3D Convolutional Neural Network-Based Denoising of Low-Count Whole-Body (18)F-Fluorodeoxyglucose and (89)Zr-Rituximab PET Scans

Acquisition time and injected activity of (18)F-fluorodeoxyglucose ((18)F-FDG) PET should ideally be reduced. However, this decreases the signal-to-noise ratio (SNR), which impairs the diagnostic value of these PET scans. In addition, (89)Zr-antibody PET is known to have a low SNR. To improve the diagnostic value of these scans, a Convolutional Neural Network (CNN) denoising method is proposed. The aim of this study was therefore to develop CNNs to increase SNR for low-count (18)F-FDG and (89)Zr-antibody PET. Super-low-count, low-count and full-count (18)F-FDG PET scans from 60 primary lung cancer patients and full-count (89)Zr-rituximab PET scans from five patients with non-Hodgkin lymphoma were acquired. CNNs were built to capture the features and to denoise the PET scans. Additionally, Gaussian smoothing (GS) and Bilateral filtering (BF) were evaluated. The performance of the denoising approaches was assessed based on the tumour recovery coefficient (TRC), coefficient of variance (COV; level of noise), and a qualitative assessment by two nuclear medicine physicians. The CNNs had a higher TRC and comparable or lower COV to GS and BF and was also the preferred method of the two observers for both (18)F-FDG and (89)Zr-rituximab PET. The CNNs improved the SNR of low-count (18)F-FDG and (89)Zr-rituximab PET, with almost similar or better clinical performance than the full-count PET, respectively. Additionally, the CNNs showed better performance than GS and BF.