Direct inference of Patlak parametric images in whole-body PET/CT imaging using convolutional neural networks

• Published in: European journal of nuclear medicine and molecular imaging Vol. 49; no. 12; pp. 4048 - 4063
• Main Authors: Zaker, Neda, Haddad, Kamal, Faghihi, Reza, Arabi, Hossein, Zaidi, Habib
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.10.2022; Springer Nature B.V
• Subjects: Advanced Image Analyses (Radiomics and Artificial Intelligence); Artificial neural networks; Bias; Brain cancer; Cardiology; Computed tomography; Data acquisition; Datasets; Deep learning; Error reduction; Feasibility studies; Fluorodeoxyglucose F18; Humans; Image acquisition; Image Processing, Computer-Assisted - methods; Imaging; Inspection; Lesions; Liver cancer; Lung cancer; Medical imaging; Medicine; Medicine & Public Health; Neural networks; Neural Networks, Computer; Neuroimaging; Nuclear Medicine; Oncology; Original; Original Article; Orthopedics; Performance evaluation; Positron emission; Positron emission tomography; Positron Emission Tomography Computed Tomography - methods; Positron-Emission Tomography - methods; Radiology; Root-mean-square errors; Signal to noise ratio; Tomography; Training; Whole Body Imaging - methods; Deep learning; Dynamic PET imaging; Patlak analysis; Lesion detectability; Clinical oncology

Purpose This study proposed and investigated the feasibility of estimating Patlak-derived influx rate constant ( K i ) from standardized uptake value (SUV) and/or dynamic PET image series. Methods Whole-body 18 F-FDG dynamic PET images of 19 subjects consisting of 13 frames or passes were employed f...