Regional 18F-fluoromisonidazole PET images generated from multiple advanced MR images using neural networks in glioblastoma

• Published in: Medicine (Baltimore) Vol. 101; no. 30; p. e29572
• Main Authors: Qin, Jianhua, Tang, Yu, Wang, Bao
• Format: Journal Article
• Language: English
• Published: Hagerstown, MD Lippincott Williams & Wilkins 29.07.2022
• Subjects: Bayes Theorem; Glioblastoma - diagnostic imaging; Glioblastoma - pathology; Humans; Misonidazole - analogs & derivatives; Neural Networks, Computer; Observational Study; Positron-Emission Tomography - methods; Radiopharmaceuticals; Tomography, X-Ray Computed; hypoxia; glioblastoma; F-FMISO; MRI; neural network
• ISSN: 1536-5964; 0025-7974; 1536-5964
• DOI: 10.1097/MD.0000000000029572

Generated 18F-fluoromisonidazole (18F-FMISO) positron emission tomography (PET) images for glioblastoma are highly sought after because 18F-FMISO can be radioactive, and the imaging procedure is not easy. This study aimed to explore the feasibility of using advanced magnetic resonance (MR) images to generate regional 18F-FMISO PET images and its predictive value for survival. Twelve kinds of advanced MR images of 28 patients from The Cancer Imaging Archive were processed. Voxel-by-voxel correlation analysis between 18F-FMISO images and advanced MR images was performed to select the MR images for generating regional 18F-FMISO images. Neural network algorithms provided by the MATLAB toolbox were used to generate regional 18F-FMISO images. The mean square error (MSE) was used to evaluate the regression effect. The prognostic value of generated 18F-FMISO images was evaluated by the Mantel-Cox test. A total of 299 831 voxels were extracted from the segmented regions of all patients. Eleven kinds of advanced MR images were selected to generate 18F-FMISO images. The best neural network algorithm was Bayesian regularization. The MSEs of the training, validation, and testing groups were 2.92E-2, 2.9E-2, and 2.92E-2, respectively. Both the maximum Tissue/Blood ratio (P = .017) and hypoxic volume (P = .023) of the generated images were predictive factors of overall survival, but only hypoxic volume (P = .029) was a predictive factor of progression-free survival. Multiple advanced MR images are feasible to generate qualified regional 18F-FMISO PET images using neural networks. The generated images also have predictive value in the prognostic evaluation of glioblastoma.