Improved unsupervised physics‐informed deep learning for intravoxel incoherent motion modeling and evaluation in pancreatic cancer patients

• Published in: Magnetic resonance in medicine Vol. 86; no. 4; pp. 2250 - 2265
• Main Authors: Kaandorp, Misha P. T., Barbieri, Sebastiano, Klaassen, Remy, Laarhoven, Hanneke W. M., Crezee, Hans, While, Peter T., Nederveen, Aart J., Gurney‐Champion, Oliver J.
• Format: Journal Article
• Language: English
• Published: United States Wiley Subscription Services, Inc 01.10.2021; John Wiley and Sons Inc
• Subjects: Adenocarcinoma; Algorithms; Artificial neural networks; Bayes Theorem; Bayesian analysis; Cancer; Chemoradiotherapy; Chemotherapy; Coefficient of variation; Computer architecture; Consistency; Deep Learning; deep neural network; Diffusion Magnetic Resonance Imaging; diffusion‐weighted magnetic resonance imaging; Humans; intravoxel incoherent motion; IVIM; Machine learning; Motion; Neural networks; Pancreatic cancer; Pancreatic Neoplasms - diagnostic imaging; Parameters; Patients; Performance evaluation; Physics; Radiation therapy; Reproducibility of Results; s—Computer Processing and Modeling; unsupervised physics‐informed deep learning; pancreatic cancer; IVIM; unsupervised physics-informed deep learning; intravoxel incoherent motion; deep neural network; diffusion-weighted magnetic resonance imaging

Purpose Earlier work showed that IVIM‐NETorig, an unsupervised physics‐informed deep neural network, was faster and more accurate than other state‐of‐the‐art intravoxel‐incoherent motion (IVIM) fitting approaches to diffusion‐weighted imaging (DWI). This study presents a substantially improved versi...