Automatic segmentation of the thalamus using a massively trained 3D convolutional neural network: higher sensitivity for the detection of reduced thalamus volume by improved inter-scanner stability.
Roland OpferJulia KrügerLothar SpiesAnn-Christin OstwaldtHagen H KitzlerSven SchipplingRalph BuchertPublished in: European radiology (2022)
• A three-dimensional convolutional neural network was trained for automatic segmentation of the thalamus with a heterogeneous sample of T1w-MRI from 1975 patients scanned on 170 different scanners. • The network provided high accuracy for thalamus segmentation with manual segmentation by experts as ground truth. • Inter-scanner variability of thalamus volume estimates across different MRI scanners was reduced by more than 50%, resulting in increased sensitivity for the detection of thalamus atrophy.
Keyphrases
- convolutional neural network
- deep learning
- deep brain stimulation
- end stage renal disease
- magnetic resonance imaging
- contrast enhanced
- newly diagnosed
- chronic kidney disease
- loop mediated isothermal amplification
- resistance training
- computed tomography
- real time pcr
- label free
- prognostic factors
- magnetic resonance