Robust Deep Learning-based Segmentation of Glioblastoma on Routine Clinical MRI Scans Using Sparsified Training.
Roelant Sjouke EijgelaarMartin VisserDomenique M J MüllerFrederik BarkhofHugo VrenkenMarcel van HerkLorenzo BelloMarco Conti NibaliMarco RossiTommaso SciortinoMitchel S BergerShawn Hervey-JumperBarbara KieselGeorg WidhalmJulia FurtnerPierre A J T RobeEmmanuel MandonnetPhilip C De Witt HamerJan C de MunckMarnix G WittePublished in: Radiology. Artificial intelligence (2020)
Accurate and automatic segmentation of glioblastoma on clinical scans is feasible using a model based on large, heterogeneous, and partially incomplete datasets. Sparsified training may boost the performance of a smaller model based on public and site-specific data.Supplemental material is available for this article.Published under a CC BY 4.0 license.
Keyphrases
- deep learning
- convolutional neural network
- computed tomography
- contrast enhanced
- artificial intelligence
- machine learning
- healthcare
- magnetic resonance imaging
- high resolution
- emergency department
- systematic review
- randomized controlled trial
- magnetic resonance
- rna seq
- mass spectrometry
- single cell
- data analysis
- diffusion weighted imaging