Joint segmentation and quantification of oncological lesions in PET/CT: Preliminary evaluation on a zeolite phantom

Elisabetta De Bernardi, Chiara Soffientini, Felicia Zito, Giuseppe Baselli

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

4 Citations (Scopus)

Abstract

In this work we propose a strategy to jointly estimate activity and borders of oncological lesions in PET/CT. The starting step is constituted by a lesion contouring on PET image volume and by a Gaussian Mixture Model (GMM) clustering which individuates a set of regions in the lesion area (lesion, uncertainty, lesion spillout, organ). A maximum likelihood (AWOSEM) reconstruction step refines regions' borders and estimates a mean convergence activity for the lesion region. It applies a model of the scanner Point Spread Function (PSF) to recover blurring and it contemporaneously works on regional basis functions and single voxels. The area outside the four regions is frozen (i.e. not updated). The algorithm was validated on an anthropomorphic phantom in which lesions have been simulated with zeolites (clinoptilolite samples, volume 0.6 - 5.2 ml) loaded with 18F-FDG. Zeolite borders for ground truth definition were derived segmenting zeolites on coregistered CT images. For each zeolite, three different initial contouring were considered, corresponding to volumes about 100%, 60% and 140% of the true volume: the GMM clustering was able to robustly delineate regions independently from the initial contouring (variations 0.75) and in estimating zeolite activity (activity error <11% for zeolites >1ml). Suboptimal results were found for zeolites at the border of the axial FOV, since the PSF model, supposed invariant to axial shift, was inadequate at the axial borders. The proposed strategy appears promising and can be proposed as a general approach for a semi-automatic quantification and segmentation of lesions previously detected on standard clinical images. It will be further validated on data sets provided with a ground truth.

Original languageEnglish
Title of host publicationIEEE Nuclear Science Symposium Conference Record
Pages3306-3310
Number of pages5
DOIs
Publication statusPublished - 2012
Event2012 IEEE Nuclear Science Symposium and Medical Imaging Conference Record, NSS/MIC 2012 - Anaheim, CA, United States
Duration: Oct 29 2012Nov 3 2012

Other

Other2012 IEEE Nuclear Science Symposium and Medical Imaging Conference Record, NSS/MIC 2012
CountryUnited States
CityAnaheim, CA
Period10/29/1211/3/12

Fingerprint

Zeolites
lesions
Joints
borders
evaluation
zeolites
ground truth
Cluster Analysis
point spread functions
Fluorodeoxyglucose F18
Uncertainty
blurring
estimates
organs
scanners
estimating
shift

ASJC Scopus subject areas

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

Cite this

De Bernardi, E., Soffientini, C., Zito, F., & Baselli, G. (2012). Joint segmentation and quantification of oncological lesions in PET/CT: Preliminary evaluation on a zeolite phantom. In IEEE Nuclear Science Symposium Conference Record (pp. 3306-3310). [6551753] https://doi.org/10.1109/NSSMIC.2012.6551753

Joint segmentation and quantification of oncological lesions in PET/CT : Preliminary evaluation on a zeolite phantom. / De Bernardi, Elisabetta; Soffientini, Chiara; Zito, Felicia; Baselli, Giuseppe.

IEEE Nuclear Science Symposium Conference Record. 2012. p. 3306-3310 6551753.

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

De Bernardi, E, Soffientini, C, Zito, F & Baselli, G 2012, Joint segmentation and quantification of oncological lesions in PET/CT: Preliminary evaluation on a zeolite phantom. in IEEE Nuclear Science Symposium Conference Record., 6551753, pp. 3306-3310, 2012 IEEE Nuclear Science Symposium and Medical Imaging Conference Record, NSS/MIC 2012, Anaheim, CA, United States, 10/29/12. https://doi.org/10.1109/NSSMIC.2012.6551753
De Bernardi E, Soffientini C, Zito F, Baselli G. Joint segmentation and quantification of oncological lesions in PET/CT: Preliminary evaluation on a zeolite phantom. In IEEE Nuclear Science Symposium Conference Record. 2012. p. 3306-3310. 6551753 https://doi.org/10.1109/NSSMIC.2012.6551753
De Bernardi, Elisabetta ; Soffientini, Chiara ; Zito, Felicia ; Baselli, Giuseppe. / Joint segmentation and quantification of oncological lesions in PET/CT : Preliminary evaluation on a zeolite phantom. IEEE Nuclear Science Symposium Conference Record. 2012. pp. 3306-3310
@inproceedings{7590b72b33164526941441700e2d5df5,
title = "Joint segmentation and quantification of oncological lesions in PET/CT: Preliminary evaluation on a zeolite phantom",
abstract = "In this work we propose a strategy to jointly estimate activity and borders of oncological lesions in PET/CT. The starting step is constituted by a lesion contouring on PET image volume and by a Gaussian Mixture Model (GMM) clustering which individuates a set of regions in the lesion area (lesion, uncertainty, lesion spillout, organ). A maximum likelihood (AWOSEM) reconstruction step refines regions' borders and estimates a mean convergence activity for the lesion region. It applies a model of the scanner Point Spread Function (PSF) to recover blurring and it contemporaneously works on regional basis functions and single voxels. The area outside the four regions is frozen (i.e. not updated). The algorithm was validated on an anthropomorphic phantom in which lesions have been simulated with zeolites (clinoptilolite samples, volume 0.6 - 5.2 ml) loaded with 18F-FDG. Zeolite borders for ground truth definition were derived segmenting zeolites on coregistered CT images. For each zeolite, three different initial contouring were considered, corresponding to volumes about 100{\%}, 60{\%} and 140{\%} of the true volume: the GMM clustering was able to robustly delineate regions independently from the initial contouring (variations 0.75) and in estimating zeolite activity (activity error <11{\%} for zeolites >1ml). Suboptimal results were found for zeolites at the border of the axial FOV, since the PSF model, supposed invariant to axial shift, was inadequate at the axial borders. The proposed strategy appears promising and can be proposed as a general approach for a semi-automatic quantification and segmentation of lesions previously detected on standard clinical images. It will be further validated on data sets provided with a ground truth.",
author = "{De Bernardi}, Elisabetta and Chiara Soffientini and Felicia Zito and Giuseppe Baselli",
year = "2012",
doi = "10.1109/NSSMIC.2012.6551753",
language = "English",
isbn = "9781467320306",
pages = "3306--3310",
booktitle = "IEEE Nuclear Science Symposium Conference Record",

}

TY - GEN

T1 - Joint segmentation and quantification of oncological lesions in PET/CT

T2 - Preliminary evaluation on a zeolite phantom

AU - De Bernardi, Elisabetta

AU - Soffientini, Chiara

AU - Zito, Felicia

AU - Baselli, Giuseppe

PY - 2012

Y1 - 2012

N2 - In this work we propose a strategy to jointly estimate activity and borders of oncological lesions in PET/CT. The starting step is constituted by a lesion contouring on PET image volume and by a Gaussian Mixture Model (GMM) clustering which individuates a set of regions in the lesion area (lesion, uncertainty, lesion spillout, organ). A maximum likelihood (AWOSEM) reconstruction step refines regions' borders and estimates a mean convergence activity for the lesion region. It applies a model of the scanner Point Spread Function (PSF) to recover blurring and it contemporaneously works on regional basis functions and single voxels. The area outside the four regions is frozen (i.e. not updated). The algorithm was validated on an anthropomorphic phantom in which lesions have been simulated with zeolites (clinoptilolite samples, volume 0.6 - 5.2 ml) loaded with 18F-FDG. Zeolite borders for ground truth definition were derived segmenting zeolites on coregistered CT images. For each zeolite, three different initial contouring were considered, corresponding to volumes about 100%, 60% and 140% of the true volume: the GMM clustering was able to robustly delineate regions independently from the initial contouring (variations 0.75) and in estimating zeolite activity (activity error <11% for zeolites >1ml). Suboptimal results were found for zeolites at the border of the axial FOV, since the PSF model, supposed invariant to axial shift, was inadequate at the axial borders. The proposed strategy appears promising and can be proposed as a general approach for a semi-automatic quantification and segmentation of lesions previously detected on standard clinical images. It will be further validated on data sets provided with a ground truth.

AB - In this work we propose a strategy to jointly estimate activity and borders of oncological lesions in PET/CT. The starting step is constituted by a lesion contouring on PET image volume and by a Gaussian Mixture Model (GMM) clustering which individuates a set of regions in the lesion area (lesion, uncertainty, lesion spillout, organ). A maximum likelihood (AWOSEM) reconstruction step refines regions' borders and estimates a mean convergence activity for the lesion region. It applies a model of the scanner Point Spread Function (PSF) to recover blurring and it contemporaneously works on regional basis functions and single voxels. The area outside the four regions is frozen (i.e. not updated). The algorithm was validated on an anthropomorphic phantom in which lesions have been simulated with zeolites (clinoptilolite samples, volume 0.6 - 5.2 ml) loaded with 18F-FDG. Zeolite borders for ground truth definition were derived segmenting zeolites on coregistered CT images. For each zeolite, three different initial contouring were considered, corresponding to volumes about 100%, 60% and 140% of the true volume: the GMM clustering was able to robustly delineate regions independently from the initial contouring (variations 0.75) and in estimating zeolite activity (activity error <11% for zeolites >1ml). Suboptimal results were found for zeolites at the border of the axial FOV, since the PSF model, supposed invariant to axial shift, was inadequate at the axial borders. The proposed strategy appears promising and can be proposed as a general approach for a semi-automatic quantification and segmentation of lesions previously detected on standard clinical images. It will be further validated on data sets provided with a ground truth.

UR - http://www.scopus.com/inward/record.url?scp=84881570806&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84881570806&partnerID=8YFLogxK

U2 - 10.1109/NSSMIC.2012.6551753

DO - 10.1109/NSSMIC.2012.6551753

M3 - Conference contribution

AN - SCOPUS:84881570806

SN - 9781467320306

SP - 3306

EP - 3310

BT - IEEE Nuclear Science Symposium Conference Record

ER -