Karl Spuhler1, Mario Serrano-Sosa1, Renee Cattell1, Christine DeLorenzo1,2, Chuan Huang1,2,3. 1. Department of Biomedical Engineering, Stony Brook University, Stony Brook, NY, USA. 2. Department of Psychiatry, Stony Brook University, Stony Brook, NY, USA. 3. Department of Radiology, Stony Brook University, Stony Brook, NY, USA.
Abstract
PURPOSE: Positron emission tomography (PET) is an essential technique in many clinical applications that allows for quantitative imaging at the molecular level. This study aims to develop a denoising method using a novel dilated convolutional neural network (CNN) to recover full-count images from low-count images. METHODS: We adopted similar hierarchical structures as the conventional U-Net and incorporated dilated kernels in each convolution to allow the network to observe larger, more robust features within the image without the requirement of downsampling and upsampling internal representations. Our dNet was trained alongside a U-Net for comparison. Both models were evaluated using a leave-one-out cross-validation procedure on a dataset of 35 subjects (~3500 slabs), which were obtained from an ongoing 18 F-Fluorodeoxyglucose (FDG) study. Low-count PET data (10% count) were generated by randomly selecting one-tenth of all events in the associated listmode file. Analysis was done on the static image from the last 10 minutes of emission data. Both low-count PET and full-count PET were reconstructed using ordered subset expectation maximization (OSEM). Objective image quality metrics, including mean absolute percent error (MAPE), peak signal-to-noise ratio (PSNR), and structural similarity index metric (SSIM), were used to analyze the deep learning methods. Both deep learning methods were further compared to a traditional Gaussian filtering method. Further, region of interest (ROI) quantitative analysis was also used to compare U-Net and dNet architectures. RESULTS: Both the U-Net and our proposed network were successfully trained to synthesize full-count PET images from the generated low-count PET images. Compared to low-count PET and Gaussian filtering, both deep learning methods improved MAPE, PSNR, and SSIM. Our dNet also systematically outperformed U-Net on all three metrics (MAPE: 4.99 ± 0.68 vs 5.31 ± 0.76, P < 0.01; PSNR: 31.55 ± 1.31 dB vs 31.05 ± 1.39, P < 0.01; SSIM: 0.9513 ± 0.0154 vs 0.9447 ± 0.0178, P < 0.01). ROI quantification showed greater quantitative improvements using dNet over U-Net. CONCLUSION: This study proposed a novel approach of using dilated convolutions for recovering full-count PET images from low-count PET images.
PURPOSE: Positron emission tomography (PET) is an essential technique in many clinical applications that allows for quantitative imaging at the molecular level. This study aims to develop a denoising method using a novel dilated convolutional neural network (CNN) to recover full-count images from low-count images. METHODS: We adopted similar hierarchical structures as the conventional U-Net and incorporated dilated kernels in each convolution to allow the network to observe larger, more robust features within the image without the requirement of downsampling and upsampling internal representations. Our dNet was trained alongside a U-Net for comparison. Both models were evaluated using a leave-one-out cross-validation procedure on a dataset of 35 subjects (~3500 slabs), which were obtained from an ongoing 18 F-Fluorodeoxyglucose (FDG) study. Low-count PET data (10% count) were generated by randomly selecting one-tenth of all events in the associated listmode file. Analysis was done on the static image from the last 10 minutes of emission data. Both low-count PET and full-count PET were reconstructed using ordered subset expectation maximization (OSEM). Objective image quality metrics, including mean absolute percent error (MAPE), peak signal-to-noise ratio (PSNR), and structural similarity index metric (SSIM), were used to analyze the deep learning methods. Both deep learning methods were further compared to a traditional Gaussian filtering method. Further, region of interest (ROI) quantitative analysis was also used to compare U-Net and dNet architectures. RESULTS: Both the U-Net and our proposed network were successfully trained to synthesize full-count PET images from the generated low-count PET images. Compared to low-count PET and Gaussian filtering, both deep learning methods improved MAPE, PSNR, and SSIM. Our dNet also systematically outperformed U-Net on all three metrics (MAPE: 4.99 ± 0.68 vs 5.31 ± 0.76, P < 0.01; PSNR: 31.55 ± 1.31 dB vs 31.05 ± 1.39, P < 0.01; SSIM: 0.9513 ± 0.0154 vs 0.9447 ± 0.0178, P < 0.01). ROI quantification showed greater quantitative improvements using dNet over U-Net. CONCLUSION: This study proposed a novel approach of using dilated convolutions for recovering full-count PET images from low-count PET images.
Authors: Gerald Bonardel; Axel Dupont; Pierre Decazes; Mathieu Queneau; Romain Modzelewski; Jeremy Coulot; Nicolas Le Calvez; Sébastien Hapdey Journal: EJNMMI Phys Date: 2022-05-11
Authors: Mario Serrano-Sosa; Jared X Van Snellenberg; Jiayan Meng; Jacob R Luceno; Karl Spuhler; Jodi J Weinstein; Anissa Abi-Dargham; Mark Slifstein; Chuan Huang Journal: J Magn Reson Imaging Date: 2021-05-10 Impact factor: 5.119
Authors: Bart M de Vries; Sandeep S V Golla; Gerben J C Zwezerijnen; Otto S Hoekstra; Yvonne W S Jauw; Marc C Huisman; Guus A M S van Dongen; Willemien C Menke-van der Houven van Oordt; Josée J M Zijlstra-Baalbergen; Liesbet Mesotten; Ronald Boellaard; Maqsood Yaqub Journal: Diagnostics (Basel) Date: 2022-02-25