Segmentation of MRI head anatomy using deep volumetric networks and multiple spatial priors

@article{Hirsch2021SegmentationOM,
  title={Segmentation of MRI head anatomy using deep volumetric networks and multiple spatial priors},
  author={Lukas Hirsch and Yu Huang and Lucas C. Parra},
  journal={Journal of Medical Imaging},
  year={2021},
  volume={8},
  pages={034001 - 034001}
}
Abstract. Purpose: Conventional automated segmentation of the head anatomy in magnetic resonance images distinguishes different brain and nonbrain tissues based on image intensities and prior tissue probability maps (TPMs). This works well for normal head anatomies but fails in the presence of unexpected lesions. Deep convolutional neural networks (CNNs) leverage instead spatial patterns and can learn to segment lesions but often ignore prior probabilities. Approach: We add three sources of… 
2 Citations
Deep Learning Achieves Neuroradiologist-Level Performance in Detecting Hydrocephalus Requiring Treatment
TLDR
Hydcephalus cases requiring treatment can be detected automatically from MRI in a heterogeneous patient population based on quantitative characterization of brain anatomy with performance comparable to that of neuroradiologists.
Deep Learning Achieves Neuroradiologist-Level Performance in Detecting Hydrocephalus Requiring Treatment
TLDR
Hydcephalus cases requiring treatment can be detected automatically from MRI in a heterogeneous patient population based on quantitative characterization of brain anatomy with performance comparable to that of neuroradiologists.

References

SHOWING 1-10 OF 58 REFERENCES
Accurate and robust segmentation of neuroanatomy in T1‐weighted MRI by combining spatial priors with deep convolutional neural networks
TLDR
This work introduces a new highly accurate and versatile method based on 3D convolutional neural networks for the automatic segmentation of neuroanatomy in T1‐weighted MRI with new state‐of‐the‐art accuracies and a high robustness to outliers.
Anatomical Priors in Convolutional Networks for Unsupervised Biomedical Segmentation
TLDR
A generative probabilistic model is introduced that employs the learned prior through a convolutional neural network to compute segmentations in an unsupervised setting to facilitate CNN-based anatomical segmentation in a range of novel clinical problems, where few or no annotations are available and thus standard networks are not trainable.
Location Sensitive Deep Convolutional Neural Networks for Segmentation of White Matter Hyperintensities
TLDR
This paper applies and compares the proposed architectures for segmentation of white matter hyperintensities in brain MR images on a large dataset and observes that the CNNs that incorporate location information substantially outperform a conventional segmentation method with handcrafted features as well asCNNs that do not integrate location information.
Cardiac Segmentation from LGE MRI Using Deep Neural Network Incorporating Shape and Spatial Priors
TLDR
A new method, based on deep neural networks (DNN), for fully automatic segmentation from late gadolinium enhancement MRI, which outperformed the conventional schemes and obtained a Dice score of 0.758 for myocardial segmentation, which compares with 0.757 from the inter-observer variations.
Anatomically Constrained Neural Networks (ACNNs): Application to Cardiac Image Enhancement and Segmentation
TLDR
This work proposes a generic training strategy that incorporates anatomical prior knowledge into CNNs through a new regularisation model, which is trained end-to-end and demonstrates how the learnt deep models of 3-D shapes can be interpreted and used as biomarkers for classification of cardiac pathologies.
Automated sub‐cortical brain structure segmentation combining spatial and deep convolutional features
Anatomical Priors for Image Segmentation via Post-Processing with Denoising Autoencoders
TLDR
Post-DAE is proposed, a post-processing method based on denoising autoencoders (DAE) trained using only segmentation masks that can improve the quality of noisy and incorrect segmentation mask obtained with a variety of standard methods, by bringing them back to a feasible space, with almost no extra computational time.
...
...