Investigation of texture quantification parameters for neurological PET image analysis

Ivan S. Klyuzhin, Stephan Blinder, Rostom Mabrouk, Arman Rahmim, Vesna Sossi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We investigate the correlation between the clinical severity of neurodegenerative disease and texture metrics (such as Haralick features) computed using PET images of the brain. Specifically, we explore how the parameters of feature computation - such as the region of interest definition method, and the direction and distance used for texture quantification - affect the correlation between texture-based image metrics and clinical disease severity. The analysis was based on an ongoing Parkinson's disease imaging study, with co-registered PET and MRI images, and tracer predominantly concentrated in the striatum. Disease duration was used as the primary clinical metric. It was found that the region of interest placement method substantially affected the correlation values. Significant correlation (p<0.01) was obtained when simple box-like regions were used instead of the anatomic MRI-based regions. The used direction affected the correlation values moderately, and distance did not have a pronounced effect. The results suggest that the Haralick features and other texture metrics that do not require kinetic modeling could be potentially used for the analysis of PET images for which the corresponding MRI data are not available. The results also show that the region of interest definition method and the direction along which metrics are computed may affect metric performance.

Original languageEnglish (US)
Title of host publication2015 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781467398626
DOIs
StatePublished - Oct 3 2016
Event2015 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2015 - San Diego, United States
Duration: Oct 31 2015Nov 7 2015

Other

Other2015 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2015
CountryUnited States
CitySan Diego
Period10/31/1511/7/15

Fingerprint

image analysis
Image analysis
textures
Textures
Magnetic resonance imaging
Neurodegenerative diseases
Neurodegenerative Diseases
Parkinson Disease
Parkinson disease
Brain
brain
tracers
boxes
Imaging techniques
Kinetics
Direction compound
kinetics

ASJC Scopus subject areas

  • Nuclear and High Energy Physics
  • Radiology Nuclear Medicine and imaging
  • Instrumentation

Cite this

Klyuzhin, I. S., Blinder, S., Mabrouk, R., Rahmim, A., & Sossi, V. (2016). Investigation of texture quantification parameters for neurological PET image analysis. In 2015 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2015 [7582053] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/NSSMIC.2015.7582053

Investigation of texture quantification parameters for neurological PET image analysis. / Klyuzhin, Ivan S.; Blinder, Stephan; Mabrouk, Rostom; Rahmim, Arman; Sossi, Vesna.

2015 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2015. Institute of Electrical and Electronics Engineers Inc., 2016. 7582053.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Klyuzhin, IS, Blinder, S, Mabrouk, R, Rahmim, A & Sossi, V 2016, Investigation of texture quantification parameters for neurological PET image analysis. in 2015 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2015., 7582053, Institute of Electrical and Electronics Engineers Inc., 2015 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2015, San Diego, United States, 10/31/15. https://doi.org/10.1109/NSSMIC.2015.7582053
Klyuzhin IS, Blinder S, Mabrouk R, Rahmim A, Sossi V. Investigation of texture quantification parameters for neurological PET image analysis. In 2015 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2015. Institute of Electrical and Electronics Engineers Inc. 2016. 7582053 https://doi.org/10.1109/NSSMIC.2015.7582053
Klyuzhin, Ivan S. ; Blinder, Stephan ; Mabrouk, Rostom ; Rahmim, Arman ; Sossi, Vesna. / Investigation of texture quantification parameters for neurological PET image analysis. 2015 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2015. Institute of Electrical and Electronics Engineers Inc., 2016.
@inproceedings{bce49c7bd74a44e1b1ddb777fefa26a7,
title = "Investigation of texture quantification parameters for neurological PET image analysis",
abstract = "We investigate the correlation between the clinical severity of neurodegenerative disease and texture metrics (such as Haralick features) computed using PET images of the brain. Specifically, we explore how the parameters of feature computation - such as the region of interest definition method, and the direction and distance used for texture quantification - affect the correlation between texture-based image metrics and clinical disease severity. The analysis was based on an ongoing Parkinson's disease imaging study, with co-registered PET and MRI images, and tracer predominantly concentrated in the striatum. Disease duration was used as the primary clinical metric. It was found that the region of interest placement method substantially affected the correlation values. Significant correlation (p<0.01) was obtained when simple box-like regions were used instead of the anatomic MRI-based regions. The used direction affected the correlation values moderately, and distance did not have a pronounced effect. The results suggest that the Haralick features and other texture metrics that do not require kinetic modeling could be potentially used for the analysis of PET images for which the corresponding MRI data are not available. The results also show that the region of interest definition method and the direction along which metrics are computed may affect metric performance.",
author = "Klyuzhin, {Ivan S.} and Stephan Blinder and Rostom Mabrouk and Arman Rahmim and Vesna Sossi",
year = "2016",
month = "10",
day = "3",
doi = "10.1109/NSSMIC.2015.7582053",
language = "English (US)",
booktitle = "2015 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2015",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Investigation of texture quantification parameters for neurological PET image analysis

AU - Klyuzhin, Ivan S.

AU - Blinder, Stephan

AU - Mabrouk, Rostom

AU - Rahmim, Arman

AU - Sossi, Vesna

PY - 2016/10/3

Y1 - 2016/10/3

N2 - We investigate the correlation between the clinical severity of neurodegenerative disease and texture metrics (such as Haralick features) computed using PET images of the brain. Specifically, we explore how the parameters of feature computation - such as the region of interest definition method, and the direction and distance used for texture quantification - affect the correlation between texture-based image metrics and clinical disease severity. The analysis was based on an ongoing Parkinson's disease imaging study, with co-registered PET and MRI images, and tracer predominantly concentrated in the striatum. Disease duration was used as the primary clinical metric. It was found that the region of interest placement method substantially affected the correlation values. Significant correlation (p<0.01) was obtained when simple box-like regions were used instead of the anatomic MRI-based regions. The used direction affected the correlation values moderately, and distance did not have a pronounced effect. The results suggest that the Haralick features and other texture metrics that do not require kinetic modeling could be potentially used for the analysis of PET images for which the corresponding MRI data are not available. The results also show that the region of interest definition method and the direction along which metrics are computed may affect metric performance.

AB - We investigate the correlation between the clinical severity of neurodegenerative disease and texture metrics (such as Haralick features) computed using PET images of the brain. Specifically, we explore how the parameters of feature computation - such as the region of interest definition method, and the direction and distance used for texture quantification - affect the correlation between texture-based image metrics and clinical disease severity. The analysis was based on an ongoing Parkinson's disease imaging study, with co-registered PET and MRI images, and tracer predominantly concentrated in the striatum. Disease duration was used as the primary clinical metric. It was found that the region of interest placement method substantially affected the correlation values. Significant correlation (p<0.01) was obtained when simple box-like regions were used instead of the anatomic MRI-based regions. The used direction affected the correlation values moderately, and distance did not have a pronounced effect. The results suggest that the Haralick features and other texture metrics that do not require kinetic modeling could be potentially used for the analysis of PET images for which the corresponding MRI data are not available. The results also show that the region of interest definition method and the direction along which metrics are computed may affect metric performance.

UR - http://www.scopus.com/inward/record.url?scp=84994128685&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84994128685&partnerID=8YFLogxK

U2 - 10.1109/NSSMIC.2015.7582053

DO - 10.1109/NSSMIC.2015.7582053

M3 - Conference contribution

AN - SCOPUS:84994128685

BT - 2015 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2015

PB - Institute of Electrical and Electronics Engineers Inc.

ER -