
keywords = "Gaussian process, Superparamagnetic relaxometry, Weak source detection",
author = "Javad Sovizi and Mathieu, {Kelsey B.} and Thrower, {Sara L.} and Wolfgang Stefan and Hazle, {John D.} and David Fuentes",
year = "2017",
month = "10",
doi = "10.1016/j.artmed.2017.07.001",
language = "English (US)",
volume = "82",
pages = "47--59",
journal = "Artificial Intelligence in Medicine",
issn = "0933-3657",
publisher = "Elsevier",

}

TY - JOUR

T1 - Gaussian process classification of superparamagnetic relaxometry data

T2 - Phantom study

AU - Sovizi, Javad

AU - Mathieu, Kelsey B.

AU - Thrower, Sara L.

AU - Stefan, Wolfgang

AU - Hazle, John D.

AU - Fuentes, David

PY - 2017/10

Y1 - 2017/10

N2 - Motivation Superparamagnetic relaxometry (SPMR) is an emerging technology that holds potential for use in early cancer detection. Measurement of the magnetic field after the excitation of cancer-bound superparamagnetic iron oxide nanoparticles (SPIONs) enables the reconstruction of SPIONs spatial distribution and hence tumor detection. However, image reconstruction often requires solving an ill-posed inverse problem that is computationally challenging and sensitive to measurement uncertainty. Moreover, an additional image processing module is required to automatically detect and localize the tumor in the reconstructed image. Objective Our goal is to examine the use of data-driven machine learning technique to detect a weak signal induced by a small cluster of SPIONs (surrogate tumor) in presence of background signal and measurement uncertainty. We aim to investigate the performance of both data-driven and image reconstruction models to characterize situations that one can replace the computationally-challenging reconstruction technique by the data-driven model. Methods We utilize Gaussian process (GP) classification model and a physics-based image reconstruction method, tailored to SPMR datasets that are obtained from (i) in silico simulations designed based on mouse cancer models and (ii) phantom experiments using MagSense system (Imagion Biosystems, Inc.). We investigate the performance of the GP classifier against the reconstruction technique, for different levels of measurement noise, different scenarios of SPIONs distribution, and different concentrations of SPIONs at the surrogate tumor. Results In our in silico source detection analysis, we were able to achieve high sensitivity results using GP model that outperformed the image reconstruction model for various choices of SPIONs concentration at the surrogate tumor and measurement noise levels. Moreover, in our phantom studies we were able to detect the surrogate tumor phantoms with 5% and 7.3% of the total used SPIONs, surrounded by 9 low-concentration phantoms with accuracies of 87.5% and 96.4%, respectively. Conclusions The GP framework provides acceptable classification accuracies when dealing with in silico and phantom SPMR datasets and can outperform an image reconstruction method for binary classification of SPMR data.

AB - Motivation Superparamagnetic relaxometry (SPMR) is an emerging technology that holds potential for use in early cancer detection. Measurement of the magnetic field after the excitation of cancer-bound superparamagnetic iron oxide nanoparticles (SPIONs) enables the reconstruction of SPIONs spatial distribution and hence tumor detection. However, image reconstruction often requires solving an ill-posed inverse problem that is computationally challenging and sensitive to measurement uncertainty. Moreover, an additional image processing module is required to automatically detect and localize the tumor in the reconstructed image. Objective Our goal is to examine the use of data-driven machine learning technique to detect a weak signal induced by a small cluster of SPIONs (surrogate tumor) in presence of background signal and measurement uncertainty. We aim to investigate the performance of both data-driven and image reconstruction models to characterize situations that one can replace the computationally-challenging reconstruction technique by the data-driven model. Methods We utilize Gaussian process (GP) classification model and a physics-based image reconstruction method, tailored to SPMR datasets that are obtained from (i) in silico simulations designed based on mouse cancer models and (ii) phantom experiments using MagSense system (Imagion Biosystems, Inc.). We investigate the performance of the GP classifier against the reconstruction technique, for different levels of measurement noise, different scenarios of SPIONs distribution, and different concentrations of SPIONs at the surrogate tumor. Results In our in silico source detection analysis, we were able to achieve high sensitivity results using GP model that outperformed the image reconstruction model for various choices of SPIONs concentration at the surrogate tumor and measurement noise levels. Moreover, in our phantom studies we were able to detect the surrogate tumor phantoms with 5% and 7.3% of the total used SPIONs, surrounded by 9 low-concentration phantoms with accuracies of 87.5% and 96.4%, respectively. Conclusions The GP framework provides acceptable classification accuracies when dealing with in silico and phantom SPMR datasets and can outperform an image reconstruction method for binary classification of SPMR data.

KW - Gaussian process

KW - Superparamagnetic relaxometry

KW - Weak source detection

UR - http://www.scopus.com/inward/record.url?scp=85028975826&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85028975826&partnerID=8YFLogxK

U2 - 10.1016/j.artmed.2017.07.001

DO - 10.1016/j.artmed.2017.07.001

M3 - Article

C2 - 28911905

AN - SCOPUS:85028975826

VL - 82

SP - 47

EP - 59

JO - Artificial Intelligence in Medicine

JF - Artificial Intelligence in Medicine

SN - 0933-3657

ER -