Unsupervised Deep Learning for Stroke Lesion Segmentation on Follow-up CT Based on Generative Adversarial Networks.
Henk van VoorstP R KonduriLaura M Van PoppelWouter van der SteenP M van der SluijsE M H SlotBart J EmmerWim H van ZwamYvo B W E M RoosCharles B L M MajoieGreg ZaharchukMatthan W A CaanHenk A Marqueringnull nullnull nullnull nullnull nullnull nullnull nullnull nullnull nullnull nullnull nullnull nullnull nullPublished in: AJNR. American journal of neuroradiology (2022)
An unsupervised generative adversarial network can be used to obtain automated infarct lesion segmentations with a moderate Dice similarity coefficient and good volumetric correspondence.
Keyphrases
- deep learning
- machine learning
- convolutional neural network
- artificial intelligence
- atrial fibrillation
- computed tomography
- image quality
- dual energy
- acute myocardial infarction
- high intensity
- contrast enhanced
- magnetic resonance imaging
- diffusion weighted imaging
- heart failure
- high throughput
- acute coronary syndrome
- cerebral ischemia