18F-FDG PET/CT heterogeneity quantification through textural features in the era of harmonisation programs: a focus on lung cancer

  • Charline Lasnon
  • Mohamed Majdoub
  • Brice Lavigne
  • Pascal Do
  • Jeannick Madelaine
  • Dimitris Visvikis
  • Mathieu Hatt
  • Nicolas Aide
Original Article

DOI: 10.1007/s00259-016-3441-2

Cite this article as:
Lasnon, C., Majdoub, M., Lavigne, B. et al. Eur J Nucl Med Mol Imaging (2016) 43: 2324. doi:10.1007/s00259-016-3441-2

Abstract

Purpose

Quantification of tumour heterogeneity in PET images has recently gained interest, but has been shown to be dependent on image reconstruction. This study aimed to evaluate the impact of the EANM/EARL accreditation program on selected 18F-FDG heterogeneity metrics.

Methods

To carry out our study, we prospectively analysed 71 tumours in 60 biopsy-proven lung cancer patient acquisitions reconstructed with unfiltered point spread function (PSF) positron emission tomography (PET) images (optimised for diagnostic purposes), PSF-reconstructed images with a 7-mm Gaussian filter (PSF7) chosen to meet European Association of Nuclear Medicine (EANM) 1.0 harmonising standards, and EANM Research Ltd. (EARL)-compliant ordered subset expectation maximisation (OSEM) images. Delineation was performed with fuzzy locally adaptive Bayesian (FLAB) algorithm on PSF images and reported on PSF7 and OSEM ones, and with a 50 % standardised uptake values (SUV)max threshold (SUVmax50%) applied independently to each image. Robust and repeatable heterogeneity metrics including 1st-order [area under the curve of the cumulative histogram (CHAUC)], 2nd-order (entropy, correlation, and dissimilarity), and 3rd-order [high-intensity larger area emphasis (HILAE) and zone percentage (ZP)] textural features (TF) were statistically compared.

Results

Volumes obtained with SUVmax50% were significantly smaller than FLAB-derived ones, and were significantly smaller in PSF images compared to OSEM and PSF7 images. PSF-reconstructed images showed significantly higher SUVmax and SUVmean values, as well as heterogeneity for CHAUC, dissimilarity, correlation, and HILAE, and a wider range of heterogeneity values than OSEM images for most of the metrics considered, especially when analysing larger tumours. Histological subtypes had no impact on TF distribution. No significant difference was observed between any of the considered metrics (SUV or heterogeneity features) that we extracted from OSEM and PSF7 reconstructions. Furthermore, the distributions of TF for OSEM and PSF7 reconstructions according to tumour volumes were similar for all ranges of volumes.

Conclusion

PSF reconstruction with Gaussian filtering chosen to meet harmonising standards resulted in similar SUV values and heterogeneity information as compared to OSEM images, which validates its use within the harmonisation strategy context. However, unfiltered PSF-reconstructed images also showed higher heterogeneity according to some metrics, as well as a wider range of heterogeneity values than OSEM images for most of the metrics considered, especially when analysing larger tumours. This suggests that, whenever available, unfiltered PSF images should also be exploited to obtain the most discriminative quantitative heterogeneity features.

Keywords

FDG PET/CT Quantification Heterogeneity Harmonisation EARL accreditation program Lung cancer 

Supplementary material

259_2016_3441_Fig7_ESM.gif (467 kb)
Supplemental Fig. 1

Illustration of the in-house software used to define a 3-D box around the tumour, aiming at enclosing the tumour as well as excluding any nearby physiological or undesired uptake. For details see the Materials and Methods section. (GIF 466 kb)

259_2016_3441_MOESM1_ESM.tif (2.4 mb)
High-resolution Image (TIF 2443 kb)
259_2016_3441_Fig8_ESM.gif (35 kb)
Supplemental Fig. 2

Relationship between quantitative values extracted from PSF or PSF7 and OSEM images, assessed using Bland-Altman plots for SUVmax (a) and SUVmean (b) in tumour lesions. (GIF 34 kb)

259_2016_3441_MOESM2_ESM.tif (5.5 mb)
High-resolution Image (TIF 5659 kb)
259_2016_3441_Fig9_ESM.gif (84 kb)
Supplemental Fig. 3

Plots of TF features (CHAUC: area under the curve of the cumulative histogram; high-intensity larger area emphasis (HILAE); ZP: zone percentage) against tumours MATV for OSEM versus PSF reconstructions. (GIF 84 kb)

259_2016_3441_MOESM3_ESM.tif (17.9 mb)
High-resolution Image (TIF 18339 kb)
259_2016_3441_Fig10_ESM.gif (73 kb)
Supplemental Fig. 4

Plots of TF features (CHAUC: area under the curve of the cumulative histogram; high-intensity larger area emphasis (HILAE); ZP: zone percentage) against tumours MATV for OSEM versus PSF7 reconstructions. (GIF 72 kb)

259_2016_3441_MOESM4_ESM.tif (17.1 mb)
High-resolution Image (TIF 17535 kb)
259_2016_3441_Fig11_ESM.gif (40 kb)
Supplemental Fig. 5

Impact of the EARL harmonisation strategy on textural features using the FLAB algorithm independently on the three sets of reconstructions to delineate lesions. Textural features for the three reconstructions used. CHAUC: area under the curve of the cumulative histogram; high-intensity larger area emphasis (HILAE); ZP zone percentage. Data is shown as Tukey boxplots (lines displaying median, 25th and 75th percentiles; cross represents the mean value).*, **, and *** indicate two-tailed P < .05, P < .01, and P < .001, respectively. ns: non significant. Data represent the eight outliers (above the 90th percentile) observed with OSEM-PSF MATV and/or PSF7-PSF MATV when using a 50 % of SUVmax threshold. (GIF 39 kb)

259_2016_3441_MOESM5_ESM.tif (10.2 mb)
High-resolution Image (TIF 10459 kb)
259_2016_3441_Fig12_ESM.gif (22 kb)
Supplemental Fig. 6

PET transverse slice of a PSF-reconstructed image showing one of the largest differences in MATV (−95 %) between FLAB and 50 % SUVmax thresholding. (GIF 22 kb)

259_2016_3441_MOESM6_ESM.tif (1.8 mb)
High-resolution Image (TIF 1845 kb)

Copyright information

© Springer-Verlag Berlin Heidelberg 2016

Authors and Affiliations

  1. 1.Nuclear Medicine DepartmentUniversity HospitalCaenFrance
  2. 2.Biologie et Thérapies Innovantes des Cancers Localement Agressifs, Université de Caen Normandie, INSERMCaenFrance
  3. 3.Normandie UniversityCaenFrance
  4. 4.LaTIM, INSERM UMR 1101BrestFrance
  5. 5.Thoracic Oncology, François Baclesse Cancer CentreCaenFrance
  6. 6.Pulmonology DepartmentCaen University HospitalCaenFrance
  7. 7.Nuclear Medicine DepartmentCaen University HospitalCaenFrance
  8. 8.CHRU Morvan, INSERM UMR 1101, Laboratoire de Traitement de l’Information Medicale (LaTIM), Groupe ‘Imagerie multi-modalité quantitative pour le diagnostic et la thérapie’BrestFrance

Personalised recommendations