Predicting peritumoral glioblastoma infiltration and subsequent recurrence using deep-learning-based analysis of multi-parametric magnetic resonance imaging.
Sunwoo KwakHamed AkbariJose A GarciaSuyash MohanYehuda DickerChiharu SakoYuji MatsumotoMacLean P NasrallahMahmoud ShalabyDonald M O'RourkeRussell Taki ShinoharaFang LiuChaitra BadveJill S Barnholtz-SloanAndrew E SloanMatthew D LeeRajan JainSantiago CepedaArnab ChakravartiJoshua David PalmerAdam P DickerGaurav ShuklaAdam E FlandersWenyin ShiGraeme F WoodworthChristos DavatzikosPublished in: Journal of medical imaging (Bellingham, Wash.) (2024)
The proposed model demonstrates that mpMRI analysis using deep learning can predict infiltration in the peri-tumoral brain region for GBM patients without needing to train a model using expert ROI drawings. Results for each institution demonstrate the model's generalizability and reproducibility.
Keyphrases
- deep learning
- magnetic resonance imaging
- end stage renal disease
- newly diagnosed
- machine learning
- artificial intelligence
- computed tomography
- prognostic factors
- magnetic resonance
- peritoneal dialysis
- convolutional neural network
- multiple sclerosis
- white matter
- patient reported outcomes
- mass spectrometry
- cerebral ischemia