Deep learning-based image quality improvement of 18F-fluorodeoxyglucose positron emission tomography: a retrospective observational study

BackgroundDeep learning (DL)-based image quality improvement is a novel technique based on convolutional neural networks. The aim of this study was to compare the clinical value of F-18-fluorodeoxyglucose positron emission tomography (F-18-FDG PET) images obtained with the DL method with those obtained using a Gaussian filter.MethodsFifty patients with a mean age of 64.4 (range, 19-88) years who underwent F-18-FDG PET/CT between April 2019 and May 2019 were included in the study. PET images were obtained with the DL method in addition to conventional images reconstructed with three-dimensional time of flight-ordered subset expectation maximization and filtered with a Gaussian filter as a baseline for comparison. The reconstructed images were reviewed by two nuclear medicine physicians and scored from 1 (poor) to 5 (excellent) for tumor delineation, overall image quality, and image noise. For the semi-quantitative analysis, standardized uptake values in tumors and healthy tissues were compared between images obtained using the DL method and those obtained with a Gaussian filter.ResultsImages acquired using the DL method scored significantly higher for tumor delineation, overall image quality, and image noise compared to baseline (P <0.001). The Fleiss' kappa value for overall inter-reader agreement was 0.78. The standardized uptake values in tumor obtained by DL were significantly higher than those acquired using a Gaussian filter (P <0.001).ConclusionsDeep learning method improves the quality of PET images.

Automated summary

The study compared the clinical value of F-18-FDG PET images obtained with a deep learning method versus a Gaussian filter method, finding that deep learning significantly improved tumor delineation, overall image quality, and image noise ratings, as well as higher standardized uptake values in tumors compared to the Gaussian filter method.