Artificial intelligence-based PET denoising could allow a two-fold reduction in [<sup>18</sup>F]FDG PET acquisition time in digital PET/CT.

Weyts, Kathleen; Lasnon, Charline; Ciappuccini, Renaud; Lequesne, Justine; Corroyer-Dulmont, Aurélien; Quak, Elske; Clarisse, Bénédicte; Roussel, Laurent et al. · Eur J Nucl Med Mol Imaging · 2022

prospective_cohort · Level II

Where this comes from

Abstract

We investigated whether artificial intelligence (AI)-based denoising halves PET acquisition time in digital PET/CT. One hundred ninety-five patients referred for [<sup>18</sup>F]FDG PET/CT were prospectively included. Body PET acquisitions were performed in list mode. Original "PET90" (90 s/bed position) was compared to reconstructed ½-duration PET (45 s/bed position) with and without AI-denoising, "PET45AI and PET45". Denoising was performed by SubtlePET™ using deep convolutional neural networks. Visual global image quality (IQ) 3-point scores and lesion detectability were evaluated. Lesion maximal and peak standardized uptake values using lean body mass (SUL<sub>max</sub> and SUL<sub>peak</sub>), metabolic volumes (MV), and liver SUL<sub>mean</sub> were measured, including both standard and EARL<sub>1</sub> (European Association of Nuclear Medicine Research Ltd) compliant SUL. Lesion-to-liver SUL ratios (LLR) and liver coefficients of variation (CV<sub>liv</sub>) were calculated. PET45 showed mediocre IQ (scored poor in 8% and moderate in 68%) and lesion concordance rate with PET90 (88.7%). In PET45AI, IQ scores were similar to PET90 (P = 0.80), good in 92% and moderate in 8% for both. The lesion concordance rate between PET90 and PET45AI was 836/856 (97.7%), with 7 lesions (0.8%) only detected in PET90 and 13 (1.5%) exclusively in PET45AI. Lesion EARL<sub>1</sub> SUL<sub>peak</sub> was not significantly different between both PET (P = 0.09). Lesion standard SUL<sub>peak</sub>, standard and EARL1 SUL<sub>max</sub>, LLR and CV<sub>liv</sub> were lower in PET45AI than in PET90 (P < 0.0001), while lesion MV and liver SUL<sub>mean</sub> were higher (P < 0.0001). Good to excellent intraclass correlation coefficients (ICC) between PET90 and PET45AI were observed for lesion SUL and MV (ICC ≥ 0.97) and for liver SUL<sub>mean</sub> (ICC ≥ 0.87). AI allows [<sup>18</sup>F]FDG PET duration in digital PET/CT to be halved, while restoring degraded ½-duration PET image quality. Future multicentric studies, including other PET radiopharmaceuticals, are warranted.

Medical subject headings

Read the full text

View on PubMed · Publisher (DOI)