Partial volume correction strategies for quantitative FDG PET in oncology

• DOI: 10.1007/s00259-010-1472-7
• Publication Date: 2010-04-26T20:20:03Z
• Authors: Nikie J. Hoetjes; Floris H. P. van ...; Otto S. Hoekstra; Corneline J. Hoek...; Nanda C. Krak; Adriaan A. Lammer...; Ronald Boellaard
• Keywords: Partial volume; Standardized uptake value

Quantitative accuracy of positron emission tomography (PET) is affected by partial volume effects resulting in increased underestimation the standardized uptake value (SUV) with decreasing tumour volume. The purpose present study was to assess and precision different correction (PVC) methods. Three methods for PVC were evaluated: (1) inclusion point spread function (PSF) within reconstruction, (2) iterative deconvolution PET images (3) calculation spill-in spill-out factors based on masks. Simulations a mathematical phantom tumours sizes shapes. Phantom experiments performed 2-D mode using National Electrical Manufacturers Association (NEMA) NU2 image quality containing six differently sized spheres. Clinical studies (2-D mode) included test-retest consisting 10 patients stage IIIB IV non-small cell lung cancer response monitoring 15 female breast patients. In all or sphere volumes interest (VOI) generated VOI adaptive relative thresholds. provided similar results. All able accurately recover true SUV 10% spheres equal larger than 1 ml. Reconstruction-based recovery, however, up twofold better image-based showed that 5–80% depending size. Test-retest variability slightly worsened from 9.8 ± 6.5 without 10.8 7.9% PVC. Finally, resulted smaller responses, i.e. −30.5% −26.3% after first cycle treatment (p < 0.01). improves (clinical) significantly it has small, but significant effect observed responses. outperforms methods, requires dedicated reconstruction software. Image-based are good alternatives because their ease implementation performance clinical studies.